Rename several directories to use MixedCase instead of lowercase.

They now mostly match the case used in #includes,
and they're consistent with the names of the .h files they contain.


git-svn-id: https://mpc-hc.svn.sourceforge.net/svnroot/mpc-hc/trunk@1648 10f7b99b-c216-0410-bff0-8a66a9350fd8

13 files changed, 7819 insertions, 0 deletions

diff --git a/src/filters/source/D2VSource/D2VSource.cpp b/src/filters/source/D2VSource/D2VSource.cpp
new file mode 100644
index 000000000..8984bf360
--- /dev/null
+++ b/src/filters/source/D2VSource/D2VSource.cpp
@@ -0,0 +1,283 @@
+/* 

+ *	Copyright (C) 2003-2006 Gabest

+ *

+ *  This Program is free software; you can redistribute it and/or modify

+ *  it under the terms of the GNU General Public License as published by

+ *  the Free Software Foundation; either version 2, or (at your option)

+ *  any later version.

+ *   

+ *  This Program is distributed in the hope that it will be useful,

+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of

+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the

+ *  GNU General Public License for more details.

+ *   

+ *  You should have received a copy of the GNU General Public License

+ *  along with GNU Make; see the file COPYING.  If not, write to

+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. 

+ *

+ */

+

+#include "stdafx.h"

+#include "D2VSource.h"

+#include "mpeg2dec.h"

+#include "../../../DSUtil/DSUtil.h"

+

+#ifdef REGISTER_FILTER

+

+const AMOVIESETUP_MEDIATYPE sudPinTypesOut[] =

+{

+	{&MEDIATYPE_Video, &MEDIASUBTYPE_YUY2}

+};

+

+const AMOVIESETUP_PIN sudOpPin[] =

+{

+	{L"Output", FALSE, TRUE, FALSE, FALSE, &CLSID_NULL, NULL, countof(sudPinTypesOut), sudPinTypesOut}

+};

+

+const AMOVIESETUP_FILTER sudFilter[] =

+{

+	{&__uuidof(CD2VSource), L"MPC - D2VSource", MERIT_NORMAL, countof(sudOpPin), sudOpPin, CLSID_LegacyAmFilterCategory}

+};

+

+CFactoryTemplate g_Templates[] =

+{

+	{sudFilter[0].strName, sudFilter[0].clsID, CreateInstance<CD2VSource>, NULL, &sudFilter[0]}

+};

+

+int g_cTemplates = countof(g_Templates);

+

+STDAPI DllRegisterServer()

+{

+	SetRegKeyValue(

+		_T("Media Type\\{e436eb83-524f-11ce-9f53-0020af0ba770}"), _T("{47CE0591-C4D5-4b41-BED7-28F59AD76228}"), 

+		_T("0"), _T("0,18,,4456443241564950726F6A65637446696C65")); // "DVD2AVIProjectFile"

+

+	SetRegKeyValue(

+		_T("Media Type\\{e436eb83-524f-11ce-9f53-0020af0ba770}"), _T("{47CE0591-C4D5-4b41-BED7-28F59AD76228}"), 

+		_T("Source Filter"), _T("{47CE0591-C4D5-4b41-BED7-28F59AD76228}"));

+

+	SetRegKeyValue(

+		_T("Media Type\\Extensions"), _T(".d2v"), 

+		_T("Source Filter"), _T("{47CE0591-C4D5-4b41-BED7-28F59AD76228}"));

+

+	return AMovieDllRegisterServer2(TRUE);

+}

+

+STDAPI DllUnregisterServer()

+{

+	DeleteRegKey(_T("Media Type\\{e436eb83-524f-11ce-9f53-0020af0ba770}"), _T("{47CE0591-C4D5-4b41-BED7-28F59AD76228}"));

+	DeleteRegKey(_T("Media Type\\Extensions"), _T(".d2v"));

+

+	return AMovieDllRegisterServer2(FALSE);

+}

+

+#include "../../FilterApp.h"

+

+CFilterApp theApp;

+

+#endif

+

+//

+// CD2VSource

+//

+

+CD2VSource::CD2VSource(LPUNKNOWN lpunk, HRESULT* phr)

+	: CBaseSource<CD2VStream>(NAME("CD2VSource"), lpunk, phr, __uuidof(this))

+{

+	if(phr) *phr = S_OK;

+}

+

+CD2VSource::~CD2VSource()

+{

+}

+

+//

+// CD2VStream

+//

+

+CD2VStream::CD2VStream(const WCHAR* fn, CSource* pParent, HRESULT* phr) 

+	: CBaseStream(NAME("D2VSourceStream"), pParent, phr)

+	, m_pFrameBuffer(NULL)

+{

+	CAutoLock cAutoLock(&m_cSharedState);

+

+	m_pDecoder.Attach(DNew CMPEG2Dec());

+	if(!m_pDecoder)

+	{

+		if(phr) *phr = E_OUTOFMEMORY;

+		return;

+	}

+

+	if(!m_pDecoder->Open(CString(fn), CMPEG2Dec::YUY2))

+	{

+		if(phr) *phr = E_FAIL;

+		return;

+	}

+

+	if(!m_pFrameBuffer.Allocate(m_pDecoder->Clip_Width*m_pDecoder->Clip_Height*4))

+	{

+		if(phr) *phr = E_OUTOFMEMORY;

+		return;

+	}

+

+	m_AvgTimePerFrame = 10000000000i64/m_pDecoder->VF_FrameRate;

+	m_rtDuration = m_rtStop = m_AvgTimePerFrame*m_pDecoder->VF_FrameLimit;

+

+	if(phr) *phr = m_rtDuration > 0 ? S_OK : E_FAIL;

+}

+

+CD2VStream::~CD2VStream()

+{

+	CAutoLock cAutoLock(&m_cSharedState);

+}

+

+HRESULT CD2VStream::DecideBufferSize(IMemAllocator* pAlloc, ALLOCATOR_PROPERTIES* pProperties)

+{

+//    CAutoLock cAutoLock(m_pFilter->pStateLock());

+

+    ASSERT(pAlloc);

+    ASSERT(pProperties);

+

+    HRESULT hr = NOERROR;

+

+	int w, h, bpp;

+	if(!GetDim(w, h, bpp))

+		return E_FAIL;

+

+	pProperties->cBuffers = 1;

+	pProperties->cbBuffer = w*h*bpp>>3;

+

+    ALLOCATOR_PROPERTIES Actual;

+    if(FAILED(hr = pAlloc->SetProperties(pProperties, &Actual))) return hr;

+

+    if(Actual.cbBuffer < pProperties->cbBuffer) return E_FAIL;

+    ASSERT(Actual.cBuffers == pProperties->cBuffers);

+

+    return NOERROR;

+}

+

+HRESULT CD2VStream::FillBuffer(IMediaSample* pSample, int nFrame, BYTE* pOut, long& len)

+{

+	if(!m_pDecoder)

+		return S_FALSE;

+

+	AM_MEDIA_TYPE* pmt;

+	if(SUCCEEDED(pSample->GetMediaType(&pmt)) && pmt)

+	{

+		CMediaType mt(*pmt);

+		SetMediaType(&mt);

+

+		DeleteMediaType(pmt);

+	}

+

+	int w, h, bpp;

+	if(!GetDim(w, h, bpp))

+		return S_FALSE;

+

+	BYTE* pIn = m_pFrameBuffer;

+

+	int pitchIn, pitchOut = 0;

+

+	pitchIn = m_pDecoder->Clip_Width*bpp>>3;

+	pitchOut = w*bpp>>3;

+

+	m_pDecoder->Decode(pIn, (unsigned long)(nFrame), pitchIn);

+

+	for(int y = 0, p = min(pitchIn, pitchOut); 

+		y < h; 

+		y++, pIn += pitchIn, pOut += pitchOut)

+	{

+		memcpy(pOut, pIn, p);

+	}

+

+	len = pitchOut*h;

+

+	return S_OK;

+}

+

+HRESULT CD2VStream::GetMediaType(int iPosition, CMediaType* pmt)

+{

+    CAutoLock cAutoLock(m_pFilter->pStateLock());

+

+    if(iPosition < 0) return E_INVALIDARG;

+    if(iPosition > 0) return VFW_S_NO_MORE_ITEMS;

+

+    pmt->SetType(&MEDIATYPE_Video);

+    pmt->SetSubtype(&MEDIASUBTYPE_YUY2);

+    pmt->SetFormatType(&FORMAT_VideoInfo);

+    pmt->SetTemporalCompression(FALSE);

+

+	VIDEOINFOHEADER* vih = (VIDEOINFOHEADER*)pmt->AllocFormatBuffer(sizeof(VIDEOINFOHEADER));

+	memset(vih, 0, sizeof(VIDEOINFOHEADER));

+	vih->AvgTimePerFrame = m_AvgTimePerFrame;

+	vih->bmiHeader.biSize = sizeof(vih->bmiHeader);

+	vih->bmiHeader.biWidth = m_pDecoder->Clip_Width;

+	vih->bmiHeader.biHeight = m_pDecoder->Clip_Height;

+	vih->bmiHeader.biPlanes = 1;

+	vih->bmiHeader.biBitCount = 16;

+	vih->bmiHeader.biCompression = '2YUY';

+	vih->bmiHeader.biSizeImage = vih->bmiHeader.biWidth*abs(vih->bmiHeader.biHeight)*vih->bmiHeader.biBitCount>>3;

+

+	pmt->SetSampleSize(vih->bmiHeader.biSizeImage);

+

+    return NOERROR;

+}

+

+HRESULT CD2VStream::SetMediaType(const CMediaType* pmt)

+{

+	if(m_pDecoder)

+	{

+		if(pmt->subtype == MEDIASUBTYPE_YUY2)

+			m_pDecoder->m_dstFormat = CMPEG2Dec::YUY2;

+		else

+			return E_FAIL;

+	}

+

+	return CSourceStream::SetMediaType(pmt);

+}

+

+HRESULT CD2VStream::CheckMediaType(const CMediaType* pmt)

+{

+	return pmt->majortype == MEDIATYPE_Video

+		&& pmt->subtype == MEDIASUBTYPE_YUY2

+		&& pmt->formattype == FORMAT_VideoInfo

+		? S_OK

+		: E_INVALIDARG;

+}

+

+STDMETHODIMP CD2VStream::Notify(IBaseFilter* pSender, Quality q)

+{

+	if(q.Late > 0 && q.Late < 100000000)

+	{

+		CAutoLock cAutoLockShared(&m_cSharedState);

+

+        m_rtSampleTime += (q.Late/m_AvgTimePerFrame)*m_AvgTimePerFrame;

+        m_rtPosition += (q.Late/m_AvgTimePerFrame)*m_AvgTimePerFrame;

+	}

+

+	return S_OK;

+}

+

+//

+

+bool CD2VStream::GetDim(int& w, int& h, int& bpp)

+{

+	if(m_mt.formattype == FORMAT_VideoInfo)

+	{

+		w = ((VIDEOINFOHEADER*)m_mt.pbFormat)->bmiHeader.biWidth;

+		h = abs(((VIDEOINFOHEADER*)m_mt.pbFormat)->bmiHeader.biHeight);

+		bpp = ((VIDEOINFOHEADER*)m_mt.pbFormat)->bmiHeader.biBitCount;

+	}

+	else if(m_mt.formattype == FORMAT_VideoInfo2)

+	{

+		w = ((VIDEOINFOHEADER2*)m_mt.pbFormat)->bmiHeader.biWidth;

+		h = abs(((VIDEOINFOHEADER2*)m_mt.pbFormat)->bmiHeader.biHeight);

+		bpp = ((VIDEOINFOHEADER2*)m_mt.pbFormat)->bmiHeader.biBitCount;

+	}

+	else

+	{

+		return(false);

+	}

+

+	return(true);

+}

diff --git a/src/filters/source/D2VSource/D2VSource.def b/src/filters/source/D2VSource/D2VSource.def
new file mode 100644
index 000000000..465cb0f61
--- /dev/null
+++ b/src/filters/source/D2VSource/D2VSource.def
@@ -0,0 +1,7 @@
+LIBRARY      "D2VSource.ax"

+

+EXPORTS

+	DllCanUnloadNow			PRIVATE

+	DllGetClassObject		PRIVATE

+	DllRegisterServer		PRIVATE

+	DllUnregisterServer		PRIVATE

diff --git a/src/filters/source/D2VSource/D2VSource.h b/src/filters/source/D2VSource/D2VSource.h
new file mode 100644
index 000000000..bf2e18f9a
--- /dev/null
+++ b/src/filters/source/D2VSource/D2VSource.h
@@ -0,0 +1,56 @@
+/* 

+ *	Copyright (C) 2003-2006 Gabest

+ *

+ *  This Program is free software; you can redistribute it and/or modify

+ *  it under the terms of the GNU General Public License as published by

+ *  the Free Software Foundation; either version 2, or (at your option)

+ *  any later version.

+ *   

+ *  This Program is distributed in the hope that it will be useful,

+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of

+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the

+ *  GNU General Public License for more details.

+ *   

+ *  You should have received a copy of the GNU General Public License

+ *  along with GNU Make; see the file COPYING.  If not, write to

+ *  the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. 

+ *

+ */

+

+#pragma once

+#include <atlbase.h>

+#include "../BaseSource/BaseSource.h"

+

+class CD2VStream;

+

+[uuid("47CE0591-C4D5-4b41-BED7-28F59AD76228")]

+class CD2VSource : public CBaseSource<CD2VStream>

+{

+public:

+	CD2VSource(LPUNKNOWN lpunk, HRESULT* phr);

+	virtual ~CD2VSource();

+};

+

+class CMPEG2Dec;

+

+class CD2VStream : public CBaseStream

+{

+private:

+	CAutoPtr<CMPEG2Dec> m_pDecoder;

+	CAutoVectorPtr<BYTE> m_pFrameBuffer;

+

+	bool GetDim(int& w, int& h, int& bpp);

+

+public:

+    CD2VStream(const WCHAR* fn, CSource* pParent, HRESULT* phr);

+	virtual ~CD2VStream();

+

+    HRESULT FillBuffer(IMediaSample* pSample, int nFrame, BYTE* pOut, long& len /*in+out*/);

+

+    HRESULT DecideBufferSize(IMemAllocator* pIMemAlloc, ALLOCATOR_PROPERTIES* pProperties);

+    HRESULT CheckMediaType(const CMediaType* pMediaType);

+    HRESULT GetMediaType(int iPosition, CMediaType* pmt);

+    HRESULT SetMediaType(const CMediaType* pmt);

+

+	STDMETHODIMP Notify(IBaseFilter* pSender, Quality q);

+};

diff --git a/src/filters/source/D2VSource/MPEG2Dec.cpp b/src/filters/source/D2VSource/MPEG2Dec.cpp
new file mode 100644
index 000000000..4be664c00
--- /dev/null
+++ b/src/filters/source/D2VSource/MPEG2Dec.cpp
@@ -0,0 +1,4490 @@
+#include "stdafx.h"

+#include "MPEG2Dec.h"

+

+int testint;

+

+struct CPU {

+	BOOL					mmx;

+	BOOL					_3dnow;

+	BOOL					ssemmx;

+	BOOL					ssefpu;

+} cpu;

+

+void CheckCPU()

+{

+	__asm

+	{

+		mov			eax, 1

+		cpuid

+		test		edx, 0x00800000		// STD MMX

+		jz			TEST_SSE

+		mov			[cpu.mmx], 1

+TEST_SSE:

+		test		edx, 0x02000000		// STD SSE

+		jz			TEST_3DNOW

+		mov			[cpu.ssemmx], 1

+		mov			[cpu.ssefpu], 1

+TEST_3DNOW:

+		mov			eax, 0x80000001

+		cpuid

+		test		edx, 0x80000000		// 3D NOW

+		jz			TEST_SSEMMX

+		mov			[cpu._3dnow], 1

+TEST_SSEMMX:

+		test		edx, 0x00400000		// SSE MMX

+		jz			TEST_END

+		mov			[cpu.ssemmx], 1

+TEST_END:

+	}

+}

+

+#pragma warning(disable:4799)	// no EMMS

+#pragma warning(disable:4731)	// ebp modified

+// idct

+extern "C" void __fastcall MMX_IDCT(short *block);

+extern "C" void __fastcall SSEMMX_IDCT(short *block);

+extern void Initialize_FPU_IDCT(void);

+extern void FPU_IDCT(short *block);

+extern void Initialize_REF_IDCT(void);

+extern void REF_IDCT(short *block);

+

+/* default intra quantization matrix */

+static unsigned char default_intra_quantizer_matrix[64] =

+{

+	8, 16, 19, 22, 26, 27, 29, 34,

+	16, 16, 22, 24, 27, 29, 34, 37,

+	19, 22, 26, 27, 29, 34, 34, 38,

+	22, 22, 26, 27, 29, 34, 37, 40,

+	22, 26, 27, 29, 32, 35, 40, 48,

+	26, 27, 29, 32, 35, 40, 48, 58,

+	26, 27, 29, 34, 38, 46, 56, 69,

+	27, 29, 35, 38, 46, 56, 69, 83

+};

+

+/* zig-zag and alternate scan patterns */

+static unsigned char scan[2][64] =

+{

+	{ /* Zig-Zag scan pattern  */

+		0,  1,  8, 16,  9,  2,  3, 10,

+	   17, 24, 32, 25, 18, 11,  4,  5,

+	   12, 19, 26, 33, 40, 48, 41, 34,

+	   27, 20, 13,  6,  7, 14, 21, 28,

+	   35, 42, 49, 56, 57, 50, 43, 36,

+	   29, 22, 15, 23, 30, 37, 44, 51,

+	   58, 59, 52, 45, 38, 31, 39, 46,

+	   53, 60, 61, 54, 47, 55, 62, 63

+	}

+	,

+	{ /* Alternate scan pattern */

+		0,  8, 16, 24,  1,  9,  2, 10,

+	   17, 25, 32, 40, 48, 56, 57, 49,

+	   41, 33, 26, 18,  3, 11, 4,  12,

+	   19, 27, 34, 42, 50, 58, 35, 43,

+	   51, 59, 20, 28,  5, 13,  6, 14,

+	   21, 29, 36, 44, 52, 60, 37, 45,

+	   53, 61, 22, 30,  7, 15, 23, 31,

+	   38, 46, 54, 62, 39, 47, 55, 63

+	}

+};

+

+/* non-linear quantization coefficient table */

+static unsigned char Non_Linear_quantizer_scale[32] =

+{

+	0, 1, 2, 3, 4, 5, 6, 7,

+	8, 10, 12, 14, 16, 18, 20, 22,

+	24, 28, 32, 36, 40, 44, 48, 52,

+	56, 64, 72, 80, 88, 96, 104, 112

+};

+

+#define ERROR_VALUE	(-1)

+

+typedef struct {

+	char run, level, len;

+} DCTtab;

+

+typedef struct {

+	char val, len;

+} VLCtab;

+

+/* Table B-10, motion_code, codes 0001 ... 01xx */

+static VLCtab MVtab0[8] =

+{

+	{ERROR_VALUE,0}, {3,3}, {2,2}, {2,2}, {1,1}, {1,1}, {1,1}, {1,1}

+};

+

+/* Table B-10, motion_code, codes 0000011 ... 000011x */

+static VLCtab MVtab1[8] =

+{

+	{ERROR_VALUE,0}, {ERROR_VALUE,0}, {ERROR_VALUE,0}, {7,6}, {6,6}, {5,6}, {4,5}, {4,5}

+};

+

+/* Table B-10, motion_code, codes 0000001100 ... 000001011x */

+static VLCtab MVtab2[12] =

+{

+	{16,9}, {15,9}, {14,9}, {13,9},

+	{12,9}, {11,9}, {10,8}, {10,8},

+	{9,8},  {9,8},  {8,8},  {8,8}

+};

+

+/* Table B-9, coded_block_pattern, codes 01000 ... 111xx */

+static VLCtab CBPtab0[32] =

+{

+	{ERROR_VALUE,0}, {ERROR_VALUE,0}, {ERROR_VALUE,0}, {ERROR_VALUE,0},

+	{ERROR_VALUE,0}, {ERROR_VALUE,0}, {ERROR_VALUE,0}, {ERROR_VALUE,0},

+	{62,5}, {2,5},  {61,5}, {1,5},  {56,5}, {52,5}, {44,5}, {28,5},

+	{40,5}, {20,5}, {48,5}, {12,5}, {32,4}, {32,4}, {16,4}, {16,4},

+	{8,4},  {8,4},  {4,4},  {4,4},  {60,3}, {60,3}, {60,3}, {60,3}

+};

+

+/* Table B-9, coded_block_pattern, codes 00000100 ... 001111xx */

+static VLCtab CBPtab1[64] =

+{

+	{ERROR_VALUE,0}, {ERROR_VALUE,0}, {ERROR_VALUE,0}, {ERROR_VALUE,0},

+	{58,8}, {54,8}, {46,8}, {30,8},

+	{57,8}, {53,8}, {45,8}, {29,8}, {38,8}, {26,8}, {37,8}, {25,8},

+	{43,8}, {23,8}, {51,8}, {15,8}, {42,8}, {22,8}, {50,8}, {14,8},

+	{41,8}, {21,8}, {49,8}, {13,8}, {35,8}, {19,8}, {11,8}, {7,8},

+	{34,7}, {34,7}, {18,7}, {18,7}, {10,7}, {10,7}, {6,7},  {6,7},

+	{33,7}, {33,7}, {17,7}, {17,7}, {9,7},  {9,7},  {5,7},  {5,7},

+	{63,6}, {63,6}, {63,6}, {63,6}, {3,6},  {3,6},  {3,6},  {3,6},

+	{36,6}, {36,6}, {36,6}, {36,6}, {24,6}, {24,6}, {24,6}, {24,6}

+};

+

+/* Table B-9, coded_block_pattern, codes 000000001 ... 000000111 */

+static VLCtab CBPtab2[8] =

+{

+	{ERROR_VALUE,0}, {0,9}, {39,9}, {27,9}, {59,9}, {55,9}, {47,9}, {31,9}

+};

+

+/* Table B-1, macroblock_address_increment, codes 00010 ... 011xx */

+static VLCtab MBAtab1[16] =

+{

+	{ERROR_VALUE,0}, {ERROR_VALUE,0}, {7,5}, {6,5}, {5,4}, {5,4}, {4,4},

+	{4,4}, {3,3}, {3,3}, {3,3}, {3,3}, {2,3}, {2,3}, {2,3}, {2,3}

+};

+

+/* Table B-1, macroblock_address_increment, codes 00000011000 ... 0000111xxxx */

+static VLCtab MBAtab2[104] =

+{

+	{33,11}, {32,11}, {31,11}, {30,11}, {29,11}, {28,11}, {27,11}, {26,11},

+	{25,11}, {24,11}, {23,11}, {22,11}, {21,10}, {21,10}, {20,10}, {20,10},

+	{19,10}, {19,10}, {18,10}, {18,10}, {17,10}, {17,10}, {16,10}, {16,10},

+	{15,8},  {15,8},  {15,8},  {15,8},  {15,8},  {15,8},  {15,8},  {15,8},

+	{14,8},  {14,8},  {14,8},  {14,8},  {14,8},  {14,8},  {14,8},  {14,8},

+	{13,8},  {13,8},  {13,8},  {13,8},  {13,8},  {13,8},  {13,8},  {13,8},

+	{12,8},  {12,8},  {12,8},  {12,8},  {12,8},  {12,8},  {12,8},  {12,8},

+	{11,8},  {11,8},  {11,8},  {11,8},  {11,8},  {11,8},  {11,8},  {11,8},

+	{10,8},  {10,8},  {10,8},  {10,8},  {10,8},  {10,8},  {10,8},  {10,8},

+	{9,7},   {9,7},   {9,7},   {9,7},   {9,7},   {9,7},   {9,7},   {9,7},

+	{9,7},   {9,7},   {9,7},   {9,7},   {9,7},   {9,7},   {9,7},   {9,7},

+	{8,7},   {8,7},   {8,7},   {8,7},   {8,7},   {8,7},   {8,7},   {8,7},

+	{8,7},   {8,7},   {8,7},   {8,7},   {8,7},   {8,7},   {8,7},   {8,7}

+};

+

+/* Table B-12, dct_dc_size_luminance, codes 00xxx ... 11110 */

+static VLCtab DClumtab0[32] =

+{

+	{1, 2}, {1, 2}, {1, 2}, {1, 2}, {1, 2}, {1, 2}, {1, 2}, {1, 2},

+	{2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2},

+	{0, 3}, {0, 3}, {0, 3}, {0, 3}, {3, 3}, {3, 3}, {3, 3}, {3, 3},

+	{4, 3}, {4, 3}, {4, 3}, {4, 3}, {5, 4}, {5, 4}, {6, 5}, {ERROR_VALUE, 0}

+};

+

+/* Table B-12, dct_dc_size_luminance, codes 111110xxx ... 111111111 */

+static VLCtab DClumtab1[16] =

+{

+	{7, 6}, {7, 6}, {7, 6}, {7, 6}, {7, 6}, {7, 6}, {7, 6}, {7, 6},

+	{8, 7}, {8, 7}, {8, 7}, {8, 7}, {9, 8}, {9, 8}, {10,9}, {11,9}

+};

+

+/* Table B-13, dct_dc_size_chrominance, codes 00xxx ... 11110 */

+static VLCtab DCchromtab0[32] =

+{

+	{0, 2}, {0, 2}, {0, 2}, {0, 2}, {0, 2}, {0, 2}, {0, 2}, {0, 2},

+	{1, 2}, {1, 2}, {1, 2}, {1, 2}, {1, 2}, {1, 2}, {1, 2}, {1, 2},

+	{2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2}, {2, 2},

+	{3, 3}, {3, 3}, {3, 3}, {3, 3}, {4, 4}, {4, 4}, {5, 5}, {ERROR_VALUE, 0}

+};

+

+/* Table B-13, dct_dc_size_chrominance, codes 111110xxxx ... 1111111111 */

+static VLCtab DCchromtab1[32] =

+{

+	{6, 6}, {6, 6}, {6, 6}, {6, 6}, {6, 6}, {6, 6}, {6, 6}, {6, 6},

+	{6, 6}, {6, 6}, {6, 6}, {6, 6}, {6, 6}, {6, 6}, {6, 6}, {6, 6},

+	{7, 7}, {7, 7}, {7, 7}, {7, 7}, {7, 7}, {7, 7}, {7, 7}, {7, 7},

+	{8, 8}, {8, 8}, {8, 8}, {8, 8}, {9, 9}, {9, 9}, {10,10}, {11,10}

+};

+

+/* Table B-14, DCT coefficients table zero,

+ * codes 0100 ... 1xxx (used for first (DC) coefficient)

+ */

+static DCTtab DCTtabfirst[12] =

+{

+	{0,2,4}, {2,1,4}, {1,1,3}, {1,1,3},

+	{0,1,1}, {0,1,1}, {0,1,1}, {0,1,1},

+	{0,1,1}, {0,1,1}, {0,1,1}, {0,1,1}

+};

+

+/* Table B-14, DCT coefficients table zero,

+ * codes 0100 ... 1xxx (used for all other coefficients)

+ */

+static DCTtab DCTtabnext[12] =

+{

+	{0,2,4},  {2,1,4},  {1,1,3},  {1,1,3},

+	{64,0,2}, {64,0,2}, {64,0,2}, {64,0,2}, /* EOB */

+	{0,1,2},  {0,1,2},  {0,1,2},  {0,1,2}

+};

+

+/* Table B-14, DCT coefficients table zero,

+ * codes 000001xx ... 00111xxx

+ */

+static DCTtab DCTtab0[60] =

+{

+	{65,0,6}, {65,0,6}, {65,0,6}, {65,0,6}, /* Escape */

+	{2,2,7}, {2,2,7}, {9,1,7}, {9,1,7},

+	{0,4,7}, {0,4,7}, {8,1,7}, {8,1,7},

+	{7,1,6}, {7,1,6}, {7,1,6}, {7,1,6},

+	{6,1,6}, {6,1,6}, {6,1,6}, {6,1,6},

+	{1,2,6}, {1,2,6}, {1,2,6}, {1,2,6},

+	{5,1,6}, {5,1,6}, {5,1,6}, {5,1,6},

+	{13,1,8}, {0,6,8}, {12,1,8}, {11,1,8},

+	{3,2,8}, {1,3,8}, {0,5,8}, {10,1,8},

+	{0,3,5}, {0,3,5}, {0,3,5}, {0,3,5},

+	{0,3,5}, {0,3,5}, {0,3,5}, {0,3,5},

+	{4,1,5}, {4,1,5}, {4,1,5}, {4,1,5},

+	{4,1,5}, {4,1,5}, {4,1,5}, {4,1,5},

+	{3,1,5}, {3,1,5}, {3,1,5}, {3,1,5},

+	{3,1,5}, {3,1,5}, {3,1,5}, {3,1,5}

+};

+

+/* Table B-15, DCT coefficients table one,

+ * codes 000001xx ... 11111111

+*/

+static DCTtab DCTtab0a[252] =

+{

+	{65,0,6}, {65,0,6}, {65,0,6}, {65,0,6}, /* Escape */

+	{7,1,7}, {7,1,7}, {8,1,7}, {8,1,7},

+	{6,1,7}, {6,1,7}, {2,2,7}, {2,2,7},

+	{0,7,6}, {0,7,6}, {0,7,6}, {0,7,6},

+	{0,6,6}, {0,6,6}, {0,6,6}, {0,6,6},

+	{4,1,6}, {4,1,6}, {4,1,6}, {4,1,6},

+	{5,1,6}, {5,1,6}, {5,1,6}, {5,1,6},

+	{1,5,8}, {11,1,8}, {0,11,8}, {0,10,8},

+	{13,1,8}, {12,1,8}, {3,2,8}, {1,4,8},

+	{2,1,5}, {2,1,5}, {2,1,5}, {2,1,5},

+	{2,1,5}, {2,1,5}, {2,1,5}, {2,1,5},

+	{1,2,5}, {1,2,5}, {1,2,5}, {1,2,5},

+	{1,2,5}, {1,2,5}, {1,2,5}, {1,2,5},

+	{3,1,5}, {3,1,5}, {3,1,5}, {3,1,5},

+	{3,1,5}, {3,1,5}, {3,1,5}, {3,1,5},

+	{1,1,3}, {1,1,3}, {1,1,3}, {1,1,3},

+	{1,1,3}, {1,1,3}, {1,1,3}, {1,1,3},

+	{1,1,3}, {1,1,3}, {1,1,3}, {1,1,3},

+	{1,1,3}, {1,1,3}, {1,1,3}, {1,1,3},

+	{1,1,3}, {1,1,3}, {1,1,3}, {1,1,3},

+	{1,1,3}, {1,1,3}, {1,1,3}, {1,1,3},

+	{1,1,3}, {1,1,3}, {1,1,3}, {1,1,3},

+	{1,1,3}, {1,1,3}, {1,1,3}, {1,1,3},

+	{64,0,4}, {64,0,4}, {64,0,4}, {64,0,4}, /* EOB */

+	{64,0,4}, {64,0,4}, {64,0,4}, {64,0,4},

+	{64,0,4}, {64,0,4}, {64,0,4}, {64,0,4},

+	{64,0,4}, {64,0,4}, {64,0,4}, {64,0,4},

+	{0,3,4}, {0,3,4}, {0,3,4}, {0,3,4},

+	{0,3,4}, {0,3,4}, {0,3,4}, {0,3,4},

+	{0,3,4}, {0,3,4}, {0,3,4}, {0,3,4},

+	{0,3,4}, {0,3,4}, {0,3,4}, {0,3,4},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,1,2}, {0,1,2}, {0,1,2}, {0,1,2},

+	{0,2,3}, {0,2,3}, {0,2,3}, {0,2,3},

+	{0,2,3}, {0,2,3}, {0,2,3}, {0,2,3},

+	{0,2,3}, {0,2,3}, {0,2,3}, {0,2,3},

+	{0,2,3}, {0,2,3}, {0,2,3}, {0,2,3},

+	{0,2,3}, {0,2,3}, {0,2,3}, {0,2,3},

+	{0,2,3}, {0,2,3}, {0,2,3}, {0,2,3},

+	{0,2,3}, {0,2,3}, {0,2,3}, {0,2,3},

+	{0,2,3}, {0,2,3}, {0,2,3}, {0,2,3},

+	{0,4,5}, {0,4,5}, {0,4,5}, {0,4,5},

+	{0,4,5}, {0,4,5}, {0,4,5}, {0,4,5},

+	{0,5,5}, {0,5,5}, {0,5,5}, {0,5,5},

+	{0,5,5}, {0,5,5}, {0,5,5}, {0,5,5},

+	{9,1,7}, {9,1,7}, {1,3,7}, {1,3,7},

+	{10,1,7}, {10,1,7}, {0,8,7}, {0,8,7},

+	{0,9,7}, {0,9,7}, {0,12,8}, {0,13,8},

+	{2,3,8}, {4,2,8}, {0,14,8}, {0,15,8}

+};

+

+/* Table B-14, DCT coefficients table zero,

+ * codes 0000001000 ... 0000001111

+ */

+static DCTtab DCTtab1[8] =

+{

+	{16,1,10}, {5,2,10}, {0,7,10}, {2,3,10},

+	{1,4,10}, {15,1,10}, {14,1,10}, {4,2,10}

+};

+

+/* Table B-15, DCT coefficients table one,

+ * codes 000000100x ... 000000111x

+ */

+static DCTtab DCTtab1a[8] =

+{

+	{5,2,9}, {5,2,9}, {14,1,9}, {14,1,9},

+	{2,4,10}, {16,1,10}, {15,1,9}, {15,1,9}

+};

+

+/* Table B-14/15, DCT coefficients table zero / one,

+ * codes 000000010000 ... 000000011111

+ */

+static DCTtab DCTtab2[16] =

+{

+	{0,11,12}, {8,2,12}, {4,3,12}, {0,10,12},

+	{2,4,12}, {7,2,12}, {21,1,12}, {20,1,12},

+	{0,9,12}, {19,1,12}, {18,1,12}, {1,5,12},

+	{3,3,12}, {0,8,12}, {6,2,12}, {17,1,12}

+};

+

+/* Table B-14/15, DCT coefficients table zero / one,

+ * codes 0000000010000 ... 0000000011111

+ */

+static DCTtab DCTtab3[16] =

+{

+	{10,2,13}, {9,2,13}, {5,3,13}, {3,4,13},

+	{2,5,13}, {1,7,13}, {1,6,13}, {0,15,13},

+	{0,14,13}, {0,13,13}, {0,12,13}, {26,1,13},

+	{25,1,13}, {24,1,13}, {23,1,13}, {22,1,13}

+};

+

+/* Table B-14/15, DCT coefficients table zero / one,

+ * codes 00000000010000 ... 00000000011111

+ */

+static DCTtab DCTtab4[16] =

+{

+	{0,31,14}, {0,30,14}, {0,29,14}, {0,28,14},

+	{0,27,14}, {0,26,14}, {0,25,14}, {0,24,14},

+	{0,23,14}, {0,22,14}, {0,21,14}, {0,20,14},

+	{0,19,14}, {0,18,14}, {0,17,14}, {0,16,14}

+};

+

+/* Table B-14/15, DCT coefficients table zero / one,

+ * codes 000000000010000 ... 000000000011111

+ */

+static DCTtab DCTtab5[16] =

+{

+	{0,40,15}, {0,39,15}, {0,38,15}, {0,37,15},

+	{0,36,15}, {0,35,15}, {0,34,15}, {0,33,15},

+	{0,32,15}, {1,14,15}, {1,13,15}, {1,12,15},

+	{1,11,15}, {1,10,15}, {1,9,15}, {1,8,15}

+};

+

+/* Table B-14/15, DCT coefficients table zero / one,

+ * codes 0000000000010000 ... 0000000000011111

+ */

+static DCTtab DCTtab6[16] =

+{

+	{1,18,16}, {1,17,16}, {1,16,16}, {1,15,16},

+	{6,3,16}, {16,2,16}, {15,2,16}, {14,2,16},

+	{13,2,16}, {12,2,16}, {11,2,16}, {31,1,16},

+	{30,1,16}, {29,1,16}, {28,1,16}, {27,1,16}

+};

+

+/* Table B-3, macroblock_type in P-pictures, codes 001..1xx */

+static VLCtab PMBtab0[8] =

+{

+	{ERROR_VALUE,0},

+	{MACROBLOCK_MOTION_FORWARD,3},

+	{MACROBLOCK_PATTERN,2}, {MACROBLOCK_PATTERN,2},

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_PATTERN,1}, 

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_PATTERN,1},

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_PATTERN,1}, 

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_PATTERN,1}

+};

+

+/* Table B-3, macroblock_type in P-pictures, codes 000001..00011x */

+static VLCtab PMBtab1[8] =

+{

+	{ERROR_VALUE,0},

+	{MACROBLOCK_QUANT|MACROBLOCK_INTRA,6},

+	{MACROBLOCK_QUANT|MACROBLOCK_PATTERN,5}, {MACROBLOCK_QUANT|MACROBLOCK_PATTERN,5},

+	{MACROBLOCK_QUANT|MACROBLOCK_MOTION_FORWARD|MACROBLOCK_PATTERN,5}, {MACROBLOCK_QUANT|MACROBLOCK_MOTION_FORWARD|MACROBLOCK_PATTERN,5},

+	{MACROBLOCK_INTRA,5}, {MACROBLOCK_INTRA,5}

+};

+

+/* Table B-4, macroblock_type in B-pictures, codes 0010..11xx */

+static VLCtab BMBtab0[16] =

+{

+	{ERROR_VALUE,0}, 

+	{ERROR_VALUE,0},

+	{MACROBLOCK_MOTION_FORWARD,4},

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_PATTERN,4},

+	{MACROBLOCK_MOTION_BACKWARD,3}, 

+	{MACROBLOCK_MOTION_BACKWARD,3},

+	{MACROBLOCK_MOTION_BACKWARD|MACROBLOCK_PATTERN,3}, 

+	{MACROBLOCK_MOTION_BACKWARD|MACROBLOCK_PATTERN,3},

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD,2}, 

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD,2},

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD,2}, 

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD,2},

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD|MACROBLOCK_PATTERN,2},

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD|MACROBLOCK_PATTERN,2},

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD|MACROBLOCK_PATTERN,2},

+	{MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD|MACROBLOCK_PATTERN,2}

+};

+

+/* Table B-4, macroblock_type in B-pictures, codes 000001..00011x */

+static VLCtab BMBtab1[8] =

+{

+	{ERROR_VALUE,0},

+	{MACROBLOCK_QUANT|MACROBLOCK_INTRA,6},

+	{MACROBLOCK_QUANT|MACROBLOCK_MOTION_BACKWARD|MACROBLOCK_PATTERN,6},

+	{MACROBLOCK_QUANT|MACROBLOCK_MOTION_FORWARD|MACROBLOCK_PATTERN,6},

+	{MACROBLOCK_QUANT|MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD|MACROBLOCK_PATTERN,5},

+	{MACROBLOCK_QUANT|MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD|MACROBLOCK_PATTERN,5},

+	{MACROBLOCK_INTRA,5}, 

+	{MACROBLOCK_INTRA,5}

+};

+

+//

+// getbit

+//

+

+void CMPEG2Dec::Initialize_Buffer()

+{

+	Rdptr = Rdbfr + BUFFER_SIZE;

+	Rdmax = Rdptr;

+

+	if (SystemStream_Flag)

+	{

+		if (Rdptr >= Rdmax)

+			Next_Packet();

+		CurrentBfr = *Rdptr++ << 24;

+

+		if (Rdptr >= Rdmax)

+			Next_Packet();

+		CurrentBfr += *Rdptr++ << 16;

+

+		if (Rdptr >= Rdmax)

+			Next_Packet();

+		CurrentBfr += *Rdptr++ << 8;

+

+		if (Rdptr >= Rdmax)

+			Next_Packet();

+		CurrentBfr += *Rdptr++;

+

+		Fill_Next();

+	}

+	else

+	{

+		Fill_Buffer();

+

+		CurrentBfr = (*Rdptr << 24) + (*(Rdptr+1) << 16) + (*(Rdptr+2) << 8) + *(Rdptr+3);

+		Rdptr += 4;

+

+		Fill_Next();

+	}

+

+	BitsLeft = 32;

+}

+

+unsigned int CMPEG2Dec::Get_Bits_All(unsigned int N)

+{

+	N -= BitsLeft;

+	Val = (CurrentBfr << (32 - BitsLeft)) >> (32 - BitsLeft);

+

+	if (N != 0)

+		Val = (Val << N) + (NextBfr >> (32 - N));

+

+	CurrentBfr = NextBfr;

+	BitsLeft = 32 - N;

+	Fill_Next();

+

+	return Val;

+}

+

+void CMPEG2Dec::Flush_Buffer_All(unsigned int N)

+{

+	CurrentBfr = NextBfr;

+	BitsLeft = BitsLeft + 32 - N;

+	Fill_Next();

+}

+

+void CMPEG2Dec::Next_Packet()

+{

+	unsigned int code, Packet_Length, Packet_Header_Length;

+

+	for (;;)

+	{

+		code = Get_Short();

+		code = (code<<16) + Get_Short();

+

+		// remove system layer byte stuffing

+		while ((code & 0xffffff00) != 0x00000100)

+			code = (code<<8) + Get_Byte();

+

+		switch (code)

+		{

+			case PACK_START_CODE:

+				Rdptr += 8;

+				break;

+

+			case VIDEO_ELEMENTARY_STREAM:   

+				Packet_Length = Get_Short();

+				Rdmax = Rdptr + Packet_Length;

+

+				code = Get_Byte();

+

+				if ((code & 0xc0)==0x80)

+				{

+					code = Get_Byte();

+					Packet_Header_Length = Get_Byte();

+

+					Rdptr += Packet_Header_Length;

+					return;

+				}

+				else

+					Rdptr += Packet_Length-1;

+				break;

+

+			default:

+				if (code>=SYSTEM_START_CODE)

+				{

+					code = Get_Short();

+					Rdptr += code;

+				}

+				break;

+		}

+	}

+}

+

+void CMPEG2Dec::Fill_Buffer()

+{

+	Read = _read(Infile[File_Flag], Rdbfr, BUFFER_SIZE);

+

+	if (Read < BUFFER_SIZE)

+		Next_File();

+

+	if (KeyOp_Flag && (Rdbfr[20] & 0x10))

+	{

+		BufferOp(Rdbfr, lfsr0, lfsr1);

+		Rdbfr[20] &= ~0x10;

+	}

+

+	Rdptr = Rdbfr;

+

+	if (SystemStream_Flag)

+		Rdmax -= BUFFER_SIZE;

+}

+

+void CMPEG2Dec::Next_File()

+{

+	if (File_Flag < File_Limit-1)

+		File_Flag ++;

+

+	_lseeki64(Infile[File_Flag], 0, SEEK_SET);

+	_read(Infile[File_Flag], Rdbfr + Read, BUFFER_SIZE - Read);

+}

+

+

+unsigned int CMPEG2Dec::Show_Bits(unsigned int N)

+{

+	if (N <= BitsLeft)

+		return (CurrentBfr << (32 - BitsLeft)) >> (32 - N);

+	else

+	{

+		N -= BitsLeft;

+		return (((CurrentBfr << (32 - BitsLeft)) >> (32 - BitsLeft)) << N) + (NextBfr >> (32 - N));

+	}

+}

+

+unsigned int CMPEG2Dec::Get_Bits(unsigned int N)

+{

+	if (N < BitsLeft)

+	{

+		Val = (CurrentBfr << (32 - BitsLeft)) >> (32 - N);

+		BitsLeft -= N;

+		return Val;

+	}

+	else

+		return Get_Bits_All(N);

+}

+

+void CMPEG2Dec::Flush_Buffer(unsigned int N)

+{

+	if (N < BitsLeft)

+		BitsLeft -= N;

+	else

+		Flush_Buffer_All(N);	

+}

+

+void CMPEG2Dec::Fill_Next()

+{

+	if (SystemStream_Flag && Rdptr>=Rdmax-4)

+	{

+		if (Rdptr >= Rdmax)

+			Next_Packet();

+		NextBfr = Get_Byte() << 24;

+

+		if (Rdptr >= Rdmax)

+			Next_Packet();

+		NextBfr += Get_Byte() << 16;

+

+		if (Rdptr >= Rdmax)

+			Next_Packet();

+		NextBfr += Get_Byte() << 8;

+

+		if (Rdptr >= Rdmax)

+			Next_Packet();

+		NextBfr += Get_Byte();

+	}

+	else if (Rdptr < Rdbfr+BUFFER_SIZE-4)

+	{

+		NextBfr = (*Rdptr << 24) + (*(Rdptr+1) << 16) + (*(Rdptr+2) << 8) + *(Rdptr+3);

+		Rdptr += 4;

+	}

+	else

+	{

+		if (Rdptr >= Rdbfr+BUFFER_SIZE)

+			Fill_Buffer();

+		NextBfr = *Rdptr++ << 24;

+

+		if (Rdptr >= Rdbfr+BUFFER_SIZE)

+			Fill_Buffer();

+		NextBfr += *Rdptr++ << 16;

+

+		if (Rdptr >= Rdbfr+BUFFER_SIZE)

+			Fill_Buffer();

+		NextBfr += *Rdptr++ << 8;

+

+		if (Rdptr >= Rdbfr+BUFFER_SIZE)

+			Fill_Buffer();

+		NextBfr += *Rdptr++;

+	}

+}

+

+unsigned int CMPEG2Dec::Get_Byte()

+{

+	while (Rdptr >= (Rdbfr + BUFFER_SIZE))

+	{

+		Read = _read(Infile[File_Flag], Rdbfr, BUFFER_SIZE);

+

+		if (Read < BUFFER_SIZE)

+			Next_File();

+

+		if (KeyOp_Flag && (Rdbfr[20] & 0x10))

+		{

+			BufferOp(Rdbfr, lfsr0, lfsr1);

+			Rdbfr[20] &= ~0x10;

+		}

+

+		Rdptr -= BUFFER_SIZE;

+		Rdmax -= BUFFER_SIZE;

+	}

+

+	return *Rdptr++;

+}

+

+unsigned int CMPEG2Dec::Get_Short()

+{

+	unsigned int i = Get_Byte();

+	return (i<<8) + Get_Byte();

+}

+

+void CMPEG2Dec::next_start_code()

+{

+	Flush_Buffer(BitsLeft & 7);

+

+	while (Show_Bits(24) != 1)

+		Flush_Buffer(8);

+}

+

+//

+// gethdr

+//

+

+int CMPEG2Dec::Get_Hdr()

+{

+	for (;;)

+	{

+		/* look for next_start_code */

+		next_start_code();

+

+		switch (Get_Bits(32))

+		{

+			case SEQUENCE_HEADER_CODE:

+				sequence_header();

+				break;

+

+			case GROUP_START_CODE:

+				group_of_pictures_header();

+				break;

+

+			case PICTURE_START_CODE:

+				picture_header();

+				return 1;

+		}

+	}

+}

+

+/* decode group of pictures header */

+/* ISO/IEC 13818-2 section 6.2.2.6 */

+void CMPEG2Dec::group_of_pictures_header()

+{

+	int gop_hour;

+	int gop_minute;

+	int gop_sec;

+	int gop_frame;

+

+	int drop_flag;

+	int closed_gop;

+	int broken_link;

+

+	drop_flag   = Get_Bits(1);

+	gop_hour    = Get_Bits(5);

+	gop_minute  = Get_Bits(6);

+	Flush_Buffer(1);	// marker bit

+	gop_sec     = Get_Bits(6);

+	gop_frame	= Get_Bits(6);

+	closed_gop  = Get_Bits(1);

+	broken_link = Get_Bits(1);

+

+	extension_and_user_data();

+}

+

+/* decode picture header */

+/* ISO/IEC 13818-2 section 6.2.3 */

+void CMPEG2Dec::picture_header()

+{

+	int vbv_delay;

+	int full_pel_forward_vector;

+	int forward_f_code;

+	int full_pel_backward_vector;

+	int backward_f_code;

+	int Extra_Information_Byte_Count;

+

+	temporal_reference  = Get_Bits(10);

+	picture_coding_type = Get_Bits(3);

+	vbv_delay           = Get_Bits(16);

+

+	if (picture_coding_type==P_TYPE || picture_coding_type==B_TYPE)

+	{

+		full_pel_forward_vector = Get_Bits(1);

+		forward_f_code = Get_Bits(3);

+	}

+

+	if (picture_coding_type==B_TYPE)

+	{

+		full_pel_backward_vector = Get_Bits(1);

+		backward_f_code = Get_Bits(3);

+	}

+

+	Extra_Information_Byte_Count = extra_bit_information();

+	extension_and_user_data();

+}

+

+/* decode sequence header */

+void CMPEG2Dec::sequence_header()

+{

+	int frame_rate_code;

+	int vbv_buffer_size;

+	int aspect_ratio_information;

+	int bit_rate_value;

+

+	int constrained_parameters_flag;

+	int i;

+

+	horizontal_size             = Get_Bits(12);

+	vertical_size               = Get_Bits(12);

+	aspect_ratio_information    = Get_Bits(4);

+	frame_rate_code             = Get_Bits(4);

+	bit_rate_value              = Get_Bits(18);

+	Flush_Buffer(1);	// marker bit

+	vbv_buffer_size             = Get_Bits(10);

+	constrained_parameters_flag = Get_Bits(1);

+

+	if (load_intra_quantizer_matrix = Get_Bits(1))

+	{

+		for (i=0; i<64; i++)

+			intra_quantizer_matrix[scan[ZIG_ZAG][i]] = Get_Bits(8);

+	}

+	else

+	{

+		for (i=0; i<64; i++)

+			intra_quantizer_matrix[i] = default_intra_quantizer_matrix[i];

+	}

+

+	if (load_non_intra_quantizer_matrix = Get_Bits(1))

+	{

+		for (i=0; i<64; i++)

+			non_intra_quantizer_matrix[scan[ZIG_ZAG][i]] = Get_Bits(8);

+	}

+	else

+	{

+		for (i=0; i<64; i++)

+			non_intra_quantizer_matrix[i] = 16;

+	}

+

+	/* copy luminance to chrominance matrices */

+	for (i=0; i<64; i++)

+	{

+		chroma_intra_quantizer_matrix[i] = intra_quantizer_matrix[i];

+		chroma_non_intra_quantizer_matrix[i] = non_intra_quantizer_matrix[i];

+	}

+	extension_and_user_data();

+}

+

+/* decode slice header */

+/* ISO/IEC 13818-2 section 6.2.4 */

+int CMPEG2Dec::slice_header()

+{

+	int slice_vertical_position_extension;

+	int quantizer_scale_code;

+	int slice_picture_id_enable = 0;

+	int slice_picture_id = 0;

+	int extra_information_slice = 0;

+

+	slice_vertical_position_extension = vertical_size>2800 ? Get_Bits(3) : 0;

+

+	quantizer_scale_code = Get_Bits(5);

+	quantizer_scale = q_scale_type ? Non_Linear_quantizer_scale[quantizer_scale_code] : quantizer_scale_code<<1;

+

+	/* slice_id introduced in March 1995 as part of the video corridendum

+	   (after the IS was drafted in November 1994) */

+	if (Get_Bits(1))

+	{

+		Get_Bits(1);	// intra slice

+

+		slice_picture_id_enable = Get_Bits(1);

+		slice_picture_id = Get_Bits(6);

+

+		extra_information_slice = extra_bit_information();

+	}

+

+	return slice_vertical_position_extension;

+}

+

+/* decode extension and user data */

+/* ISO/IEC 13818-2 section 6.2.2.2 */

+void CMPEG2Dec::extension_and_user_data()

+{

+	int code, ext_ID;

+

+	next_start_code();

+

+	while ((code = Show_Bits(32))==EXTENSION_START_CODE || code==USER_DATA_START_CODE)

+	{

+		if (code==EXTENSION_START_CODE)

+		{

+			Flush_Buffer(32);

+			ext_ID = Get_Bits(4);

+

+			switch (ext_ID)

+			{

+				case SEQUENCE_EXTENSION_ID:

+					sequence_extension();

+					break;

+

+				case SEQUENCE_DISPLAY_EXTENSION_ID:

+					sequence_display_extension();

+					break;

+

+				case QUANT_MATRIX_EXTENSION_ID:

+					quant_matrix_extension();

+					break;

+

+				case PICTURE_DISPLAY_EXTENSION_ID:

+					picture_display_extension();

+					break;

+

+				case PICTURE_CODING_EXTENSION_ID:

+					picture_coding_extension();

+					break;

+

+				case COPYRIGHT_EXTENSION_ID:

+					copyright_extension();

+					break;

+			}

+			next_start_code();

+		}

+		else

+		{

+			Flush_Buffer(32);

+			next_start_code();

+		}

+	}

+}

+

+/* decode sequence extension */

+/* ISO/IEC 13818-2 section 6.2.2.3 */

+void CMPEG2Dec::sequence_extension()

+{

+	int profile_and_level_indication;

+	int low_delay;

+	int frame_rate_extension_n;

+	int frame_rate_extension_d;

+

+	int horizontal_size_extension;

+	int vertical_size_extension;

+	int bit_rate_extension;

+	int vbv_buffer_size_extension;

+

+	profile_and_level_indication = Get_Bits(8);

+	progressive_sequence         = Get_Bits(1);

+	chroma_format                = Get_Bits(2);

+	horizontal_size_extension    = Get_Bits(2);

+	vertical_size_extension      = Get_Bits(2);

+	bit_rate_extension           = Get_Bits(12);

+	Flush_Buffer(1);	// marker bit

+	vbv_buffer_size_extension    = Get_Bits(8);

+	low_delay                    = Get_Bits(1);

+ 

+	frame_rate_extension_n       = Get_Bits(2);

+	frame_rate_extension_d       = Get_Bits(5);

+

+	horizontal_size = (horizontal_size_extension<<12) | (horizontal_size&0x0fff);

+	vertical_size = (vertical_size_extension<<12) | (vertical_size&0x0fff);

+}

+

+/* decode sequence display extension */

+void CMPEG2Dec::sequence_display_extension()

+{

+	int video_format;  

+	int color_description;

+	int color_primaries;

+	int transfer_characteristics;

+	int matrix_coefficients;

+	int display_horizontal_size;

+	int display_vertical_size;

+

+	video_format      = Get_Bits(3);

+	color_description = Get_Bits(1);

+

+	if (color_description)

+	{

+		color_primaries          = Get_Bits(8);

+		transfer_characteristics = Get_Bits(8);

+		matrix_coefficients      = Get_Bits(8);

+	}

+

+	display_horizontal_size = Get_Bits(14);

+	Flush_Buffer(1);	// marker bit

+	display_vertical_size   = Get_Bits(14);

+}

+

+/* decode quant matrix entension */

+/* ISO/IEC 13818-2 section 6.2.3.2 */

+void CMPEG2Dec::quant_matrix_extension()

+{

+	int i;

+

+	if (load_intra_quantizer_matrix = Get_Bits(1))

+		for (i=0; i<64; i++)

+			chroma_intra_quantizer_matrix[scan[ZIG_ZAG][i]]

+				= intra_quantizer_matrix[scan[ZIG_ZAG][i]] = Get_Bits(8);

+

+	if (load_non_intra_quantizer_matrix = Get_Bits(1))

+		for (i=0; i<64; i++)

+			chroma_non_intra_quantizer_matrix[scan[ZIG_ZAG][i]]

+				= non_intra_quantizer_matrix[scan[ZIG_ZAG][i]] = Get_Bits(8);

+

+	if (load_chroma_intra_quantizer_matrix = Get_Bits(1))

+		for (i=0; i<64; i++)

+			chroma_intra_quantizer_matrix[scan[ZIG_ZAG][i]] = Get_Bits(8);

+

+	if (load_chroma_non_intra_quantizer_matrix = Get_Bits(1))

+		for (i=0; i<64; i++)

+			chroma_non_intra_quantizer_matrix[scan[ZIG_ZAG][i]] = Get_Bits(8);

+}

+

+/* decode picture display extension */

+/* ISO/IEC 13818-2 section 6.2.3.3. */

+void CMPEG2Dec::picture_display_extension()

+{

+	int frame_center_horizontal_offset[3];

+	int frame_center_vertical_offset[3];

+

+	int i;

+	int number_of_frame_center_offsets;

+

+	/* based on ISO/IEC 13818-2 section 6.3.12 

+	   (November 1994) Picture display extensions */

+

+	/* derive number_of_frame_center_offsets */

+	if (progressive_sequence)

+	{

+		if (repeat_first_field)

+		{

+			if (top_field_first)

+				number_of_frame_center_offsets = 3;

+			else

+				number_of_frame_center_offsets = 2;

+		}

+		else

+			number_of_frame_center_offsets = 1;

+	}

+	else

+	{

+		if (picture_structure!=FRAME_PICTURE)

+			number_of_frame_center_offsets = 1;

+		else

+		{

+			if (repeat_first_field)

+				number_of_frame_center_offsets = 3;

+			else

+				number_of_frame_center_offsets = 2;

+		}

+	}

+

+	/* now parse */

+	for (i=0; i<number_of_frame_center_offsets; i++)

+	{

+		frame_center_horizontal_offset[i] = Get_Bits(16);

+		Flush_Buffer(1);	// marker bit

+

+		frame_center_vertical_offset[i] = Get_Bits(16);

+		Flush_Buffer(1);	// marker bit

+	}

+}

+

+/* decode picture coding extension */

+void CMPEG2Dec::picture_coding_extension()

+{

+	int chroma_420_type;

+	int composite_display_flag;

+	int v_axis;

+	int field_sequence;

+	int sub_carrier;

+	int burst_amplitude;

+	int sub_carrier_phase;

+

+	f_code[0][0] = Get_Bits(4);

+	f_code[0][1] = Get_Bits(4);

+	f_code[1][0] = Get_Bits(4);

+	f_code[1][1] = Get_Bits(4);

+

+	intra_dc_precision         = Get_Bits(2);

+	picture_structure          = Get_Bits(2);

+	top_field_first            = Get_Bits(1);

+	frame_pred_frame_dct       = Get_Bits(1);

+	concealment_motion_vectors = Get_Bits(1);

+	q_scale_type			   = Get_Bits(1);

+	intra_vlc_format           = Get_Bits(1);

+	alternate_scan			   = Get_Bits(1);

+	repeat_first_field         = Get_Bits(1);

+	chroma_420_type            = Get_Bits(1);

+	progressive_frame          = Get_Bits(1);

+	composite_display_flag     = Get_Bits(1);

+

+	pf_current = progressive_frame;

+

+	if (composite_display_flag)

+	{

+		v_axis            = Get_Bits(1);

+		field_sequence    = Get_Bits(3);

+		sub_carrier       = Get_Bits(1);

+		burst_amplitude   = Get_Bits(7);

+		sub_carrier_phase = Get_Bits(8);

+	}

+}

+

+/* decode extra bit information */

+/* ISO/IEC 13818-2 section 6.2.3.4. */

+int CMPEG2Dec::extra_bit_information()

+{

+	int Byte_Count = 0;

+

+	while (Get_Bits(1))

+	{

+		Flush_Buffer(8);

+		Byte_Count ++;

+	}

+

+	return(Byte_Count);

+}

+

+/* Copyright extension */

+/* ISO/IEC 13818-2 section 6.2.3.6. */

+/* (header added in November, 1994 to the IS document) */

+void CMPEG2Dec::copyright_extension()

+{

+	int copyright_flag;

+	int copyright_identifier;

+	int original_or_copy;

+	int copyright_number_1;

+	int copyright_number_2;

+	int copyright_number_3;

+

+	int reserved_data;

+

+	copyright_flag =       Get_Bits(1); 

+	copyright_identifier = Get_Bits(8);

+	original_or_copy =     Get_Bits(1);

+  

+	/* reserved */

+	reserved_data = Get_Bits(7);

+

+	Flush_Buffer(1); // marker bit

+	copyright_number_1 =   Get_Bits(20);

+	Flush_Buffer(1); // marker bit

+	copyright_number_2 =   Get_Bits(22);

+	Flush_Buffer(1); // marker bit

+	copyright_number_3 =   Get_Bits(22);

+}

+

+//

+// getpic

+//

+

+static const unsigned char cc_table[12] = {

+	0, 0, 0, 0, 1, 2, 1, 2, 1, 2, 1, 2

+};

+

+void CMPEG2Dec::Decode_Picture(int ref, unsigned char *dst, int pitch)

+{

+	if (picture_structure==FRAME_PICTURE && Second_Field)

+		Second_Field = 0;

+

+	if (picture_coding_type!=B_TYPE)

+	{

+		pf_forward = pf_backward;

+		pf_backward = pf_current;

+	}

+

+	Update_Picture_Buffers();

+

+	picture_data();

+

+	if (ref && (picture_structure==FRAME_PICTURE || Second_Field))

+	{

+		if (picture_coding_type==B_TYPE)

+			assembleFrame(auxframe, pf_current, dst, pitch);

+		else

+			assembleFrame(forward_reference_frame, pf_forward, dst, pitch);

+	}

+

+	if (picture_structure!=FRAME_PICTURE)

+		Second_Field = !Second_Field;

+}

+

+/* reuse old picture buffers as soon as they are no longer needed */

+void CMPEG2Dec::Update_Picture_Buffers()

+{                           

+	int cc;              /* color component index */

+	unsigned char *tmp;  /* temporary swap pointer */

+

+	for (cc=0; cc<3; cc++)

+	{

+		/* B pictures  do not need to be save for future reference */

+		if (picture_coding_type==B_TYPE)

+			current_frame[cc] = auxframe[cc];

+		else

+		{

+			if (!Second_Field)

+			{

+				/* only update at the beginning of the coded frame */

+				tmp = forward_reference_frame[cc];

+

+				/* the previously decoded reference frame is stored coincident with the 

+				   location where the backward reference frame is stored (backwards 

+				   prediction is not needed in P pictures) */

+				forward_reference_frame[cc] = backward_reference_frame[cc];

+

+				/* update pointer for potential future B pictures */

+				backward_reference_frame[cc] = tmp;

+			}

+

+			/* can erase over old backward reference frame since it is not used

+			   in a P picture, and since any subsequent B pictures will use the 

+			   previously decoded I or P frame as the backward_reference_frame */

+			current_frame[cc] = backward_reference_frame[cc];

+		}

+

+	    if (picture_structure==BOTTOM_FIELD)

+			current_frame[cc] += (cc==0) ? Coded_Picture_Width : Chroma_Width;

+	}

+}

+

+/* decode all macroblocks of the current picture */

+/* stages described in ISO/IEC 13818-2 section 7 */

+void CMPEG2Dec::picture_data()

+{

+	int MBAmax;

+

+	/* number of macroblocks per picture */

+	MBAmax = mb_width*mb_height;

+

+	if (picture_structure!=FRAME_PICTURE)

+		MBAmax>>=1;

+

+	for (;;)

+		if (slice(MBAmax)<0)

+			return;

+}

+

+/* decode all macroblocks of the current picture */

+/* ISO/IEC 13818-2 section 6.3.16 */

+/* return 0 : go to next slice */

+/* return -1: go to next picture */

+int CMPEG2Dec::slice(int MBAmax)

+{

+	int MBA = 0, MBAinc =0, macroblock_type, motion_type, dct_type, ret;

+	int dc_dct_pred[3], PMV[2][2][2], motion_vertical_field_select[2][2], dmvector[2];

+

+	if ((ret=start_of_slice(&MBA, &MBAinc, dc_dct_pred, PMV))!=1)

+		return ret;

+

+	for (;;)

+	{

+		/* this is how we properly exit out of picture */

+		if (MBA>=MBAmax) return -1;		// all macroblocks decoded

+

+		if (MBAinc==0)

+		{

+			if (!Show_Bits(23) || Fault_Flag)	// next_start_code or fault

+			{

+resync:

+				Fault_Flag = 0;

+				return 0;	// trigger: go to next slice

+			}

+			else /* neither next_start_code nor Fault_Flag */

+			{

+				/* decode macroblock address increment */

+				MBAinc = Get_macroblock_address_increment();

+				if (Fault_Flag) goto resync;

+			}

+		}

+

+		if (MBAinc==1) /* not skipped */

+		{

+			if (!decode_macroblock(&macroblock_type, &motion_type, &dct_type, PMV,

+				dc_dct_pred, motion_vertical_field_select, dmvector))

+				goto resync;

+		}

+		else /* MBAinc!=1: skipped macroblock */

+			/* ISO/IEC 13818-2 section 7.6.6 */

+			skipped_macroblock(dc_dct_pred, PMV, &motion_type, motion_vertical_field_select, &macroblock_type);

+

+		/* ISO/IEC 13818-2 section 7.6 */

+		motion_compensation(MBA, macroblock_type, motion_type, PMV,

+							motion_vertical_field_select, dmvector, dct_type);

+

+		/* advance to next macroblock */

+		MBA++; MBAinc--;

+

+		if (MBA>=MBAmax) return -1;		// all macroblocks decoded

+	}

+}

+

+/* ISO/IEC 13818-2 section 6.3.17.1: Macroblock modes */

+void CMPEG2Dec::macroblock_modes(int *pmacroblock_type, int *pmotion_type,

+							 int *pmotion_vector_count, int *pmv_format,

+							 int *pdmv, int *pmvscale, int *pdct_type)

+{

+	int macroblock_type, motion_type = 0, motion_vector_count;

+	int mv_format, dmv, mvscale, dct_type;

+

+	/* get macroblock_type */

+	macroblock_type = Get_macroblock_type();

+	if (Fault_Flag) return;

+

+	/* get frame/field motion type */

+	if (macroblock_type & (MACROBLOCK_MOTION_FORWARD|MACROBLOCK_MOTION_BACKWARD))

+	{

+		if (picture_structure==FRAME_PICTURE)

+			motion_type = frame_pred_frame_dct ? MC_FRAME : Get_Bits(2);

+		else

+			motion_type = Get_Bits(2);

+    }

+	else if ((macroblock_type & MACROBLOCK_INTRA) && concealment_motion_vectors)

+		motion_type = (picture_structure==FRAME_PICTURE) ? MC_FRAME : MC_FIELD;

+

+	/* derive motion_vector_count, mv_format and dmv, (table 6-17, 6-18) */

+	if (picture_structure==FRAME_PICTURE)

+	{

+		motion_vector_count = (motion_type==MC_FIELD) ? 2 : 1;

+		mv_format = (motion_type==MC_FRAME) ? MV_FRAME : MV_FIELD;

+	}

+	else

+	{

+		motion_vector_count = (motion_type==MC_16X8) ? 2 : 1;

+		mv_format = MV_FIELD;

+	}

+	

+	dmv = (motion_type==MC_DMV); /* dual prime */

+

+	/*

+	   field mv predictions in frame pictures have to be scaled

+	   ISO/IEC 13818-2 section 7.6.3.1 Decoding the motion vectors

+	*/

+	mvscale = (mv_format==MV_FIELD && picture_structure==FRAME_PICTURE);

+

+	/* get dct_type (frame DCT / field DCT) */

+	dct_type = (picture_structure==FRAME_PICTURE) && (!frame_pred_frame_dct)

+				&& (macroblock_type & (MACROBLOCK_PATTERN|MACROBLOCK_INTRA)) ? Get_Bits(1) : 0;

+

+	/* return values */

+	*pmacroblock_type = macroblock_type;

+	*pmotion_type = motion_type;

+	*pmotion_vector_count = motion_vector_count;

+	*pmv_format = mv_format;

+	*pdmv = dmv;

+	*pmvscale = mvscale;

+	*pdct_type = dct_type;

+}

+

+/* move/add 8x8-Block from block[comp] to backward_reference_frame */

+/* copy reconstructed 8x8 block from block[comp] to current_frame[]

+   ISO/IEC 13818-2 section 7.6.8: Adding prediction and coefficient data

+   This stage also embodies some of the operations implied by:

+   - ISO/IEC 13818-2 section 7.6.7: Combining predictions

+   - ISO/IEC 13818-2 section 6.1.3: Macroblock

+*/

+void CMPEG2Dec::Add_Block(int count, int bx, int by, int dct_type, int addflag)

+{

+	static const __int64 mmmask_128 = 0x0080008000800080;

+

+	int comp, cc, iincr, bxh, byh;

+	unsigned char *rfp;

+	short *Block_Ptr;

+

+	for (comp=0; comp<count; comp++)

+	{

+		Block_Ptr = block[comp];

+		cc = cc_table[comp];

+

+		bxh = bx; byh = by;

+

+		if (cc==0)

+		{

+			if (picture_structure==FRAME_PICTURE)

+			{

+				if (dct_type)

+				{

+					rfp = current_frame[0] + Coded_Picture_Width*(by+((comp&2)>>1)) + bx + ((comp&1)<<3);

+					iincr = Coded_Picture_Width<<1;

+				}

+				else

+				{

+					rfp = current_frame[0] + Coded_Picture_Width*(by+((comp&2)<<2)) + bx + ((comp&1)<<3);

+					iincr = Coded_Picture_Width;

+				}

+			}

+			else

+			{

+				rfp = current_frame[0] + (Coded_Picture_Width<<1)*(by+((comp&2)<<2)) + bx + ((comp&1)<<3);

+				iincr = Coded_Picture_Width<<1;

+			}

+		}

+		else

+		{

+			if (chroma_format!=CHROMA444)

+				bxh >>= 1;

+			if (chroma_format==CHROMA420)

+				byh >>= 1;

+

+			if (picture_structure==FRAME_PICTURE)

+			{

+				if (dct_type && chroma_format!=CHROMA420)

+				{

+					/* field DCT coding */

+					rfp = current_frame[cc] + Chroma_Width*(byh+((comp&2)>>1)) + bxh + (comp&8);

+					iincr = Chroma_Width<<1;

+				}

+				else

+				{

+					/* frame DCT coding */

+					rfp = current_frame[cc] + Chroma_Width*(byh+((comp&2)<<2)) + bxh + (comp&8);

+					iincr = Chroma_Width;

+				}

+			}

+			else

+			{

+				/* field picture */

+				rfp = current_frame[cc] + (Chroma_Width<<1)*(byh+((comp&2)<<2)) + bxh + (comp&8);

+				iincr = Chroma_Width<<1;

+			}

+		}

+

+		if (addflag)

+		{

+			__asm

+			{

+				pxor		mm0, mm0

+				mov			eax, [rfp]

+				mov			ebx, [Block_Ptr]

+				mov			edi, 8

+addon:

+				movq		mm2, [ebx+8]

+

+				movq		mm3, [eax]

+				movq		mm4, mm3

+

+				movq		mm1, [ebx]

+				punpckhbw	mm3, mm0

+

+				paddsw		mm3, mm2

+				packuswb	mm3, mm0

+

+				punpcklbw	mm4, mm0

+				psllq		mm3, 32

+

+				paddsw		mm4, mm1

+				packuswb	mm4, mm0

+

+				por			mm3, mm4			

+				add			ebx, 16

+

+				dec			edi

+				movq		[eax], mm3

+

+				add			eax, [iincr]

+				cmp			edi, 0x00

+				jg			addon

+			}

+		}

+		else

+		{

+			__asm

+			{

+				mov			eax, [rfp]

+				mov			ebx, [Block_Ptr]

+				mov			edi, 8

+

+				pxor		mm0, mm0

+				movq		mm7, [mmmask_128]

+addoff:

+				movq		mm3, [ebx+8]

+				movq		mm4, [ebx]

+

+				paddsw		mm3, mm7

+				paddsw		mm4, mm7

+

+				packuswb	mm3, mm0

+				packuswb	mm4, mm0

+

+				psllq		mm3, 32

+				por			mm3, mm4

+			

+				add			ebx, 16

+				dec			edi

+

+				movq		[eax], mm3

+

+				add			eax, [iincr]

+				cmp			edi, 0x00

+				jg			addoff

+			}

+		}

+	}

+}

+

+/* set scratch pad macroblock to zero */

+void CMPEG2Dec::Clear_Block(int count)

+{

+	int comp;

+	short *Block_Ptr;

+

+	for (comp=0; comp<count; comp++)

+	{

+		Block_Ptr = block[comp];

+

+		__asm

+		{

+			mov			eax, [Block_Ptr];

+			pxor		mm0, mm0;

+			movq		[eax+0 ], mm0;

+			movq		[eax+8 ], mm0;

+			movq		[eax+16], mm0;

+			movq		[eax+24], mm0;

+			movq		[eax+32], mm0;

+			movq		[eax+40], mm0;

+			movq		[eax+48], mm0;

+			movq		[eax+56], mm0;

+			movq		[eax+64], mm0;

+			movq		[eax+72], mm0;

+			movq		[eax+80], mm0;

+			movq		[eax+88], mm0;

+			movq		[eax+96], mm0;

+			movq		[eax+104],mm0;

+			movq		[eax+112],mm0;

+			movq		[eax+120],mm0;

+		}

+	}

+}

+

+/* ISO/IEC 13818-2 section 7.6 */

+void CMPEG2Dec::motion_compensation(int MBA, int macroblock_type, int motion_type, 

+								int PMV[2][2][2], int motion_vertical_field_select[2][2],

+								int dmvector[2], int dct_type)

+{

+	int bx, by;

+	int comp;

+

+	/* derive current macroblock position within picture */

+	/* ISO/IEC 13818-2 section 6.3.1.6 and 6.3.1.7 */

+	bx = 16*(MBA%mb_width);

+	by = 16*(MBA/mb_width);

+

+	/* motion compensation */

+	if (!(macroblock_type & MACROBLOCK_INTRA))

+		form_predictions(bx, by, macroblock_type, motion_type, PMV, 

+			motion_vertical_field_select, dmvector);

+

+	switch (IDCT_Flag)

+	{

+		case IDCT_MMX:

+			for (comp=0; comp<block_count; comp++)

+				MMX_IDCT(block[comp]);

+			break;

+

+		case IDCT_SSEMMX:

+			for (comp=0; comp<block_count; comp++)

+				SSEMMX_IDCT(block[comp]);

+			break;

+

+		case IDCT_FPU:

+			__asm emms;

+			for (comp=0; comp<block_count; comp++)

+				FPU_IDCT(block[comp]);

+			break;

+

+		case IDCT_REF:

+			__asm emms;

+			for (comp=0; comp<block_count; comp++)

+				REF_IDCT(block[comp]);

+			break;

+	}

+

+	Add_Block(block_count, bx, by, dct_type, (macroblock_type & MACROBLOCK_INTRA)==0);

+}

+

+/* ISO/IEC 13818-2 section 7.6.6 */

+void CMPEG2Dec::skipped_macroblock(int dc_dct_pred[3], int PMV[2][2][2], int *motion_type, 

+							   int motion_vertical_field_select[2][2], int *macroblock_type)

+{

+	Clear_Block(block_count);

+

+	/* reset intra_dc predictors */

+	/* ISO/IEC 13818-2 section 7.2.1: DC coefficients in intra blocks */

+	dc_dct_pred[0]=dc_dct_pred[1]=dc_dct_pred[2]=0;

+

+	/* reset motion vector predictors */

+	/* ISO/IEC 13818-2 section 7.6.3.4: Resetting motion vector predictors */

+	if (picture_coding_type==P_TYPE)

+		PMV[0][0][0]=PMV[0][0][1]=PMV[1][0][0]=PMV[1][0][1]=0;

+

+	/* derive motion_type */

+	if (picture_structure==FRAME_PICTURE)

+		*motion_type = MC_FRAME;

+	else

+	{

+		*motion_type = MC_FIELD;

+		motion_vertical_field_select[0][0] = motion_vertical_field_select[0][1] = 

+			(picture_structure==BOTTOM_FIELD);

+	}

+

+	/* clear MACROBLOCK_INTRA */

+	*macroblock_type&= ~MACROBLOCK_INTRA;

+}

+

+/* return==-1 means go to next picture */

+/* the expression "start of slice" is used throughout the normative

+   body of the MPEG specification */

+int CMPEG2Dec::start_of_slice(int *MBA, int *MBAinc,

+						  int dc_dct_pred[3], int PMV[2][2][2])

+{

+	unsigned int code;

+	int slice_vert_pos_ext;

+

+	next_start_code();

+	code = Get_Bits(32);

+

+	if (code<SLICE_START_CODE_MIN || code>SLICE_START_CODE_MAX)

+	{

+		// only slice headers are allowed in picture_data

+		Fault_Flag = 10;

+		return -1;

+	}

+

+	/* decode slice header (may change quantizer_scale) */

+	slice_vert_pos_ext = slice_header();

+

+	/* decode macroblock address increment */

+	*MBAinc = Get_macroblock_address_increment();

+	if (Fault_Flag) return -1;

+

+	/* set current location */

+	/* NOTE: the arithmetic used to derive macroblock_address below is

+	   equivalent to ISO/IEC 13818-2 section 6.3.17: Macroblock */

+	*MBA = ((slice_vert_pos_ext<<7) + (code&255) - 1)*mb_width + *MBAinc - 1;

+	*MBAinc = 1;	// first macroblock in slice: not skipped

+

+	/* reset all DC coefficient and motion vector predictors */

+	/* ISO/IEC 13818-2 section 7.2.1: DC coefficients in intra blocks */

+	dc_dct_pred[0]=dc_dct_pred[1]=dc_dct_pred[2]=0;

+  

+	/* ISO/IEC 13818-2 section 7.6.3.4: Resetting motion vector predictors */

+	PMV[0][0][0]=PMV[0][0][1]=PMV[1][0][0]=PMV[1][0][1]=0;

+	PMV[0][1][0]=PMV[0][1][1]=PMV[1][1][0]=PMV[1][1][1]=0;

+

+	/* successfull: trigger decode macroblocks in slice */

+	return 1;

+}

+

+/* ISO/IEC 13818-2 sections 7.2 through 7.5 */

+int CMPEG2Dec::decode_macroblock(int *macroblock_type, int *motion_type, int *dct_type,

+							 int PMV[2][2][2], int dc_dct_pred[3], 

+							 int motion_vertical_field_select[2][2], int dmvector[2])

+{

+	int quantizer_scale_code, comp, motion_vector_count, mv_format; 

+	int dmv, mvscale, coded_block_pattern;

+

+	/* ISO/IEC 13818-2 section 6.3.17.1: Macroblock modes */

+	macroblock_modes(macroblock_type, motion_type, &motion_vector_count, &mv_format,

+					 &dmv, &mvscale, dct_type);

+	if (Fault_Flag) return 0;	// trigger: go to next slice

+

+	if (*macroblock_type & MACROBLOCK_QUANT)

+	{

+		quantizer_scale_code = Get_Bits(5);

+

+		/* ISO/IEC 13818-2 section 7.4.2.2: Quantizer scale factor */

+		quantizer_scale = q_scale_type ?

+		Non_Linear_quantizer_scale[quantizer_scale_code] : (quantizer_scale_code << 1);

+	}

+

+	/* ISO/IEC 13818-2 section 6.3.17.2: Motion vectors */

+	/* decode forward motion vectors */

+	if ((*macroblock_type & MACROBLOCK_MOTION_FORWARD) 

+		|| ((*macroblock_type & MACROBLOCK_INTRA) && concealment_motion_vectors))

+		motion_vectors(PMV, dmvector, motion_vertical_field_select, 0,

+		motion_vector_count, mv_format, f_code[0][0]-1, f_code[0][1]-1, dmv, mvscale);

+	if (Fault_Flag) return 0;	// trigger: go to next slice

+

+	/* decode backward motion vectors */

+	if (*macroblock_type & MACROBLOCK_MOTION_BACKWARD)

+		motion_vectors(PMV, dmvector, motion_vertical_field_select, 1,

+		motion_vector_count,mv_format, f_code[1][0]-1, f_code[1][1]-1, 0, mvscale);

+	if (Fault_Flag) return 0;  // trigger: go to next slice

+

+	if ((*macroblock_type & MACROBLOCK_INTRA) && concealment_motion_vectors)

+		Flush_Buffer(1);	// marker bit

+

+	/* macroblock_pattern */

+	/* ISO/IEC 13818-2 section 6.3.17.4: Coded block pattern */

+	if (*macroblock_type & MACROBLOCK_PATTERN)

+	{

+		coded_block_pattern = Get_coded_block_pattern();

+

+		if (chroma_format==CHROMA422)

+			coded_block_pattern = (coded_block_pattern<<2) | Get_Bits(2);

+		else if (chroma_format==CHROMA444)

+			coded_block_pattern = (coded_block_pattern<<6) | Get_Bits(6);

+	}

+	else

+	    coded_block_pattern = (*macroblock_type & MACROBLOCK_INTRA) ? (1<<block_count)-1 : 0;

+

+	if (Fault_Flag) return 0;	// trigger: go to next slice

+

+	Clear_Block(block_count);

+

+	/* decode blocks */

+	for (comp=0; comp<block_count; comp++)

+	{

+		if (coded_block_pattern & (1<<(block_count-1-comp)))

+		{

+			if (*macroblock_type & MACROBLOCK_INTRA)

+				Decode_MPEG2_Intra_Block(comp, dc_dct_pred);

+			else

+				Decode_MPEG2_Non_Intra_Block(comp);

+			if (Fault_Flag) return 0;	// trigger: go to next slice

+		}

+	}

+

+	/* reset intra_dc predictors */

+	/* ISO/IEC 13818-2 section 7.2.1: DC coefficients in intra blocks */

+	if (!(*macroblock_type & MACROBLOCK_INTRA))

+		dc_dct_pred[0]=dc_dct_pred[1]=dc_dct_pred[2]=0;

+

+	/* reset motion vector predictors */

+	if ((*macroblock_type & MACROBLOCK_INTRA) && !concealment_motion_vectors)

+	{

+		/* intra mb without concealment motion vectors */

+		/* ISO/IEC 13818-2 section 7.6.3.4: Resetting motion vector predictors */

+		PMV[0][0][0]=PMV[0][0][1]=PMV[1][0][0]=PMV[1][0][1]=0;

+		PMV[0][1][0]=PMV[0][1][1]=PMV[1][1][0]=PMV[1][1][1]=0;

+	}

+

+	/* special "No_MC" macroblock_type case */

+	/* ISO/IEC 13818-2 section 7.6.3.5: Prediction in P pictures */

+	if ((picture_coding_type==P_TYPE) 

+		&& !(*macroblock_type & (MACROBLOCK_MOTION_FORWARD|MACROBLOCK_INTRA)))

+	{

+		/* non-intra mb without forward mv in a P picture */

+		/* ISO/IEC 13818-2 section 7.6.3.4: Resetting motion vector predictors */

+		PMV[0][0][0]=PMV[0][0][1]=PMV[1][0][0]=PMV[1][0][1]=0;

+

+		/* derive motion_type */

+		/* ISO/IEC 13818-2 section 6.3.17.1: Macroblock modes, frame_motion_type */

+		if (picture_structure==FRAME_PICTURE)

+			*motion_type = MC_FRAME;

+		else

+		{

+			*motion_type = MC_FIELD;

+			motion_vertical_field_select[0][0] = (picture_structure==BOTTOM_FIELD);

+		}

+	}

+	/* successfully decoded macroblock */

+	return 1 ;

+}

+

+/* decode one intra coded MPEG-2 block */

+void CMPEG2Dec::Decode_MPEG2_Intra_Block(int comp, int dc_dct_pred[])

+{

+	int val, i, j, sign, *qmat;

+	unsigned int code;

+	DCTtab *tab;

+	short *bp;

+

+	bp = block[comp];

+	qmat = (comp<4 || chroma_format==CHROMA420) 

+		? intra_quantizer_matrix : chroma_intra_quantizer_matrix;

+

+	/* ISO/IEC 13818-2 section 7.2.1: decode DC coefficients */

+	switch (cc_table[comp])

+	{

+		case 0:

+			val = (dc_dct_pred[0]+= Get_Luma_DC_dct_diff());

+			break;

+

+		case 1:

+			val = (dc_dct_pred[1]+= Get_Chroma_DC_dct_diff());

+			break;

+

+		case 2:

+			val = (dc_dct_pred[2]+= Get_Chroma_DC_dct_diff());

+			break;

+	}

+

+	bp[0] = val << (3-intra_dc_precision);

+

+	/* decode AC coefficients */

+	for (i=1; ; i++)

+	{

+		code = Show_Bits(16);

+

+		if (code>=16384 && !intra_vlc_format)

+			tab = &DCTtabnext[(code>>12)-4];

+		else if (code>=1024)

+		{

+			if (intra_vlc_format)

+				tab = &DCTtab0a[(code>>8)-4];

+			else

+				tab = &DCTtab0[(code>>8)-4];

+		}

+		else if (code>=512)

+		{

+			if (intra_vlc_format)

+				tab = &DCTtab1a[(code>>6)-8];

+			else

+				tab = &DCTtab1[(code>>6)-8];

+		}

+		else if (code>=256)

+			tab = &DCTtab2[(code>>4)-16];

+		else if (code>=128)

+			tab = &DCTtab3[(code>>3)-16];

+		else if (code>=64)

+			tab = &DCTtab4[(code>>2)-16];

+		else if (code>=32)

+			tab = &DCTtab5[(code>>1)-16];

+		else if (code>=16)

+			tab = &DCTtab6[code-16];

+		else

+		{

+			Fault_Flag = 1;

+			return;

+		}

+

+		Flush_Buffer(tab->len);

+

+		if (tab->run<64)

+		{

+			i+= tab->run;

+			val = tab->level;

+			sign = Get_Bits(1);

+		}

+		else if (tab->run==64) /* end_of_block */

+			return;

+		else /* escape */

+		{

+			i+= Get_Bits(6);

+			val = Get_Bits(12);

+

+			if (sign = (val>=2048))

+				val = 4096 - val;

+		}

+

+		j = scan[alternate_scan][i];

+

+		val = (val * quantizer_scale * qmat[j]) >> 4;

+		bp[j] = sign ? -val : val;

+	}

+}

+

+/* decode one non-intra coded MPEG-2 block */

+void CMPEG2Dec::Decode_MPEG2_Non_Intra_Block(int comp)

+{

+	int val, i, j, sign, *qmat;

+	unsigned int code;

+	DCTtab *tab;

+	short *bp;

+

+	bp = block[comp];

+	qmat = (comp<4 || chroma_format==CHROMA420) 

+		? non_intra_quantizer_matrix : chroma_non_intra_quantizer_matrix;

+

+	/* decode AC coefficients */

+	for (i=0; ; i++)

+	{

+		code = Show_Bits(16);

+

+		if (code>=16384)

+		{

+			if (i==0)

+				tab = &DCTtabfirst[(code>>12)-4];

+			else

+				tab = &DCTtabnext[(code>>12)-4];

+		}

+		else if (code>=1024)

+			tab = &DCTtab0[(code>>8)-4];

+		else if (code>=512)

+			tab = &DCTtab1[(code>>6)-8];

+		else if (code>=256)

+			tab = &DCTtab2[(code>>4)-16];

+		else if (code>=128)

+			tab = &DCTtab3[(code>>3)-16];

+		else if (code>=64)

+			tab = &DCTtab4[(code>>2)-16];

+		else if (code>=32)

+			tab = &DCTtab5[(code>>1)-16];

+		else if (code>=16)

+			tab = &DCTtab6[code-16];

+		else

+		{

+			Fault_Flag = 1;

+			return;

+		}

+

+		Flush_Buffer(tab->len);

+

+		if (tab->run<64)

+		{

+			i+= tab->run;

+			val = tab->level;

+			sign = Get_Bits(1);

+		}

+		else if (tab->run==64) /* end_of_block */

+			return;

+		else /* escape */

+		{

+			i+= Get_Bits(6);

+			val = Get_Bits(12);

+

+			if (sign = (val>=2048))

+				val = 4096 - val;

+		}

+

+		j = scan[alternate_scan][i];

+

+		val = (((val<<1)+1) * quantizer_scale * qmat[j]) >> 5;

+		bp[j] = sign ? -val : val;

+	}

+}

+

+int CMPEG2Dec::Get_macroblock_type()

+{

+	int macroblock_type;

+

+	switch (picture_coding_type)

+	{

+		case I_TYPE:

+			macroblock_type = Get_I_macroblock_type();

+			break;

+

+		case P_TYPE:

+			macroblock_type = Get_P_macroblock_type();

+			break;

+

+		case B_TYPE:

+			macroblock_type = Get_B_macroblock_type();

+			break;

+	}

+

+	return macroblock_type;

+}

+

+int CMPEG2Dec::Get_I_macroblock_type()

+{

+	if (Get_Bits(1))

+		return 1;

+

+	if (!Get_Bits(1))

+		Fault_Flag = 2;

+

+	return 17;

+}

+

+int CMPEG2Dec::Get_P_macroblock_type()

+{

+	int code;

+

+	if ((code = Show_Bits(6))>=8)

+	{

+		code >>= 3;

+		Flush_Buffer(PMBtab0[code].len);

+

+		return PMBtab0[code].val;

+	}

+

+	if (code==0)

+	{

+		Fault_Flag = 2;

+		return 0;

+	}

+

+	Flush_Buffer(PMBtab1[code].len);

+

+	return PMBtab1[code].val;

+}

+

+int CMPEG2Dec::Get_B_macroblock_type()

+{

+	int code;

+

+	if ((code = Show_Bits(6))>=8)

+	{

+		code >>= 2;

+		Flush_Buffer(BMBtab0[code].len);

+

+		return BMBtab0[code].val;

+	}

+

+	if (code==0)

+	{

+		Fault_Flag = 2;

+		return 0;

+	}

+

+	Flush_Buffer(BMBtab1[code].len);

+

+	return BMBtab1[code].val;

+}

+

+int CMPEG2Dec::Get_coded_block_pattern()

+{

+	int code;

+

+	if ((code = Show_Bits(9))>=128)

+	{

+		code >>= 4;

+		Flush_Buffer(CBPtab0[code].len);

+

+		return CBPtab0[code].val;

+	}

+

+	if (code>=8)

+	{

+		code >>= 1;

+		Flush_Buffer(CBPtab1[code].len);

+

+		return CBPtab1[code].val;

+	}

+

+	if (code<1)

+	{

+		Fault_Flag = 3;

+		return 0;

+	}

+

+	Flush_Buffer(CBPtab2[code].len);

+

+	return CBPtab2[code].val;

+}

+

+int CMPEG2Dec::Get_macroblock_address_increment()

+{

+	int code, val;

+

+	val = 0;

+

+	while ((code = Show_Bits(11))<24)

+	{

+		if (code!=15) /* if not macroblock_stuffing */

+		{

+			if (code==8) /* if macroblock_escape */

+				val+= 33;

+			else

+			{

+				Fault_Flag = 4;

+				return 1;

+			}

+		}

+		Flush_Buffer(11);

+	}

+

+	/* macroblock_address_increment == 1 */

+	/* ('1' is in the MSB position of the lookahead) */

+	if (code>=1024)

+	{

+		Flush_Buffer(1);

+		return val + 1;

+	}

+

+	/* codes 00010 ... 011xx */

+	if (code>=128)

+	{

+		/* remove leading zeros */

+		code >>= 6;

+		Flush_Buffer(MBAtab1[code].len);

+    

+		return val + MBAtab1[code].val;

+	}

+  

+	/* codes 00000011000 ... 0000111xxxx */

+	code-= 24; /* remove common base */

+	Flush_Buffer(MBAtab2[code].len);

+

+	return val + MBAtab2[code].val;

+}

+

+/*

+   parse VLC and perform dct_diff arithmetic.

+   MPEG-2:  ISO/IEC 13818-2 section 7.2.1 

+

+   Note: the arithmetic here is presented more elegantly than

+   the spec, yet the results, dct_diff, are the same.

+*/

+int CMPEG2Dec::Get_Luma_DC_dct_diff()

+{

+	int code, size, dct_diff;

+

+	/* decode length */

+	code = Show_Bits(5);

+

+	if (code<31)

+	{

+		size = DClumtab0[code].val;

+		Flush_Buffer(DClumtab0[code].len);

+	}

+	else

+	{

+		code = Show_Bits(9) - 0x1f0;

+		size = DClumtab1[code].val;

+		Flush_Buffer(DClumtab1[code].len);

+	}

+

+	if (size==0)

+		dct_diff = 0;

+	else

+	{

+		dct_diff = Get_Bits(size);

+

+		if ((dct_diff & (1<<(size-1)))==0)

+			dct_diff-= (1<<size) - 1;

+	}

+

+	return dct_diff;

+}

+

+int CMPEG2Dec::Get_Chroma_DC_dct_diff()

+{

+	int code, size, dct_diff;

+

+	/* decode length */

+	code = Show_Bits(5);

+

+	if (code<31)

+	{

+		size = DCchromtab0[code].val;

+		Flush_Buffer(DCchromtab0[code].len);

+	}

+	else

+	{

+		code = Show_Bits(10) - 0x3e0;

+		size = DCchromtab1[code].val;

+		Flush_Buffer(DCchromtab1[code].len);

+	}

+

+	if (size==0)

+		dct_diff = 0;

+	else

+	{

+		dct_diff = Get_Bits(size);

+

+		if ((dct_diff & (1<<(size-1)))==0)

+			dct_diff-= (1<<size) - 1;

+	}

+

+	return dct_diff;

+}

+

+/*

+static int currentfield;

+static unsigned char **predframe;

+static int DMV[2][2];

+static int stw;

+*/

+

+void CMPEG2Dec::form_predictions(int bx, int by, int macroblock_type, int motion_type,

+					  int PMV[2][2][2], int motion_vertical_field_select[2][2],

+					  int dmvector[2])

+{

+	int currentfield;

+	unsigned char **predframe;

+	int DMV[2][2];

+	int stw;

+	

+	stw = 0;

+

+	if ((macroblock_type & MACROBLOCK_MOTION_FORWARD) || (picture_coding_type==P_TYPE))

+	{

+		if (picture_structure==FRAME_PICTURE)

+		{

+			if ((motion_type==MC_FRAME) || !(macroblock_type & MACROBLOCK_MOTION_FORWARD))

+			{

+				/* frame-based prediction (broken into top and bottom halves

+				   for spatial scalability prediction purposes) */

+				form_prediction(forward_reference_frame, 0, current_frame, 0, Coded_Picture_Width, 

+					Coded_Picture_Width<<1, 16, 8, bx, by, PMV[0][0][0], PMV[0][0][1], stw);

+

+				form_prediction(forward_reference_frame, 1, current_frame, 1, Coded_Picture_Width, 

+					Coded_Picture_Width<<1, 16, 8, bx, by, PMV[0][0][0], PMV[0][0][1], stw);

+			}

+			else if (motion_type==MC_FIELD) /* field-based prediction */

+			{

+				/* top field prediction */

+				form_prediction(forward_reference_frame, motion_vertical_field_select[0][0], 

+					current_frame, 0, Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8,

+					bx, by>>1, PMV[0][0][0], PMV[0][0][1]>>1, stw);

+

+				/* bottom field prediction */

+				form_prediction(forward_reference_frame, motion_vertical_field_select[1][0], 

+					current_frame, 1, Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8,

+					bx, by>>1, PMV[1][0][0], PMV[1][0][1]>>1, stw);

+			}

+			else if (motion_type==MC_DMV) /* dual prime prediction */

+			{

+				/* calculate derived motion vectors */

+				Dual_Prime_Arithmetic(DMV, dmvector, PMV[0][0][0], PMV[0][0][1]>>1);

+

+				/* predict top field from top field */

+				form_prediction(forward_reference_frame, 0, current_frame, 0, 

+					Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8, bx, by>>1,

+					PMV[0][0][0], PMV[0][0][1]>>1, 0);

+

+				/* predict and add to top field from bottom field */

+				form_prediction(forward_reference_frame, 1, current_frame, 0,

+					Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8, bx, by>>1,

+					DMV[0][0], DMV[0][1], 1);

+

+				/* predict bottom field from bottom field */

+				form_prediction(forward_reference_frame, 1, current_frame, 1,

+					Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8, bx, by>>1,

+					PMV[0][0][0], PMV[0][0][1]>>1, 0);

+

+				/* predict and add to bottom field from top field */

+				form_prediction(forward_reference_frame, 0, current_frame, 1,

+					Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8, bx, by>>1,

+					DMV[1][0], DMV[1][1], 1);

+			}

+			else

+				Fault_Flag = 5;

+		}

+		else

+		{

+			/* field picture */

+			currentfield = (picture_structure==BOTTOM_FIELD);

+

+			/* determine which frame to use for prediction */

+			if (picture_coding_type==P_TYPE && Second_Field && currentfield!=motion_vertical_field_select[0][0])

+				predframe = backward_reference_frame;

+			else

+				predframe = forward_reference_frame;

+

+			if ((motion_type==MC_FIELD) || !(macroblock_type & MACROBLOCK_MOTION_FORWARD))

+			{

+				form_prediction(predframe, motion_vertical_field_select[0][0], current_frame, 0, 

+					Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 16, bx, by,

+					PMV[0][0][0], PMV[0][0][1], stw);

+			}

+			else if (motion_type==MC_16X8)

+			{

+				form_prediction(predframe, motion_vertical_field_select[0][0], current_frame, 0, 

+					Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8, bx, by,

+					PMV[0][0][0], PMV[0][0][1], stw);

+

+				if (picture_coding_type==P_TYPE && Second_Field && currentfield!=motion_vertical_field_select[1][0])

+					predframe = backward_reference_frame;

+				else

+					predframe = forward_reference_frame;

+

+				form_prediction(predframe, motion_vertical_field_select[1][0], current_frame, 

+					0, Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8, bx, by+8,

+					PMV[1][0][0], PMV[1][0][1], stw);

+			}

+			else if (motion_type==MC_DMV)

+			{

+				if (Second_Field)

+					predframe = backward_reference_frame;

+				else

+					predframe = forward_reference_frame;

+

+				/* calculate derived motion vectors */

+				Dual_Prime_Arithmetic(DMV, dmvector, PMV[0][0][0], PMV[0][0][1]);

+

+				/* predict from field of same parity */

+				form_prediction(forward_reference_frame, currentfield, current_frame, 0, 

+					Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 16, bx, by,

+					PMV[0][0][0], PMV[0][0][1], 0);

+

+				/* predict from field of opposite parity */

+				form_prediction(predframe, !currentfield, current_frame, 0,

+					Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 16, bx, by,

+					DMV[0][0], DMV[0][1], 1);

+			}

+			else

+				Fault_Flag = 5;

+		}

+

+		stw = 1;

+	}

+

+	if (macroblock_type & MACROBLOCK_MOTION_BACKWARD)

+	{

+		if (picture_structure==FRAME_PICTURE)

+		{

+			if (motion_type==MC_FRAME)

+			{

+				/* frame-based prediction */

+				form_prediction(backward_reference_frame, 0, current_frame, 0,

+					Coded_Picture_Width, Coded_Picture_Width<<1, 16, 8, bx, by,

+					PMV[0][1][0], PMV[0][1][1], stw);

+

+				form_prediction(backward_reference_frame, 1, current_frame, 1,

+					Coded_Picture_Width, Coded_Picture_Width<<1, 16, 8, bx, by,

+					PMV[0][1][0], PMV[0][1][1], stw);

+			}

+			else /* field-based prediction */

+			{

+				/* top field prediction */

+				form_prediction(backward_reference_frame, motion_vertical_field_select[0][1], 

+					current_frame, 0, Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8,

+					bx, by>>1, PMV[0][1][0], PMV[0][1][1]>>1, stw);

+

+				/* bottom field prediction */

+				form_prediction(backward_reference_frame, motion_vertical_field_select[1][1], 

+					current_frame, 1, Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8,

+					bx, by>>1, PMV[1][1][0], PMV[1][1][1]>>1, stw);

+			}

+		}

+		else

+		{

+			/* field picture */

+			if (motion_type==MC_FIELD)

+			{

+				/* field-based prediction */

+				form_prediction(backward_reference_frame, motion_vertical_field_select[0][1], 

+					current_frame, 0, Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 16,

+					bx, by, PMV[0][1][0], PMV[0][1][1], stw);

+			}

+			else if (motion_type==MC_16X8)

+			{

+				form_prediction(backward_reference_frame, motion_vertical_field_select[0][1],

+					current_frame, 0, Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8,

+					bx, by, PMV[0][1][0], PMV[0][1][1], stw);

+

+				form_prediction(backward_reference_frame, motion_vertical_field_select[1][1],

+					current_frame, 0, Coded_Picture_Width<<1, Coded_Picture_Width<<1, 16, 8,

+					bx, by+8, PMV[1][1][0], PMV[1][1][1], stw);

+			}

+			else

+				Fault_Flag = 5;

+		}

+	}

+}

+

+void CMPEG2Dec::form_prediction(unsigned char *src[], int sfield, unsigned char *dst[],

+							int dfield, int lx, int lx2, int w, int h, int x, int y,

+							int dx, int dy, int average_flag)

+{

+	form_component_prediction(src[0]+(sfield?lx2>>1:0), dst[0]+(dfield?lx2>>1:0),

+		lx, lx2, w, h, x, y, dx, dy, average_flag);

+

+	if (chroma_format!=CHROMA444)

+	{

+		lx>>=1; lx2>>=1; w>>=1; x>>=1; dx/=2;

+	}

+

+	if (chroma_format==CHROMA420)

+	{

+		h>>=1; y>>=1; dy/=2;

+	}

+

+	/* Cb */

+	form_component_prediction(src[1]+(sfield?lx2>>1:0), dst[1]+(dfield?lx2>>1:0),

+		lx, lx2, w, h, x, y, dx, dy, average_flag);

+

+	/* Cr */

+	form_component_prediction(src[2]+(sfield?lx2>>1:0), dst[2]+(dfield?lx2>>1:0),

+		lx, lx2, w, h, x, y, dx, dy, average_flag);

+}

+

+/* ISO/IEC 13818-2 section 7.6.4: Forming predictions */

+void CMPEG2Dec::form_component_prediction(unsigned char *src, unsigned char *dst,

+										  int lx, int lx2, int w, int h, int x, int y,

+										  int dx, int dy, int average_flag)

+{

+	static const __int64 mmmask_0001 = 0x0001000100010001;

+	static const __int64 mmmask_0002 = 0x0002000200020002;

+	static const __int64 mmmask_0003 = 0x0003000300030003;

+	static const __int64 mmmask_0006 = 0x0006000600060006;

+

+	unsigned char *s = src + lx * (y + (dy>>1)) + x + (dx>>1);

+	unsigned char *d = dst + lx * y + x;

+	int flag = (average_flag<<2) + ((dx & 1)<<1) + (dy & 1);

+

+	switch (flag)

+	{

+		case 0:

+			// d[i] = s[i];

+			__asm

+			{

+				mov			eax, [s]

+				mov			ebx, [d]

+				mov			esi, 0x00

+				mov			edi, [h]

+mc0:

+				movq		mm1, [eax+esi]

+				add			esi, 0x08

+				cmp			esi, [w]

+				movq		[ebx+esi-8], mm1

+				jl			mc0

+

+				add			eax, [lx2]

+				add			ebx, [lx2]

+				dec			edi

+				mov			esi, 0x00

+				cmp			edi, 0x00

+				jg			mc0

+			}

+			break;

+

+		case 1:

+			// d[i] = (s[i]+s[i+lx]+1)>>1;

+			__asm

+			{

+				pxor		mm0, mm0

+				movq		mm7, [mmmask_0001]

+				mov			eax, [s]

+				mov			ebx, [d]

+				mov			ecx, eax

+				add			ecx, [lx]

+				mov			esi, 0x00

+				mov			edi, [h]

+mc1:

+				movq		mm1, [eax+esi]

+				movq		mm2, [ecx+esi]

+

+				movq		mm3, mm1

+				movq		mm4, mm2

+

+				punpcklbw	mm1, mm0

+				punpckhbw	mm3, mm0

+				punpcklbw	mm2, mm0

+				punpckhbw	mm4, mm0

+

+				paddsw		mm1, mm2

+				paddsw		mm3, mm4

+

+				paddsw		mm1, mm7

+				paddsw		mm3, mm7

+

+				psrlw		mm1, 1

+				psrlw		mm3, 1

+

+				packuswb	mm1, mm0

+				packuswb	mm3, mm0

+

+				psllq		mm3, 32

+				por			mm1, mm3

+

+				add			esi, 0x08

+				cmp			esi, [w]

+				movq		[ebx+esi-8], mm1

+				jl			mc1

+

+				add			eax, [lx2]

+				add			ebx, [lx2]

+				add			ecx, [lx2]

+				dec			edi

+				mov			esi, 0x00

+				cmp			edi, 0x00

+				jg			mc1

+			}

+			break;

+

+		case 2:

+			// d[i] = (s[i]+s[i+1]+1)>>1;

+			__asm

+			{

+				pxor		mm0, mm0

+				movq		mm7, [mmmask_0001]

+				mov			eax, [s]

+				mov			ebx, [d]

+				mov			esi, 0x00

+				mov			edi, [h]

+mc2:

+				movq		mm1, [eax+esi]

+				movq		mm2, [eax+esi+1]

+

+				movq		mm3, mm1

+				movq		mm4, mm2

+

+				punpcklbw	mm1, mm0

+				punpckhbw	mm3, mm0

+

+				punpcklbw	mm2, mm0

+				punpckhbw	mm4, mm0

+

+				paddsw		mm1, mm2

+				paddsw		mm3, mm4

+

+				paddsw		mm1, mm7

+				paddsw		mm3, mm7

+

+				psrlw		mm1, 1

+				psrlw		mm3, 1

+

+				packuswb	mm1, mm0

+				packuswb	mm3, mm0

+

+				psllq		mm3, 32

+				por			mm1, mm3

+

+				add			esi, 0x08

+				cmp			esi, [w]

+				movq		[ebx+esi-8], mm1

+				jl			mc2

+

+				add			eax, [lx2]

+				add			ebx, [lx2]

+				dec			edi

+				mov			esi, 0x00

+				cmp			edi, 0x00

+				jg			mc2

+			}

+			break;

+

+		case 3:

+			// d[i] = (s[i]+s[i+1]+s[i+lx]+s[i+lx+1]+2)>>2;

+			__asm

+			{

+				pxor		mm0, mm0

+				movq		mm7, [mmmask_0002]

+				mov			eax, [s]

+				mov			ebx, [d]

+				mov			ecx, eax

+				add			ecx, [lx]

+				mov			esi, 0x00

+				mov			edi, [h]

+mc3:

+				movq		mm1, [eax+esi]

+				movq		mm2, [eax+esi+1]

+				movq		mm3, mm1

+				movq		mm4, mm2

+

+				punpcklbw	mm1, mm0

+				punpckhbw	mm3, mm0

+

+				punpcklbw	mm2, mm0

+				punpckhbw	mm4, mm0

+

+				paddsw		mm1, mm2

+				paddsw		mm3, mm4

+

+				movq		mm5, [ecx+esi]

+				paddsw		mm1, mm7

+

+				movq		mm6, [ecx+esi+1]

+				paddsw		mm3, mm7

+

+				movq		mm2, mm5

+				movq		mm4, mm6

+

+				punpcklbw	mm2, mm0

+				punpckhbw	mm5, mm0

+

+				punpcklbw	mm4, mm0

+				punpckhbw	mm6, mm0

+				

+				paddsw		mm2, mm4

+				paddsw		mm5, mm6

+

+				paddsw		mm1, mm2

+				paddsw		mm3, mm5

+

+				psrlw		mm1, 2

+				psrlw		mm3, 2

+

+				packuswb	mm1, mm0

+				packuswb	mm3, mm0

+

+				psllq		mm3, 32

+				por			mm1, mm3

+

+				add			esi, 0x08

+				cmp			esi, [w]

+				movq		[ebx+esi-8], mm1

+				jl			mc3

+

+				add			eax, [lx2]

+				add			ebx, [lx2]

+				add			ecx, [lx2]

+				dec			edi

+				mov			esi, 0x00

+				cmp			edi, 0x00

+				jg			mc3

+			}

+			break;

+

+		case 4:

+			// d[i] = (s[i]+d[i]+1)>>1;

+			__asm

+			{

+				pxor		mm0, mm0

+				movq		mm7, [mmmask_0001]

+				mov			eax, [s]

+				mov			ebx, [d]

+				mov			esi, 0x00

+				mov			edi, [h]

+mc4:

+				movq		mm1, [eax+esi]

+				movq		mm2, [ebx+esi]

+				movq		mm3, mm1

+				movq		mm4, mm2

+

+				punpcklbw	mm1, mm0

+				punpckhbw	mm3, mm0

+

+				punpcklbw	mm2, mm0

+				punpckhbw	mm4, mm0

+

+				paddsw		mm1, mm2

+				paddsw		mm3, mm4

+

+				paddsw		mm1, mm7

+				paddsw		mm3, mm7

+

+				psrlw		mm1, 1

+				psrlw		mm3, 1

+

+				packuswb	mm1, mm0

+				packuswb	mm3, mm0

+

+				psllq		mm3, 32

+				por			mm1, mm3

+

+				add			esi, 0x08

+				cmp			esi, [w]

+				movq		[ebx+esi-8], mm1

+				jl			mc4

+

+				add			eax, [lx2]

+				add			ebx, [lx2]

+				dec			edi

+				mov			esi, 0x00

+				cmp			edi, 0x00

+				jg			mc4

+			}

+			break;

+

+		case 5:

+			// d[i] = ((d[i]<<1) + s[i]+s[i+lx] + 3)>>2;

+			__asm

+			{

+				pxor		mm0, mm0

+				movq		mm7, [mmmask_0003]

+				mov			eax, [s]

+				mov			ebx, [d]

+				mov			ecx, eax

+				add			ecx, [lx]

+				mov			esi, 0x00

+				mov			edi, [h]

+mc5:

+				movq		mm1, [eax+esi]

+				movq		mm2, [ecx+esi]

+				movq		mm3, mm1

+				movq		mm4, mm2

+

+				punpcklbw	mm1, mm0

+				punpckhbw	mm3, mm0

+

+				punpcklbw	mm2, mm0

+				punpckhbw	mm4, mm0

+

+				paddsw		mm1, mm2

+				paddsw		mm3, mm4

+

+				movq		mm5, [ebx+esi]

+

+				paddsw		mm1, mm7

+				paddsw		mm3, mm7

+

+				movq		mm6, mm5

+				punpcklbw	mm5, mm0

+				punpckhbw	mm6, mm0

+

+				psllw		mm5, 1

+				psllw		mm6, 1

+

+				paddsw		mm1, mm5

+				paddsw		mm3, mm6

+

+				psrlw		mm1, 2

+				psrlw		mm3, 2

+

+				packuswb	mm1, mm0

+				packuswb	mm3, mm0

+

+				psllq		mm3, 32

+				por			mm1, mm3

+

+				add			esi, 0x08

+				cmp			esi, [w]

+				movq		[ebx+esi-8], mm1

+				jl			mc5

+

+				add			eax, [lx2]

+				add			ebx, [lx2]

+				add			ecx, [lx2]

+				dec			edi

+				mov			esi, 0x00

+				cmp			edi, 0x00

+				jg			mc5

+			}

+			break;

+

+		case 6:

+			// d[i] = ((d[i]<<1) + s[i]+s[i+1] + 3) >> 2;

+			__asm

+			{

+				pxor		mm0, mm0

+				movq		mm7, [mmmask_0003]

+				mov			eax, [s]

+				mov			ebx, [d]

+				mov			esi, 0x00

+				mov			edi, [h]

+mc6:

+				movq		mm1, [eax+esi]

+				movq		mm2, [eax+esi+1]

+				movq		mm3, mm1

+				movq		mm4, mm2

+

+				punpcklbw	mm1, mm0

+				punpckhbw	mm3, mm0

+

+				punpcklbw	mm2, mm0

+				punpckhbw	mm4, mm0

+

+				paddsw		mm1, mm2

+				paddsw		mm3, mm4

+

+				movq		mm5, [ebx+esi]

+

+				paddsw		mm1, mm7

+				paddsw		mm3, mm7

+

+				movq		mm6, mm5

+				punpcklbw	mm5, mm0

+				punpckhbw	mm6, mm0

+

+				psllw		mm5, 1

+				psllw		mm6, 1

+

+				paddsw		mm1, mm5

+				paddsw		mm3, mm6

+

+				psrlw		mm1, 2

+				psrlw		mm3, 2

+

+				packuswb	mm1, mm0

+				packuswb	mm3, mm0

+

+				psllq		mm3, 32

+				por			mm1, mm3

+

+				add			esi, 0x08

+				cmp			esi, [w]

+				movq		[ebx+esi-8], mm1

+				jl			mc6

+

+				add			eax, [lx2]

+				add			ebx, [lx2]

+				dec			edi

+				mov			esi, 0x00

+				cmp			edi, 0x00

+				jg			mc6

+			}

+			break;

+

+		case 7:

+			// d[i] = ((d[i]<<2) + s[i]+s[i+1]+s[i+lx]+s[i+lx+1] + 6)>>3;

+			__asm

+			{

+				pxor		mm0, mm0

+				movq		mm7, [mmmask_0006]

+				mov			eax, [s]

+				mov			ebx, [d]

+				mov			ecx, eax

+				add			ecx, [lx]

+				mov			esi, 0x00

+				mov			edi, [h]

+mc7:

+				movq		mm1, [eax+esi]

+				movq		mm2, [eax+esi+1]

+				movq		mm3, mm1

+				movq		mm4, mm2

+

+				punpcklbw	mm1, mm0

+				punpckhbw	mm3, mm0

+

+				punpcklbw	mm2, mm0

+				punpckhbw	mm4, mm0

+

+				paddsw		mm1, mm2

+				paddsw		mm3, mm4

+

+				movq		mm5, [ecx+esi]

+				paddsw		mm1, mm7

+

+				movq		mm6, [ecx+esi+1]

+				paddsw		mm3, mm7

+

+				movq		mm2, mm5

+				movq		mm4, mm6

+

+				punpcklbw	mm2, mm0

+				punpckhbw	mm5, mm0

+

+				punpcklbw	mm4, mm0

+				punpckhbw	mm6, mm0

+				

+				paddsw		mm2, mm4

+				paddsw		mm5, mm6

+

+				paddsw		mm1, mm2

+				paddsw		mm3, mm5

+

+				movq		mm6, [ebx+esi]

+

+				movq		mm4, mm6

+				punpcklbw	mm4, mm0

+				punpckhbw	mm6, mm0

+

+				psllw		mm4, 2

+				psllw		mm6, 2

+

+				paddsw		mm1, mm4

+				paddsw		mm3, mm6

+

+				psrlw		mm1, 3

+				psrlw		mm3, 3

+

+				packuswb	mm1, mm0

+				packuswb	mm3, mm0

+

+				psllq		mm3, 32

+				por			mm1, mm3

+

+				add			esi, 0x08

+				cmp			esi, [w]

+				movq		[ebx+esi-8], mm1

+				jl			mc7

+

+				add			eax, [lx2]

+				add			ebx, [lx2]

+				add			ecx, [lx2]

+				dec			edi

+				mov			esi, 0x00

+				cmp			edi, 0x00

+				jg			mc7

+			}

+			break;

+	}

+}

+

+//

+// motion

+//

+

+/* ISO/IEC 13818-2 sections 6.2.5.2, 6.3.17.2, and 7.6.3: Motion vectors */

+void CMPEG2Dec::motion_vectors(int PMV[2][2][2],int dmvector[2],

+					int motion_vertical_field_select[2][2], int s,

+					int motion_vector_count, int mv_format, int h_r_size,

+					int v_r_size, int dmv, int mvscale)

+{

+	if (motion_vector_count==1)

+	{

+		if (mv_format==MV_FIELD && !dmv)

+			motion_vertical_field_select[1][s] =

+			motion_vertical_field_select[0][s] = Get_Bits(1);

+

+		motion_vector(PMV[0][s],dmvector,h_r_size,v_r_size,dmv,mvscale,0);

+

+		/* update other motion vector predictors */

+		PMV[1][s][0] = PMV[0][s][0];

+		PMV[1][s][1] = PMV[0][s][1];

+	}

+	else

+	{

+		motion_vertical_field_select[0][s] = Get_Bits(1);

+		motion_vector(PMV[0][s],dmvector,h_r_size,v_r_size,dmv,mvscale,0);

+		motion_vertical_field_select[1][s] = Get_Bits(1);

+		motion_vector(PMV[1][s],dmvector,h_r_size,v_r_size,dmv,mvscale,0);

+	}

+}

+

+

+/* ISO/IEC 13818-2 section 7.6.3.6: Dual prime additional arithmetic */

+void CMPEG2Dec::Dual_Prime_Arithmetic(int DMV[][2],int *dmvector, int mvx,int mvy)

+{

+	if (picture_structure==FRAME_PICTURE)

+	{

+		if (top_field_first)

+		{

+			/* vector for prediction of top field from bottom field */

+			DMV[0][0] = ((mvx  +(mvx>0))>>1) + dmvector[0];

+			DMV[0][1] = ((mvy  +(mvy>0))>>1) + dmvector[1] - 1;

+

+			/* vector for prediction of bottom field from top field */

+			DMV[1][0] = ((3*mvx+(mvx>0))>>1) + dmvector[0];

+			DMV[1][1] = ((3*mvy+(mvy>0))>>1) + dmvector[1] + 1;

+		}

+		else

+		{

+			/* vector for prediction of top field from bottom field */

+			DMV[0][0] = ((3*mvx+(mvx>0))>>1) + dmvector[0];

+			DMV[0][1] = ((3*mvy+(mvy>0))>>1) + dmvector[1] - 1;

+

+			/* vector for prediction of bottom field from top field */

+			DMV[1][0] = ((mvx  +(mvx>0))>>1) + dmvector[0];

+			DMV[1][1] = ((mvy  +(mvy>0))>>1) + dmvector[1] + 1;

+		}

+	}

+	else

+	{

+		/* vector for prediction from field of opposite 'parity' */

+		DMV[0][0] = ((mvx+(mvx>0))>>1) + dmvector[0];

+		DMV[0][1] = ((mvy+(mvy>0))>>1) + dmvector[1];

+

+		/* correct for vertical field shift */

+		if (picture_structure==TOP_FIELD)

+			DMV[0][1]--;

+		else

+			DMV[0][1]++;

+	}

+}

+

+/* get and decode motion vector and differential motion vector for one prediction */

+void CMPEG2Dec::motion_vector(int *PMV, int *dmvector, int h_r_size, int v_r_size,

+				   int dmv, int mvscale, int full_pel_vector)

+{

+	int motion_code, motion_residual;

+

+	/* horizontal component */

+	/* ISO/IEC 13818-2 Table B-10 */

+	motion_code = Get_motion_code();

+

+	motion_residual = (h_r_size!=0 && motion_code!=0) ? Get_Bits(h_r_size) : 0;

+

+	decode_motion_vector(&PMV[0],h_r_size,motion_code,motion_residual,full_pel_vector);

+

+	if (dmv)

+		dmvector[0] = Get_dmvector();

+

+	/* vertical component */

+	motion_code     = Get_motion_code();

+	motion_residual = (v_r_size!=0 && motion_code!=0) ? Get_Bits(v_r_size) : 0;

+

+	if (mvscale)

+		PMV[1] >>= 1; /* DIV 2 */

+

+	decode_motion_vector(&PMV[1],v_r_size,motion_code,motion_residual,full_pel_vector);

+

+	if (mvscale)

+		PMV[1] <<= 1;

+

+	if (dmv)

+		dmvector[1] = Get_dmvector();

+}

+

+/* calculate motion vector component */

+/* ISO/IEC 13818-2 section 7.6.3.1: Decoding the motion vectors */

+/* Note: the arithmetic here is more elegant than that which is shown 

+   in 7.6.3.1.  The end results (PMV[][][]) should, however, be the same.  */

+void CMPEG2Dec::decode_motion_vector(int *pred, int r_size, int motion_code,

+								 int motion_residual, int full_pel_vector)

+{

+	int lim, vec;

+

+	lim = 16<<r_size;

+	vec = full_pel_vector ? (*pred >> 1) : (*pred);

+

+	if (motion_code>0)

+	{

+		vec+= ((motion_code-1)<<r_size) + motion_residual + 1;

+		if (vec>=lim)

+			vec-= lim + lim;

+	}

+	else if (motion_code<0)

+	{

+		vec-= ((-motion_code-1)<<r_size) + motion_residual + 1;

+		if (vec<-lim)

+			vec+= lim + lim;

+	}

+

+	*pred = full_pel_vector ? (vec<<1) : vec;

+}

+

+int CMPEG2Dec::Get_motion_code()

+{

+	int code;

+

+	if (Get_Bits(1))

+		return 0;

+

+	if ((code = Show_Bits(9))>=64)

+	{

+		code >>= 6;

+		Flush_Buffer(MVtab0[code].len);

+

+		return Get_Bits(1)?-MVtab0[code].val:MVtab0[code].val;

+	}

+

+	if (code>=24)

+	{

+		code >>= 3;

+		Flush_Buffer(MVtab1[code].len);

+

+		return Get_Bits(1)?-MVtab1[code].val:MVtab1[code].val;

+	}

+

+	if ((code-=12)<0)

+	{

+		Fault_Flag = 10;

+		return 0;

+	}

+

+	Flush_Buffer(MVtab2[code].len);

+

+	return Get_Bits(1) ? -MVtab2[code].val : MVtab2[code].val;

+}

+

+/* get differential motion vector (for dual prime prediction) */

+int CMPEG2Dec::Get_dmvector()

+{

+	if (Get_Bits(1))

+		return Get_Bits(1) ? -1 : 1;

+	else

+		return 0;

+}

+

+//

+// store

+//

+

+static const __int64 mmmask_0001 = 0x0001000100010001;

+static const __int64 mmmask_0002 = 0x0002000200020002;

+static const __int64 mmmask_0003 = 0x0003000300030003;

+static const __int64 mmmask_0004 = 0x0004000400040004;

+static const __int64 mmmask_0005 = 0x0005000500050005;

+static const __int64 mmmask_0007 = 0x0007000700070007;

+static const __int64 mmmask_0016 = 0x0010001000100010;

+static const __int64 mmmask_0040 = 0x0040004000400040;

+static const __int64 mmmask_0128 = 0x0080008000800080;

+

+void CMPEG2Dec::assembleFrame(unsigned char *src[], int pf, unsigned char *dst, int pitch)

+{

+	unsigned char *y444;

+

+	if (Fault_Flag)

+		Fault_Flag = 0;

+

+	if (Luminance_Flag)

+	{

+		Luminance_Filter(src[0], lum);

+		y444 = lum;

+	}

+	else

+		y444 = src[0];

+

+	if (chroma_format==CHROMA420)

+	{

+		conv420to422(src[1], u422, pf);

+		conv420to422(src[2], v422, pf);

+

+		if (!dstYUY2())

+		{

+			conv422to444(u422, u444);

+			conv422to444(v422, v444);

+		}

+	}

+	else if (!dstYUY2())

+	{

+		conv422to444(src[1], u444);

+		conv422to444(src[2], v444);

+	}

+

+	if (dstYUY2())

+		conv422toYUY2(y444, u422, v422, dst, pitch);

+	else

+ 		conv444toRGB24(y444, u444, v444, dst, pitch);

+}

+

+void CMPEG2Dec::Luminance_Filter(unsigned char *src, unsigned char *dst)

+{

+	src += CLIP_AREA;

+	dst += CLIP_AREA;

+

+	__asm

+	{

+		mov			edx, this

+		mov			eax, [src]

+		mov			ebx, [dst]

+		mov			esi, 0x00

+		mov			edi, [edx].LUM_AREA

+		pxor		mm0, mm0

+		movq		mm5, [edx].LumOffsetMask

+		movq		mm6, [edx].LumGainMask

+		movq		mm7, mmmask_0040

+

+lumconv:

+		movq		mm1, [eax+esi]

+		movq		mm2, mm1

+

+		punpcklbw	mm1, mm0

+		punpckhbw	mm2, mm0

+

+		pmullw		mm1, mm6

+		pmullw		mm2, mm6

+

+		paddw		mm1, mm7

+		paddw		mm2, mm7

+

+		psrlw		mm1, 7

+		psrlw		mm2, 7

+

+		paddw		mm1, mm5

+		paddw		mm2, mm5

+

+		packuswb	mm1, mm0

+		packuswb	mm2, mm0

+

+		add			esi, 0x08

+		cmp			esi, edi

+		movq		[ebx+esi-8], mm1

+		movq		[ebx+esi-4], mm2

+		jl			lumconv

+	}

+}

+

+void CMPEG2Dec::conv422to444(unsigned char *src, unsigned char *dst)

+{

+	src += HALF_CLIP_AREA;

+	dst += CLIP_AREA;

+

+	__asm

+	{

+		mov			edx, this

+		mov			eax, [src]

+		mov			ebx, [dst]

+		mov			edi, [edx].Clip_Height

+

+		movq		mm1, [mmmask_0001]

+		pxor		mm0, mm0

+

+convyuv444init:

+		movq		mm7, [eax]

+		mov			esi, 0x00

+

+convyuv444:

+		movq		mm2, mm7

+		movq		mm7, [eax+esi+8]

+		movq		mm3, mm2

+		movq		mm4, mm7

+

+		psrlq		mm3, 8

+		psllq		mm4, 56

+		por			mm3, mm4

+

+		movq		mm4, mm2

+		movq		mm5, mm3

+

+		punpcklbw	mm4, mm0

+		punpcklbw	mm5, mm0

+

+		movq		mm6, mm4

+		paddusw		mm4, mm1

+		paddusw		mm4, mm5

+		psrlw		mm4, 1

+		psllq		mm4, 8

+		por			mm4, mm6

+

+		punpckhbw	mm2, mm0

+		punpckhbw	mm3, mm0

+

+		movq		mm6, mm2

+		paddusw		mm2, mm1

+		paddusw		mm2, mm3

+

+		movq		[ebx+esi*2], mm4

+

+		psrlw		mm2, 1

+		psllq		mm2, 8

+		por			mm2, mm6

+

+		add			esi, 0x08

+		cmp			esi, [edx].HALF_WIDTH_D8

+		movq		[ebx+esi*2-8], mm2

+		jl			convyuv444

+

+		movq		mm2, mm7

+		punpcklbw	mm2, mm0

+		movq		mm3, mm2

+

+		psllq		mm2, 8

+		por			mm2, mm3

+

+		movq		[ebx+esi*2], mm2

+

+		punpckhbw	mm7, mm0

+		movq		mm6, mm7

+

+		psllq		mm6, 8

+		por			mm6, mm7

+

+		movq		[ebx+esi*2+8], mm6

+

+		add			eax, [edx].HALF_WIDTH

+		add			ebx, [edx].Coded_Picture_Width

+		dec			edi

+		cmp			edi, 0x00

+		jg			convyuv444init

+	}

+}

+

+void CMPEG2Dec::conv420to422(unsigned char *src, unsigned char *dst, int frame_type)

+{

+	if (frame_type)

+	{

+		__asm

+		{

+			push		ebp

+			mov			eax, [src]

+			mov			ebx, [dst]

+			mov			ebp, this

+			mov			ecx, ebx

+			add			ecx, ds:[ebp].HALF_WIDTH

+			mov			esi, 0x00

+			movq		mm3, [mmmask_0003]

+			pxor		mm0, mm0

+			movq		mm4, [mmmask_0002]

+

+			mov			edx, eax

+			add			edx, ds:[ebp].HALF_WIDTH

+convyuv422topp:

+			movd		mm1, [eax+esi]

+			movd		mm2, [edx+esi]

+			movd		[ebx+esi], mm1

+			punpcklbw	mm1, mm0

+			pmullw		mm1, mm3

+			paddusw		mm1, mm4

+			punpcklbw	mm2, mm0

+			paddusw		mm2, mm1

+			psrlw		mm2, 0x02

+			packuswb	mm2, mm0

+

+			add			esi, 0x04

+			cmp			esi, ds:[ebp].HALF_WIDTH

+			movd		[ecx+esi-4], mm2

+			jl			convyuv422topp

+

+			add			eax, ds:[ebp].HALF_WIDTH

+			add			ebx, ds:[ebp].Coded_Picture_Width

+			add			ecx, ds:[ebp].Coded_Picture_Width

+			mov			esi, 0x00

+

+			mov			edi, ds:[ebp].PROGRESSIVE_HEIGHT

+convyuv422p:

+			movd		mm1, [eax+esi]

+

+			punpcklbw	mm1, mm0

+			mov			edx, eax

+

+			pmullw		mm1, mm3

+			sub			edx, ds:[ebp].HALF_WIDTH

+

+			movd		mm5, [edx+esi]

+			movd		mm2, [edx+esi]

+

+			punpcklbw	mm5, mm0

+			punpcklbw	mm2, mm0

+			paddusw		mm5, mm1

+			paddusw		mm2, mm1

+			paddusw		mm5, mm4

+			paddusw		mm2, mm4

+			psrlw		mm5, 0x02

+			psrlw		mm2, 0x02

+			packuswb	mm5, mm0

+			packuswb	mm2, mm0

+

+			mov			edx, eax

+			add			edx, ds:[ebp].HALF_WIDTH

+			add			esi, 0x04

+			cmp			esi, ds:[ebp].HALF_WIDTH

+			movd		[ebx+esi-4], mm5

+			movd		[ecx+esi-4], mm2

+

+			jl			convyuv422p

+

+			add			eax, ds:[ebp].HALF_WIDTH

+			add			ebx, ds:[ebp].Coded_Picture_Width

+			add			ecx, ds:[ebp].Coded_Picture_Width

+			mov			esi, 0x00

+			dec			edi

+			cmp			edi, 0x00

+			jg			convyuv422p

+

+			mov			edx, eax

+			sub			edx, ds:[ebp].HALF_WIDTH

+convyuv422bottomp:

+			movd		mm1, [eax+esi]

+			movd		mm5, [edx+esi]

+			punpcklbw	mm5, mm0

+			movd		[ecx+esi], mm1

+

+			punpcklbw	mm1, mm0

+			pmullw		mm1, mm3

+			paddusw		mm5, mm1

+			paddusw		mm5, mm4

+			psrlw		mm5, 0x02

+			packuswb	mm5, mm0

+

+			add			esi, 0x04

+			cmp			esi, ds:[ebp].HALF_WIDTH

+			movd		[ebx+esi-4], mm5

+			jl			convyuv422bottomp

+			pop			ebp

+		}

+	}

+	else

+	{

+		__asm

+		{

+			push		ebp

+			mov			eax, [src]

+			mov			ecx, [dst]

+			mov			ebp, this

+			mov			esi, 0x00

+			pxor		mm0, mm0

+			movq		mm3, [mmmask_0003]

+			movq		mm4, [mmmask_0004]

+			movq		mm5, [mmmask_0005]

+

+convyuv422topi:

+			movd		mm1, [eax+esi]

+			mov			ebx, eax

+			add			ebx, ds:[ebp].HALF_WIDTH

+			movd		mm2, [ebx+esi]

+			movd		[ecx+esi], mm1

+			punpcklbw	mm1, mm0

+			movq		mm6, mm1

+			pmullw		mm1, mm3

+

+			punpcklbw	mm2, mm0

+			movq		mm7, mm2

+			pmullw		mm2, mm5

+			paddusw		mm2, mm1

+			paddusw		mm2, mm4

+			psrlw		mm2, 0x03

+			packuswb	mm2, mm0

+

+			mov			edx, ecx

+			add			edx, ds:[ebp].HALF_WIDTH

+			pmullw		mm6, mm5

+			movd		[edx+esi], mm2

+

+			add			ebx, ds:[ebp].HALF_WIDTH

+			movd		mm2, [ebx+esi]

+			punpcklbw	mm2, mm0

+			pmullw		mm2, mm3

+			paddusw		mm2, mm6

+			paddusw		mm2, mm4

+			psrlw		mm2, 0x03

+			packuswb	mm2, mm0

+

+			add			edx, ds:[ebp].HALF_WIDTH

+			add			ebx, ds:[ebp].HALF_WIDTH

+			pmullw		mm7, [mmmask_0007]

+			movd		[edx+esi], mm2

+

+			movd		mm2, [ebx+esi]

+			punpcklbw	mm2, mm0

+			paddusw		mm2, mm7

+			paddusw		mm2, mm4

+			psrlw		mm2, 0x03

+			packuswb	mm2, mm0

+

+			add			edx, ds:[ebp].HALF_WIDTH

+			add			esi, 0x04

+			cmp			esi, ds:[ebp].HALF_WIDTH

+			movd		[edx+esi-4], mm2

+

+			jl			convyuv422topi

+

+			add			eax, ds:[ebp].Coded_Picture_Width

+			add			ecx, ds:[ebp].DOUBLE_WIDTH

+			mov			esi, 0x00

+

+			mov			edi, ds:[ebp].INTERLACED_HEIGHT

+convyuv422i:

+			movd		mm1, [eax+esi]

+			punpcklbw	mm1, mm0

+			movq		mm6, mm1

+			mov			ebx, eax

+			sub			ebx, ds:[ebp].Coded_Picture_Width

+			movd		mm3, [ebx+esi]

+			pmullw		mm1, [mmmask_0007]

+			punpcklbw	mm3, mm0

+			paddusw		mm3, mm1

+			paddusw		mm3, mm4

+			psrlw		mm3, 0x03

+			packuswb	mm3, mm0

+

+			add			ebx, ds:[ebp].HALF_WIDTH

+			movq		mm1, [ebx+esi]

+			add			ebx, ds:[ebp].Coded_Picture_Width

+			movd		[ecx+esi], mm3

+

+			movq		mm3, [mmmask_0003]

+			movd		mm2, [ebx+esi]

+

+			punpcklbw	mm1, mm0

+			pmullw		mm1, mm3

+			punpcklbw	mm2, mm0

+			movq		mm7, mm2

+			pmullw		mm2, mm5

+			paddusw		mm2, mm1

+			paddusw		mm2, mm4

+			psrlw		mm2, 0x03

+			packuswb	mm2, mm0

+

+			pmullw		mm6, mm5

+			mov			edx, ecx

+			add			edx, ds:[ebp].HALF_WIDTH

+			movd		[edx+esi], mm2

+

+			add			ebx, ds:[ebp].HALF_WIDTH

+			movd		mm2, [ebx+esi]

+			punpcklbw	mm2, mm0

+			pmullw		mm2, mm3

+			paddusw		mm2, mm6

+			paddusw		mm2, mm4

+			psrlw		mm2, 0x03

+			packuswb	mm2, mm0

+

+			pmullw		mm7, [mmmask_0007]

+			add			edx, ds:[ebp].HALF_WIDTH

+			add			ebx, ds:[ebp].HALF_WIDTH

+ 			movd		[edx+esi], mm2

+

+			movd		mm2, [ebx+esi]

+			punpcklbw	mm2, mm0

+			paddusw		mm2, mm7

+			paddusw		mm2, mm4

+			psrlw		mm2, 0x03

+			packuswb	mm2, mm0

+

+			add			edx, ds:[ebp].HALF_WIDTH

+			add			esi, 0x04

+			cmp			esi, ds:[ebp].HALF_WIDTH

+			movd		[edx+esi-4], mm2

+

+			jl			convyuv422i

+			add			eax, ds:[ebp].Coded_Picture_Width

+			add			ecx, ds:[ebp].DOUBLE_WIDTH

+			mov			esi, 0x00

+			dec			edi

+			cmp			edi, 0x00

+			jg			convyuv422i

+

+convyuv422bottomi:

+			movd		mm1, [eax+esi]

+			movq		mm6, mm1

+			punpcklbw	mm1, mm0

+			mov			ebx, eax

+			sub			ebx, ds:[ebp].Coded_Picture_Width

+			movd		mm3, [ebx+esi]

+			punpcklbw	mm3, mm0

+			pmullw		mm1, [mmmask_0007]

+			paddusw		mm3, mm1

+			paddusw		mm3, mm4

+			psrlw		mm3, 0x03

+			packuswb	mm3, mm0

+

+			add			ebx, ds:[ebp].HALF_WIDTH

+			movq		mm1, [ebx+esi]

+			punpcklbw	mm1, mm0

+			movd		[ecx+esi], mm3

+

+			pmullw		mm1, [mmmask_0003]

+			add			ebx, ds:[ebp].Coded_Picture_Width

+			movd		mm2, [ebx+esi]

+			punpcklbw	mm2, mm0

+			movq		mm7, mm2

+			pmullw		mm2, mm5

+			paddusw		mm2, mm1

+			paddusw		mm2, mm4

+			psrlw		mm2, 0x03

+			packuswb	mm2, mm0

+

+			mov			edx, ecx

+			add			edx, ds:[ebp].HALF_WIDTH

+			pmullw		mm7, [mmmask_0007]

+			movd		[edx+esi], mm2

+

+			add			edx, ds:[ebp].HALF_WIDTH

+ 			movd		[edx+esi], mm6

+

+			punpcklbw	mm6, mm0

+			paddusw		mm6, mm7

+			paddusw		mm6, mm4

+			psrlw		mm6, 0x03

+			packuswb	mm6, mm0

+

+			add			edx, ds:[ebp].HALF_WIDTH

+			add			esi, 0x04

+			cmp			esi, ds:[ebp].HALF_WIDTH

+			movd		[edx+esi-4], mm6

+

+			jl			convyuv422bottomi

+			pop			ebp

+		}

+	}

+}

+

+void CMPEG2Dec::conv444toRGB24(unsigned char *py, unsigned char *pu, unsigned char *pv, unsigned char *dst, int pitch)

+{

+	int PWIDTH = pitch - DSTBYTES;

+

+	py += CLIP_STEP;

+	pu += CLIP_STEP;

+	pv += CLIP_STEP;

+

+	int Clip_Height = this->Clip_Height;

+	__int64 RGB_Offset = this->RGB_Offset;

+	__int64 RGB_Scale = this->RGB_Scale;

+	__int64 RGB_CBU = this->RGB_CBU;

+	__int64 RGB_CRV = this->RGB_CRV;

+	__int64 RGB_CGX = this->RGB_CGX;

+	int Clip_Width = this->Clip_Width;

+	int Coded_Picture_Width = this->Coded_Picture_Width;

+

+	__asm

+	{

+		mov			eax, [py]

+		mov			ebx, [pu]

+		mov			ecx, [pv]

+		mov			edx, [dst]

+		mov			edi, Clip_Height

+		mov			esi, 0x00

+		pxor		mm0, mm0

+

+convRGB24:

+		movd		mm1, [eax+esi]

+		movd		mm3, [ebx+esi]

+		punpcklbw	mm1, mm0

+		punpcklbw	mm3, mm0

+		movd		mm5, [ecx+esi]

+		punpcklbw	mm5, mm0

+		movq		mm7, [mmmask_0128]

+		psubw		mm3, mm7

+		psubw		mm5, mm7

+

+		psubw		mm1, RGB_Offset

+		movq		mm2, mm1

+		movq		mm7, [mmmask_0001]

+		punpcklwd	mm1, mm7

+		punpckhwd	mm2, mm7

+		movq		mm7, RGB_Scale

+		pmaddwd		mm1, mm7

+		pmaddwd		mm2, mm7

+

+		movq		mm4, mm3

+		punpcklwd	mm3, mm0

+		punpckhwd	mm4, mm0

+		movq		mm7, RGB_CBU

+		pmaddwd		mm3, mm7

+		pmaddwd		mm4, mm7

+		paddd		mm3, mm1

+		paddd		mm4, mm2

+		psrld		mm3, 13

+		psrld		mm4, 13

+		packuswb	mm3, mm0

+		packuswb	mm4, mm0

+

+		movq		mm6, mm5

+		punpcklwd	mm5, mm0

+		punpckhwd	mm6, mm0

+		movq		mm7, RGB_CRV

+		pmaddwd		mm5, mm7

+		pmaddwd		mm6, mm7

+		paddd		mm5, mm1

+		paddd		mm6, mm2

+		psrld		mm5, 13

+		psrld		mm6, 13

+		packuswb	mm5, mm0

+		packuswb	mm6, mm0

+

+		punpcklbw	mm3, mm5

+		punpcklbw	mm4, mm6

+		movq		mm5, mm3

+		movq		mm6, mm4

+		psrlq		mm5, 16

+		psrlq		mm6, 16

+		por			mm3, mm5

+		por			mm4, mm6

+

+		movd		mm5, [ebx+esi]

+		movd		mm6, [ecx+esi]

+		punpcklbw	mm5, mm0

+		punpcklbw	mm6, mm0

+		movq		mm7, [mmmask_0128]

+		psubw		mm5, mm7

+		psubw		mm6, mm7

+

+		movq		mm7, mm6

+		punpcklwd	mm6, mm5

+		punpckhwd	mm7, mm5		

+		movq		mm5, RGB_CGX

+		pmaddwd		mm6, mm5

+		pmaddwd		mm7, mm5

+		paddd		mm6, mm1

+		paddd		mm7, mm2

+

+		psrld		mm6, 13

+		psrld		mm7, 13

+		packuswb	mm6, mm0

+		packuswb	mm7, mm0

+

+		punpcklbw	mm3, mm6

+		punpcklbw	mm4, mm7

+

+		movq		mm1, mm3

+		movq		mm5, mm4

+		movq		mm6, mm4

+

+		psrlq		mm1, 32

+		psllq		mm1, 24

+		por			mm1, mm3

+

+		psrlq		mm3, 40

+		psllq		mm6, 16

+		por			mm3, mm6

+		movd		[edx], mm1

+

+		psrld		mm4, 16

+		psrlq		mm5, 24

+		por			mm5, mm4

+		movd		[edx+4], mm3

+

+		add			edx, 0x0c

+		add			esi, 0x04

+		cmp			esi, Clip_Width

+		movd		[edx-4], mm5

+

+		jl			convRGB24

+

+		add			eax, Coded_Picture_Width

+		add			ebx, Coded_Picture_Width

+		add			ecx, Coded_Picture_Width

+		add			edx, PWIDTH

+		mov			esi, 0x00

+		dec			edi

+		cmp			edi, 0x00

+		jg			convRGB24

+

+		emms

+	}

+}

+

+// YUV 4:2:2 Format:

+// YUYV YUYV ...

+void CMPEG2Dec::conv422toYUY2(unsigned char *py, unsigned char *pu, unsigned char *pv, unsigned char *dst, int pitch)

+{

+	py += CLIP_STEP;

+	pu += CLIP_STEP;

+	pv += CLIP_STEP;

+

+	int y = this->Clip_Height;

+	int Clip_Width_2 = this->Clip_Width / 2;

+	int Coded_Picture_Width = this->Coded_Picture_Width;

+	int Coded_Picture_Width_2 = this->Coded_Picture_Width / 2;

+

+	__asm

+	{

+		emms

+		mov			eax, [py]

+		mov			ebx, [pu]

+		mov			ecx, [pv]

+		mov			edx, [dst]

+		mov			edi, Clip_Width_2

+	yloop:

+		xor			esi, esi

+	xloop:

+		movd		mm1, [eax+esi*2]		;0000YYYY

+		movd		mm2, [ebx+esi]			;0000UUUU

+		movd		mm3, [ecx+esi]			;0000VVVV

+		;interleave this to VYUYVYUY

+		punpcklbw	mm2, mm3				;VUVUVUVU

+		punpcklbw	mm1, mm2				;VYUYVYUY

+		movq		[edx+esi*4], mm1

+		movd		mm1, [eax+esi*2+4]		;0000YYYY

+		punpckhdq	mm2, mm2				;xxxxVUVU

+		punpcklbw	mm1, mm2				;VYUYVYUY

+		movq		[edx+esi*4+8], mm1

+		add			esi, 4

+		cmp			esi, edi

+		jb			xloop

+		add			edx, pitch

+		add			eax, Coded_Picture_Width

+		add			ebx, Coded_Picture_Width_2

+		add			ecx, Coded_Picture_Width_2

+		dec			y

+		jnz			yloop

+		emms

+	}

+}

+

+//

+// codec

+//

+

+static const int ChromaFormat[4] = {

+	0, 6, 8, 12

+};

+

+CMPEG2Dec::CMPEG2Dec()

+{

+  VF_File = 0;

+  VF_FrameLimit = VF_FrameBound = VF_GOPLimit = VF_GOPNow = VF_GOPSize =

+	VF_OldFrame = VF_OldRef = 0;

+  VF_FrameSize = VF_FrameRate = 0;

+  memset(Rdbfr, 0, sizeof(Rdbfr));

+  Rdptr = Rdmax = 0;

+  CurrentBfr = NextBfr = BitsLeft = Val = Read = 0;

+  Fault_Flag = File_Flag = File_Limit = FO_Flag = IDCT_Flag = SystemStream_Flag = 0;

+  Luminance_Flag = Resize_Flag = KeyOp_Flag = lfsr0 = lfsr1 = 0;

+  BufferOp = 0;

+  memset(intra_quantizer_matrix, 0, sizeof(intra_quantizer_matrix));

+  memset(non_intra_quantizer_matrix, 0, sizeof(non_intra_quantizer_matrix));

+  memset(chroma_intra_quantizer_matrix, 0, sizeof(chroma_intra_quantizer_matrix));

+  memset(chroma_non_intra_quantizer_matrix, 0, sizeof(chroma_non_intra_quantizer_matrix));

+  load_intra_quantizer_matrix =

+  load_non_intra_quantizer_matrix =

+  load_chroma_intra_quantizer_matrix =

+  load_chroma_non_intra_quantizer_matrix = 0;

+  q_scale_type =

+  alternate_scan =

+  quantizer_scale = 0;

+

+	int i;

+	for (i=0; i<MAX_FILE_NUMBER; i++) Infilename[i] = NULL;

+	for (i=0; i<8; i++) p_block[i] = block[i] = NULL;

+	p_fTempArray = fTempArray = NULL;

+	for (i=0; i<3; i++) backward_reference_frame[i] = forward_reference_frame[i] =  auxframe[i] = NULL;

+	lum = NULL;

+	u422 = v422 = u444 = v444 = dstFrame = NULL;

+	hLibrary = NULL;

+

+  CheckCPU();

+}

+

+static char* myfgets(char* buff, int len, FILE* file)

+{

+	char* ret = buff;

+

+	ret[0] = 0;

+	

+	while(ret = fgets(buff, len, file))

+	{

+		while(isspace(*ret)) ret++;

+		if(*ret) break;

+	}

+

+	return(ret);

+}

+

+int CMPEG2Dec::Open(LPCTSTR path, DstFormat dstFormat)

+{

+	m_dstFormat = dstFormat;

+	char ID[19], PASS[19] = "DVD2AVIProjectFile";

+	DWORD i, j, size, code, type, tff, rff, film, ntsc, gop, top, bottom, mapping;

+	int repeat_on, repeat_off, repeat_init;

+	int Clip_Top, Clip_Bottom, Clip_Left, Clip_Right, Squeeze_Width, Squeeze_Height;

+

+	HKEY key; DWORD value = REG_SZ; DWORD length = 256;

+	char *ext, buffer[256];

+

+	CMPEG2Dec* out = this;

+

+	out->VF_File = _tfopen(path, _T("r"));

+	if (out->VF_File==NULL)

+		return 0;

+	if (fgets(ID, 19, out->VF_File)==NULL)

+		return 0;

+	if (strcmp(ID, PASS))

+		return 0;

+

+	// load DLL

+	if (RegOpenKeyExA(HKEY_CURRENT_USER, "Software\\VFPlugin", 0, KEY_ALL_ACCESS, &key)==ERROR_SUCCESS)

+	{

+		RegQueryValueExA(key, "DVD2AVI", NULL, &value, (unsigned char*)buffer, &length);

+

+		ext = strrchr(buffer, '\\');

+		sprintf(buffer + (int)(ext-buffer) + 1, "OpenDVD.dll");

+		RegCloseKey(key);

+	}

+

+	if ((hLibrary = LoadLibraryA(buffer)) != NULL)

+		BufferOp = (PBufferOp) GetProcAddress(hLibrary, "BufferOp");

+

+	for (i=0; i<MAX_FILE_NUMBER; i++)

+		Infilename[i] = DNew char[_MAX_PATH];

+

+	if(1 != fscanf(out->VF_File, "%d", &File_Limit))

+		return 0;

+

+	i = File_Limit;

+	while (i)

+	{

+		if(1 != fscanf(out->VF_File, "%d ", &j))

+			return 0;

+		fgets(Infilename[File_Limit-i], j+1, out->VF_File);

+		if ((Infile[File_Limit-i] = _open(Infilename[File_Limit-i], _O_RDONLY | _O_BINARY))==-1)

+			return 0;

+		i--;

+	}

+

+	if(3 != sscanf(myfgets(buffer, sizeof(buffer), out->VF_File), "\nStream_Type=%d,%X,%X\n", &SystemStream_Flag, &lfsr0, &lfsr1))

+		return 0;

+	if (lfsr0 || lfsr1)

+		KeyOp_Flag = 1;

+	else

+		KeyOp_Flag = 0;

+

+	if(1 != sscanf(myfgets(buffer, sizeof(buffer), out->VF_File), "iDCT_Algorithm=%d\n", &IDCT_Flag))

+		return 0;

+

+	switch (IDCT_Flag)

+	{

+		case IDCT_SSEMMX:

+			if (!cpu.ssemmx)

+				IDCT_Flag = IDCT_MMX;

+			break;

+

+		case IDCT_FPU:

+			Initialize_FPU_IDCT();

+			break;

+

+		case IDCT_REF:

+			Initialize_REF_IDCT();

+			break;

+	}

+

+	File_Flag = 0;

+	_lseeki64(Infile[0], 0, SEEK_SET);

+	Initialize_Buffer();

+

+	do

+	{

+		next_start_code();

+		code = Get_Bits(32);

+	}

+	while (code!=SEQUENCE_HEADER_CODE);

+

+	sequence_header();

+

+	mb_width = (horizontal_size+15)/16;

+	mb_height = progressive_sequence ? (vertical_size+15)/16 : 2*((vertical_size+31)/32);

+

+	Coded_Picture_Width = 16 * mb_width;

+	Coded_Picture_Height = 16 * mb_height;

+

+	Chroma_Width = (chroma_format==CHROMA444) ? Coded_Picture_Width : Coded_Picture_Width>>1;

+	Chroma_Height = (chroma_format!=CHROMA420) ? Coded_Picture_Height : Coded_Picture_Height>>1;

+

+	block_count = ChromaFormat[chroma_format];

+

+	for (i=0; i<8; i++)

+	{

+		p_block[i] = (short *)DNew BYTE[sizeof(short)*64 + 64];

+		block[i]   = (short *)((long)p_block[i] + 64 - (long)p_block[i]%64);

+	}

+

+	p_fTempArray = (void *)DNew BYTE[sizeof(float)*128 + 64];

+	fTempArray = (void *)((long)p_fTempArray + 64 - (long)p_fTempArray%64);

+

+	for (i=0; i<3; i++)

+	{

+		if (i==0)

+			size = Coded_Picture_Width * Coded_Picture_Height;

+		else

+			size = Chroma_Width * Chroma_Height;

+

+		backward_reference_frame[i] = DNew unsigned char[size];

+		forward_reference_frame[i] = DNew unsigned char[size];

+		auxframe[i] = DNew unsigned char[size];

+	}

+

+	if(1 != sscanf(myfgets(buffer, sizeof(buffer), out->VF_File), "YUVRGB_Scale=%d\n", &i))

+		return 0;

+

+	if (i)

+	{

+		RGB_Scale = 0x1000254310002543;

+		RGB_Offset = 0x0010001000100010;

+		RGB_CBU = 0x0000408D0000408D;

+		RGB_CGX = 0xF377E5FCF377E5FC;

+		RGB_CRV = 0x0000331300003313;

+	}

+	else

+	{

+		RGB_Scale = 0x1000200010002000;

+		RGB_Offset = 0x0000000000000000;

+		RGB_CBU = 0x000038B4000038B4;

+		RGB_CGX = 0xF4FDE926F4FDE926;

+		RGB_CRV = 0x00002CDD00002CDD;

+	}

+

+	char* tmp = myfgets(buffer, sizeof(buffer), out->VF_File);

+	if(2 != sscanf(tmp, "Luminance=%d,%d\n", &i, &j))

+	{

+		if(2 != sscanf(tmp, "Luminance_Filter=%d,%d\n", &i, &j))

+			return 0;

+		i=128; j=0;

+	}

+

+	if (i==128 && j==0)

+		Luminance_Flag = 0;

+	else

+	{

+		Luminance_Flag = 1;

+		LumGainMask = ((__int64)i<<48) + ((__int64)i<<32) + ((__int64)i<<16) + (__int64)i;

+		LumOffsetMask = ((__int64)j<<48) + ((__int64)j<<32) + ((__int64)j<<16) + (__int64)j;

+

+		lum = DNew unsigned char[Coded_Picture_Width * Coded_Picture_Height];

+	}

+

+	if(6 != sscanf(myfgets(buffer, sizeof(buffer), out->VF_File), "Picture_Size=%d,%d,%d,%d,%d,%d\n", 

+		&Clip_Top, &Clip_Bottom, &Clip_Left, &Clip_Right, &Squeeze_Width, &Squeeze_Height))

+		return 0;

+

+	Resize_Flag = 0;

+	Resize_Width = Clip_Width = Coded_Picture_Width;

+	Resize_Height = Clip_Height = Coded_Picture_Height;

+	CLIP_AREA = HALF_CLIP_AREA = CLIP_STEP = 0;

+

+	if (Clip_Top || Clip_Bottom || Clip_Left || Clip_Right)

+	{

+		Clip_Width -= Clip_Left+Clip_Right;

+		Clip_Height -= Clip_Top+Clip_Bottom;

+		Resize_Width = Clip_Width;

+		Resize_Height = Clip_Height;

+

+		CLIP_AREA = Coded_Picture_Width * Clip_Top;

+		HALF_CLIP_AREA = (Coded_Picture_Width>>1) * Clip_Top;

+		CLIP_STEP = Coded_Picture_Width * Clip_Top + Clip_Left;

+	}

+

+	if (Squeeze_Width || Squeeze_Height)

+	{

+		Resize_Flag = 1;

+		Resize_Width -= Squeeze_Width;

+		Resize_Height -= Squeeze_Height;

+	}

+

+	DSTBYTES = Clip_Width * (dstRGB24() ? 3 : 2);

+	DSTBYTES2 = DSTBYTES * 2;

+	LUM_AREA = Coded_Picture_Width * Clip_Height;

+	PROGRESSIVE_HEIGHT = (Coded_Picture_Height>>1) - 2;

+	INTERLACED_HEIGHT = (Coded_Picture_Height>>2) - 2;

+	HALF_WIDTH = Coded_Picture_Width>>1;

+	HALF_WIDTH_D8 = (Coded_Picture_Width>>1) - 8;

+	DOUBLE_WIDTH = Coded_Picture_Width<<1;

+

+	u422 = DNew unsigned char[Coded_Picture_Width * Coded_Picture_Height / 2];

+	v422 = DNew unsigned char[Coded_Picture_Width * Coded_Picture_Height / 2];

+	u444 = DNew unsigned char[Coded_Picture_Width * Coded_Picture_Height];

+	v444 = DNew unsigned char[Coded_Picture_Width * Coded_Picture_Height];

+	dstFrame = DNew unsigned char[Clip_Width * Clip_Height * 4];  // max value (super set)

+

+	if(1 != sscanf(myfgets(buffer, sizeof(buffer), out->VF_File), "Field_Operation=%d\n", &FO_Flag))

+		return 0;

+	if(1 != sscanf(myfgets(buffer, sizeof(buffer), out->VF_File), "Frame_Rate=%d\n", &(out->VF_FrameRate)))

+		return 0;

+	if(4 != sscanf(myfgets(buffer, sizeof(buffer), out->VF_File), "Location=%d,%X,%d,%X\n", &i, &j, &i, &j))

+		return 0;

+

+	ntsc = film = top = bottom = gop = mapping = repeat_on = repeat_off = repeat_init = 0;

+

+	while (1 == fscanf(out->VF_File, "%d", &type) && type<9)

+	{

+		if (type==7)	// I frame

+		{

+			GOPList[gop] = reinterpret_cast<GOPLIST*>(calloc(1, sizeof(GOPLIST)));

+			GOPList[gop]->number = film;

+			if(2 != fscanf(out->VF_File, "%d %X", &(GOPList[gop]->file), &j))

+				break;

+

+			GOPList[gop]->position = (__int64)j*BUFFER_SIZE;

+			gop ++;

+

+			if(1 != fscanf(out->VF_File, "%d", &j))

+				break;

+

+			tff = j>>1;

+			rff = j & 1;

+		}

+		else	// P, B frame

+		{

+			tff = type>>1;

+			rff = type & 1;

+		}

+

+		if (!film)

+		{

+			if (tff)

+				Field_Order = 1;

+			else

+				Field_Order = 0;

+		}

+

+		if (FO_Flag==FO_FILM)

+		{

+			if (rff)

+				repeat_on++;

+			else

+				repeat_off++;

+

+			if (repeat_init)

+			{

+				if (repeat_off-repeat_on == 5)

+				{

+					repeat_on = repeat_off = 0;

+				}

+				else

+				{

+					FrameList[mapping] = reinterpret_cast<FRAMELIST*>(calloc(1, sizeof(FRAMELIST)));

+					FrameList[mapping]->top = FrameList[mapping]->bottom = film;

+					mapping ++;

+				}

+

+				if (repeat_on-repeat_off == 5)

+				{

+					repeat_on = repeat_off = 0;

+					FrameList[mapping] = reinterpret_cast<FRAMELIST*>(calloc(1, sizeof(FRAMELIST)));

+					FrameList[mapping]->top = FrameList[mapping]->bottom = film;

+					mapping ++;

+				}

+			}

+			else

+			{

+				if (repeat_off-repeat_on == 3)

+				{

+					repeat_on = repeat_off = 0;

+					repeat_init = 1;

+				}

+				else

+				{

+					FrameList[mapping] = reinterpret_cast<FRAMELIST*>(calloc(1, sizeof(FRAMELIST)));

+					FrameList[mapping]->top = FrameList[mapping]->bottom = film;

+					mapping ++;

+				}

+

+				if (repeat_on-repeat_off == 3)

+				{

+					repeat_on = repeat_off = 0;

+					repeat_init = 1;

+

+					FrameList[mapping] = reinterpret_cast<FRAMELIST*>(calloc(1, sizeof(FRAMELIST)));

+					FrameList[mapping]->top = FrameList[mapping]->bottom = film;

+					mapping ++;

+				}

+			}

+		}

+		else

+		{

+			if (top)

+			{

+				FrameList[ntsc]->bottom = film;

+				ntsc ++;

+				FrameList[ntsc] = reinterpret_cast<FRAMELIST*>(calloc(1, sizeof(FRAMELIST)));

+				FrameList[ntsc]->top = film;

+			}

+			else if (bottom)

+			{

+				FrameList[ntsc]->top = film;

+				ntsc ++;

+				FrameList[ntsc] = reinterpret_cast<FRAMELIST*>(calloc(1, sizeof(FRAMELIST)));

+				FrameList[ntsc]->bottom = film;

+			}

+			else

+			{

+				FrameList[ntsc] = reinterpret_cast<FRAMELIST*>(calloc(1, sizeof(FRAMELIST)));

+				FrameList[ntsc]->top = film;

+				FrameList[ntsc]->bottom = film;

+				ntsc ++;

+			}

+

+			if (rff)

+			{

+				if (!top && !bottom)

+					FrameList[ntsc] = reinterpret_cast<FRAMELIST*>(calloc(1, sizeof(FRAMELIST)));

+

+				if (tff)

+				{

+					FrameList[ntsc]->top = film;

+					top = 1;

+				}

+				else

+				{

+					FrameList[ntsc]->bottom = film;

+					bottom = 1;

+				}

+

+				if (top && bottom)

+				{

+					top = bottom = 0;

+					ntsc ++;

+				}

+			}

+		}

+

+		film ++;

+	}

+

+	out->VF_FrameBound = film;

+	film -= 2;

+

+	if (FO_Flag==FO_FILM)

+	{

+		while (FrameList[mapping-1]->top >= film)

+			mapping --;

+

+		out->VF_FrameLimit = mapping;

+	}

+	else

+	{

+		if (FO_Flag==FO_SWAP)

+		{

+			Field_Order = !Field_Order;

+

+			if (Field_Order)

+				for (i=0; i<ntsc-1; i++)

+					FrameList[i]->bottom = FrameList[i+1]->bottom;

+			else

+				for (i=0; i<ntsc-1; i++)

+					FrameList[i]->top = FrameList[i+1]->top;

+		}

+

+		while ((FrameList[ntsc-1]->top >= film) || (FrameList[ntsc-1]->bottom >= film))

+			ntsc --;

+

+		out->VF_FrameLimit = ntsc;

+

+		for (i=0; i<out->VF_FrameLimit-1; i++)

+			if (FrameList[i]->top==FrameList[i+1]->top || FrameList[i]->top==FrameList[i+1]->bottom ||

+				FrameList[i]->bottom==FrameList[i+1]->top || FrameList[i]->bottom==FrameList[i+1]->bottom)

+			{

+				FrameList[i]->forward = 1;

+				FrameList[i+1]->backward = 1;

+			}

+	}

+

+	Full_Frame = 1;

+	for (i=0; i<out->VF_FrameLimit; i++)

+		if (FrameList[i]->top!=FrameList[i]->bottom)

+		{

+			Full_Frame = 0;

+			break;

+		}

+

+	out->VF_GOPNow = out->VF_GOPLimit = gop;

+	out->VF_OldFrame = out->VF_FrameLimit;

+	out->VF_FrameSize = Clip_Width * Clip_Height * 3;

+

+	return 1;

+}

+

+void CMPEG2Dec::Decode(unsigned char *dst, DWORD frame, int pitch)

+{

+	DWORD i, now, size, origin, ref, fo;

+	int remain;

+

+	CMPEG2Dec* in = this;

+

+	if (FO_Flag==FO_FILM)

+	{

+		fo = 0;

+		frame = FrameList[frame]->top;

+	}

+

+	origin = frame;

+

+	if (FO_Flag!=FO_FILM)

+	{

+		if (FrameList[frame]->top == FrameList[frame]->bottom)

+		{

+			fo = 0;

+			frame = FrameList[frame]->top;

+		}

+		else if (FrameList[frame]->top < FrameList[frame]->bottom)

+		{

+			fo = 1;

+			frame = FrameList[frame]->top;

+		}

+		else

+		{

+			fo = 2;

+			frame = FrameList[frame]->bottom;

+		}

+	}

+

+	ref = frame;

+

+	if (frame >= GOPList[in->VF_GOPLimit-1]->number)

+	{

+		now = in->VF_GOPLimit-1;

+		ref -= GOPList[in->VF_GOPLimit-1]->number;

+		size = in->VF_FrameBound - GOPList[in->VF_GOPLimit-1]->number + 1;

+	}

+	else

+		for (now = 0; now < (in->VF_GOPLimit-1); now++)

+		{

+			if (frame>=GOPList[now]->number && frame<GOPList[now+1]->number)

+			{

+				ref -= GOPList[now]->number;

+				size = GOPList[now+1]->number - GOPList[now]->number + 1;

+				break;

+			}

+		}

+

+	if (fo)

+		ref ++;

+

+	if (now != in->VF_GOPNow)

+	{

+		if ((in->VF_OldFrame + 1)==origin)

+		{

+			if (Full_Frame)

+			{

+				Get_Hdr();

+				Decode_Picture(1, dst, pitch);

+

+				if (picture_structure!=FRAME_PICTURE)

+				{

+					Get_Hdr();

+					Decode_Picture(1, dst, pitch);

+				}

+			}

+			else

+				switch (fo)

+				{

+					case 0:

+						if (!FrameList[origin]->backward)

+						{

+							Get_Hdr();

+							Decode_Picture(1, dst, pitch);

+

+							if (picture_structure!=FRAME_PICTURE)

+							{

+								Get_Hdr();

+								Decode_Picture(1, dst, pitch);

+							}

+

+							if (FrameList[origin]->forward)

+							{

+								if (Field_Order)

+									Copyodd(dst, dstFrame, pitch, 1);

+								else

+									Copyeven(dst, dstFrame, pitch, 1);

+							}

+						}

+						else

+						{

+							Copyodd(dstFrame, dst, pitch, 0);

+							Copyeven(dstFrame, dst, pitch, 0);

+						}

+						break;

+

+					case 1:

+						Copyodd(dstFrame, dst, pitch, 0);

+

+						Get_Hdr();

+						Decode_Picture(1, dstFrame, DSTBYTES);

+

+						if (picture_structure!=FRAME_PICTURE)

+						{

+							Get_Hdr();

+							Decode_Picture(1, dstFrame, DSTBYTES);

+						}

+

+						Copyeven(dstFrame, dst, pitch, 0);

+						break;

+

+					case 2:	

+						Copyeven(dstFrame, dst, pitch, 0);

+

+						Get_Hdr();

+						Decode_Picture(1, dstFrame, DSTBYTES);

+

+						if (picture_structure!=FRAME_PICTURE)

+						{

+							Get_Hdr();

+							Decode_Picture(1, dstFrame, DSTBYTES);

+						}

+

+						Copyodd(dstFrame, dst, pitch, 0);

+						break;

+				}

+

+			if (in->VF_GOPSize)

+			{

+				for (i=0; i < in->VF_GOPSize; i++)

+					free(GOPBuffer[i]);

+

+				in->VF_GOPSize = 0;

+			}

+

+			in->VF_GOPNow = in->VF_GOPLimit;

+			in->VF_OldFrame = origin;

+			return;

+		}

+

+		remain = ref;

+		in->VF_OldRef = ref;

+		in->VF_GOPNow = now;

+		Second_Field = 0;

+

+		if (size < in->VF_GOPSize)

+			for (i=0; i < (in->VF_GOPSize - size); i++)

+				free(GOPBuffer[size+i]);

+		else if (size > in->VF_GOPSize)

+			for (i=0; i < (size - in->VF_GOPSize); i++)

+				GOPBuffer[in->VF_GOPSize+i] = reinterpret_cast<unsigned char*>(malloc(in->VF_FrameSize));

+

+		in->VF_GOPSize = size;

+

+		File_Flag = GOPList[now]->file;

+		_lseeki64(Infile[GOPList[now]->file], GOPList[now]->position, SEEK_SET);

+		Initialize_Buffer();

+

+		while (Get_Hdr() && picture_coding_type!=I_TYPE);

+

+		Decode_Picture(0, dst, pitch);

+

+		while (Get_Hdr() && picture_coding_type==B_TYPE);

+

+		if (picture_structure!=FRAME_PICTURE)

+		{

+			Decode_Picture(0, dst, pitch);

+			Get_Hdr();

+		}

+

+		Decode_Picture(1, dst, pitch);

+

+		if (picture_structure!=FRAME_PICTURE)

+		{

+			Get_Hdr();

+			Decode_Picture(1, dst, pitch);

+		}

+

+		Copyodd(dst, GOPBuffer[0], pitch, 1);

+		Copyeven(dst, GOPBuffer[0], pitch, 1);

+

+		while (remain && Get_Hdr())

+		{

+			Decode_Picture(1, dst, pitch);

+

+			if (picture_structure!=FRAME_PICTURE)

+			{

+				Get_Hdr();

+				Decode_Picture(1, dst, pitch);

+			}

+

+			Copyodd(dst, GOPBuffer[ref - remain + 1], pitch, 1);

+			Copyeven(dst, GOPBuffer[ref - remain + 1], pitch, 1);

+

+			remain--;

+		}

+

+		if (!Full_Frame && ref>=(size-2))

+		{

+			Copyodd(dst, dstFrame, pitch, 1);

+			Copyeven(dst, dstFrame, pitch, 1);

+		}

+	}

+	else

+	{

+		remain = ref - in->VF_OldRef;

+

+		if (remain > 0)

+		{

+			in->VF_OldRef = ref;

+

+			while (remain && Get_Hdr())

+			{

+				Decode_Picture(1, dst, pitch);

+

+				if (picture_structure!=FRAME_PICTURE)

+				{

+					Get_Hdr();

+					Decode_Picture(1, dst, pitch);

+				}

+

+				Copyodd(dst, GOPBuffer[ref - remain + 1], pitch, 1);

+				Copyeven(dst, GOPBuffer[ref - remain + 1], pitch, 1);

+

+				remain--;

+			}

+

+			if (!Full_Frame && ref>=(size-2))

+			{

+				Copyodd(dst, dstFrame, pitch, 1);

+				Copyeven(dst, dstFrame, pitch, 1);

+			}

+		}

+	}

+

+	switch (fo)

+	{

+		case 0:

+			Copyodd(GOPBuffer[ref], dst, pitch, 0);

+			Copyeven(GOPBuffer[ref], dst, pitch, 0);

+			break;

+

+		case 1:

+			Copyodd(GOPBuffer[ref-1], dst, pitch, 0);

+			Copyeven(GOPBuffer[ref], dst, pitch, 0);

+			break;

+

+		case 2:

+			Copyodd(GOPBuffer[ref], dst, pitch, 0);

+			Copyeven(GOPBuffer[ref-1], dst, pitch, 0);

+			break;

+	}

+

+	in->VF_OldFrame = origin;

+}

+

+void CMPEG2Dec::Close()

+{

+	int i;

+

+	for(i = 0; i < VF_GOPLimit; i++) free(GOPList[i]);

+	for(i = 0; i < VF_FrameLimit; i++) free(FrameList[i]);

+

+	if (VF_File != NULL)

+		fclose(VF_File);

+

+	while (VF_GOPSize)

+	{

+		VF_GOPSize--;

+		free(GOPBuffer[VF_GOPSize]);

+	}

+

+	while (File_Limit)

+	{

+		File_Limit--;

+		_close(Infile[File_Limit]);

+	}

+

+	for (i=0; i<MAX_FILE_NUMBER; i++)

+		delete [] Infilename[i];

+

+	for (i=0; i<3; i++)

+	{

+		delete [] backward_reference_frame[i];

+		delete [] forward_reference_frame[i];

+		delete [] auxframe[i];

+	}

+

+	delete [] u422;

+	delete [] v422;

+	delete [] u444;

+	delete [] v444;

+	delete [] dstFrame;

+

+	if(Luminance_Flag)

+		delete [] lum;

+

+	for (i=0; i<8; i++)

+		delete [] p_block[i];

+

+	delete [] p_fTempArray;

+

+	if (hLibrary)

+		FreeLibrary(hLibrary);

+}

+

+void CMPEG2Dec::Copyodd(unsigned char *src, unsigned char *dst, int pitch, int forward)

+{

+	int i;

+	int PWIDTH = forward ? (pitch<<1) : DSTBYTES2;

+	int QWIDTH = forward ? DSTBYTES2 : (pitch<<1);

+

+	for (i=0; i<(Clip_Height>>1); i++)

+	{

+		memcpy (dst, src, DSTBYTES);

+		src += PWIDTH;

+		dst += QWIDTH;

+	}

+}

+

+void CMPEG2Dec::Copyeven(unsigned char *src, unsigned char *dst, int pitch, int forward)

+{

+	int i;

+	int PWIDTH = forward ? (pitch<<1) : DSTBYTES2;

+	int QWIDTH = forward ? DSTBYTES2 : (pitch<<1);

+	src += forward ? pitch : DSTBYTES;

+	dst += forward ? DSTBYTES : pitch;

+

+	for (i=0; i<(Clip_Height>>1); i++)

+	{

+		memcpy (dst, src, DSTBYTES);

+		src += PWIDTH;

+		dst += QWIDTH;

+	}

+}

diff --git a/src/filters/source/D2VSource/MPEG2Dec.h b/src/filters/source/D2VSource/MPEG2Dec.h
new file mode 100644
index 000000000..34368208b
--- /dev/null
+++ b/src/filters/source/D2VSource/MPEG2Dec.h
@@ -0,0 +1,304 @@
+#pragma once

+

+#include <windows.h>

+#include <winreg.h>

+#include <stdio.h>

+#include <io.h>

+#include <fcntl.h>

+

+/* code definition */

+#define PICTURE_START_CODE			0x100

+#define SLICE_START_CODE_MIN		0x101

+#define SLICE_START_CODE_MAX		0x1AF

+#define USER_DATA_START_CODE		0x1B2

+#define SEQUENCE_HEADER_CODE		0x1B3

+#define EXTENSION_START_CODE		0x1B5

+#define SEQUENCE_END_CODE			0x1B7

+#define GROUP_START_CODE			0x1B8

+

+#define SYSTEM_END_CODE				0x1B9

+#define PACK_START_CODE				0x1BA

+#define SYSTEM_START_CODE			0x1BB

+#define PRIVATE_STREAM_1			0x1BD

+#define VIDEO_ELEMENTARY_STREAM		0x1E0

+

+/* extension start code IDs */

+#define SEQUENCE_EXTENSION_ID					1

+#define SEQUENCE_DISPLAY_EXTENSION_ID			2

+#define QUANT_MATRIX_EXTENSION_ID				3

+#define COPYRIGHT_EXTENSION_ID					4

+#define PICTURE_DISPLAY_EXTENSION_ID			7

+#define PICTURE_CODING_EXTENSION_ID				8

+

+#define ZIG_ZAG									0

+#define MB_WEIGHT								32

+#define MB_CLASS4								64

+

+#define I_TYPE			1

+#define P_TYPE			2

+#define B_TYPE			3

+

+#define TOP_FIELD		1

+#define BOTTOM_FIELD	2

+#define FRAME_PICTURE	3

+

+#define MACROBLOCK_INTRA				1

+#define MACROBLOCK_PATTERN				2

+#define MACROBLOCK_MOTION_BACKWARD		4

+#define MACROBLOCK_MOTION_FORWARD		8

+#define MACROBLOCK_QUANT				16

+

+#define MC_FIELD		1

+#define MC_FRAME		2

+#define MC_16X8			2

+#define MC_DMV			3

+

+#define MV_FIELD		0

+#define MV_FRAME		1

+

+#define CHROMA420		1

+#define CHROMA422		2

+#define CHROMA444		3

+

+#define BUFFER_SIZE			2048

+#define MAX_FILE_NUMBER		256

+

+#define IDCT_MMX		1

+#define IDCT_SSEMMX		2

+#define	IDCT_FPU		3

+#define IDCT_REF		4

+

+#define FO_NONE			0

+#define FO_FILM			1

+#define FO_SWAP			2

+

+

+typedef void (WINAPI *PBufferOp) (unsigned char*, int, int);

+

+#define MAX_FRAME_NUMBER	1000000

+#define MAX_GOP_SIZE		1024

+

+

+class CMPEG2Dec  

+{

+protected:

+

+  // getbit.cpp

+  void Initialize_Buffer();

+  void Fill_Buffer();

+  void Next_Packet();

+  void Flush_Buffer_All(unsigned int N);

+  unsigned int Get_Bits_All(unsigned int N);

+  void Next_File();

+

+  unsigned int Show_Bits(unsigned int N);

+  unsigned int Get_Bits(unsigned int N);

+  void Flush_Buffer(unsigned int N);

+  void Fill_Next();

+  unsigned int Get_Byte();

+  unsigned int Get_Short();

+  void next_start_code();

+

+  unsigned char Rdbfr[BUFFER_SIZE], *Rdptr, *Rdmax;

+  unsigned int CurrentBfr, NextBfr, BitsLeft, Val, Read;

+

+  // gethdr.cpp

+  int Get_Hdr();

+  void sequence_header();

+  int slice_header();

+private:

+  void group_of_pictures_header();

+  void picture_header();

+  void sequence_extension();

+  void sequence_display_extension();

+  void quant_matrix_extension();

+  void picture_display_extension();

+  void picture_coding_extension();

+  void copyright_extension();

+  int  extra_bit_information();

+  void extension_and_user_data();

+

+protected:

+  // getpic.cpp

+  void Decode_Picture(int ref, unsigned char *dst, int pitch);

+private:

+  void Update_Picture_Buffers();

+  void picture_data();

+  int slice(int MBAmax);

+  void macroblock_modes(int *pmacroblock_type, int *pmotion_type, 

+  	int *pmotion_vector_count, int *pmv_format, int *pdmv, int *pmvscale, int *pdct_type);

+  void Clear_Block(int count);

+  void Add_Block(int count, int bx, int by, int dct_type, int addflag);

+  void motion_compensation(int MBA, int macroblock_type, int motion_type,

+	  int PMV[2][2][2], int motion_vertical_field_select[2][2], int dmvector[2], int dct_type);

+  void skipped_macroblock(int dc_dct_pred[3], int PMV[2][2][2], 

+  	int *motion_type, int motion_vertical_field_select[2][2], int *macroblock_type);

+  int start_of_slice(int *MBA, int *MBAinc, int dc_dct_pred[3], int PMV[2][2][2]);

+  int decode_macroblock(int *macroblock_type, int *motion_type, int *dct_type,

+	  int PMV[2][2][2], int dc_dct_pred[3], int motion_vertical_field_select[2][2], int dmvector[2]);

+  void Decode_MPEG2_Intra_Block(int comp, int dc_dct_pred[]);

+  void Decode_MPEG2_Non_Intra_Block(int comp);

+

+  int Get_macroblock_type();

+  int Get_I_macroblock_type();

+  int Get_P_macroblock_type();

+  int Get_B_macroblock_type();

+  int Get_D_macroblock_type();

+  int Get_coded_block_pattern();

+  int Get_macroblock_address_increment();

+  int Get_Luma_DC_dct_diff();

+  int Get_Chroma_DC_dct_diff();

+

+  void form_predictions(int bx, int by, int macroblock_type, int motion_type, 

+	  int PMV[2][2][2], int motion_vertical_field_select[2][2], int dmvector[2]);

+  void form_prediction(unsigned char *src[], int sfield, unsigned char *dst[], int dfield, 

+	  int lx, int lx2, int w, int h, int x, int y, int dx, int dy, int average_flag);

+  void form_component_prediction(unsigned char *src, unsigned char *dst,

+	  int lx, int lx2, int w, int h, int x, int y, int dx, int dy, int average_flag);

+

+  // motion.cpp

+  void motion_vectors(int PMV[2][2][2], int dmvector[2], int motion_vertical_field_select[2][2], 

+	  int s, int motion_vector_count, int mv_format, 

+	  int h_r_size, int v_r_size, int dmv, int mvscale);

+  void Dual_Prime_Arithmetic(int DMV[][2], int *dmvector, int mvx, int mvy);

+private:

+  void motion_vector(int *PMV, int *dmvector, int h_r_size, int v_r_size, 

+  	int dmv, int mvscale, int full_pel_vector);

+  void decode_motion_vector(int *pred, int r_size, int motion_code,

+  	int motion_residualesidual, int full_pel_vector);

+  int Get_motion_code();

+  int Get_dmvector();

+

+protected:

+  // store.cpp

+  void assembleFrame(unsigned char *src[], int pf, unsigned char *dst, int pitch);

+private:

+  void Luminance_Filter(unsigned char *src, unsigned char *dst);

+  void conv420to422(unsigned char *src, unsigned char *dst, int frame_type);

+  void conv422to444(unsigned char *src, unsigned char *dst);

+  void conv444toRGB24(unsigned char *py, unsigned char *pu, unsigned char *pv, unsigned char *dst, int pitch);

+  void conv422toYUY2(unsigned char *py, unsigned char *pu, unsigned char *pv, unsigned char *dst, int pitch);

+

+protected:

+  // decoder operation control flags

+  int Fault_Flag;

+  int File_Flag;

+  int File_Limit;

+  int FO_Flag;

+  int IDCT_Flag;

+  int SystemStream_Flag;

+

+  int Luminance_Flag;

+  int Resize_Flag;

+

+  int KeyOp_Flag;

+  int lfsr0, lfsr1;

+  PBufferOp BufferOp;

+

+  int Infile[MAX_FILE_NUMBER];

+  char *Infilename[MAX_FILE_NUMBER];

+

+  int intra_quantizer_matrix[64];

+  int non_intra_quantizer_matrix[64];

+  int chroma_intra_quantizer_matrix[64];

+  int chroma_non_intra_quantizer_matrix[64];

+  

+  int load_intra_quantizer_matrix;

+  int load_non_intra_quantizer_matrix;

+  int load_chroma_intra_quantizer_matrix;

+  int load_chroma_non_intra_quantizer_matrix;

+

+  int q_scale_type;

+  int alternate_scan;

+  int quantizer_scale;

+

+  void *fTempArray, *p_fTempArray;

+  short *block[8], *p_block[8];

+  int pf_backward, pf_forward, pf_current;

+

+  // global values

+  unsigned char *backward_reference_frame[3], *forward_reference_frame[3];

+  unsigned char *auxframe[3], *current_frame[3];

+  unsigned char *u422, *v422, *u444, *v444, /* *rgb24,*/ *lum;

+  unsigned char *dstFrame;	// replaces rgb24

+  __int64 RGB_Scale, RGB_Offset, RGB_CRV, RGB_CBU, RGB_CGX, LumOffsetMask, LumGainMask;

+

+  int HALF_WIDTH, PROGRESSIVE_HEIGHT, INTERLACED_HEIGHT, DOUBLE_WIDTH;

+  int /*TWIDTH, SWIDTH,*/ HALF_WIDTH_D8, LUM_AREA, CLIP_AREA, HALF_CLIP_AREA, CLIP_STEP;

+  int DSTBYTES, DSTBYTES2;	// these replace TWIDTH and SWIDTH

+public:

+  int Clip_Width, Clip_Height, Resize_Width, Resize_Height;

+protected:

+

+  int Coded_Picture_Width, Coded_Picture_Height, Chroma_Width, Chroma_Height;

+  int block_count, Second_Field;

+  int horizontal_size, vertical_size, mb_width, mb_height;

+

+  /* ISO/IEC 13818-2 section 6.2.2.3:  sequence_extension() */

+  int progressive_sequence;

+  int chroma_format;

+

+  /* ISO/IEC 13818-2 section 6.2.3: picture_header() */

+  int picture_coding_type;

+  int temporal_reference;

+

+  /* ISO/IEC 13818-2 section 6.2.3.1: picture_coding_extension() header */

+  int f_code[2][2];

+  int picture_structure;

+  int frame_pred_frame_dct;

+  int progressive_frame;

+  int concealment_motion_vectors;

+  int intra_dc_precision;

+  int top_field_first;

+  int repeat_first_field;

+  int intra_vlc_format;

+

+  // interface

+  typedef struct {

+	DWORD		number;

+	int			file;

+	__int64		position;

+  }	GOPLIST;

+  GOPLIST *GOPList[MAX_FRAME_NUMBER];

+

+  typedef struct {

+	DWORD			top;

+	DWORD			bottom;

+	char			forward;

+	char			backward;

+  }	FRAMELIST;

+  FRAMELIST *FrameList[MAX_FRAME_NUMBER];

+

+  unsigned char *GOPBuffer[MAX_GOP_SIZE];

+public:

+  BOOL Field_Order, Full_Frame;

+protected:

+  HINSTANCE hLibrary;

+

+  void Copyodd(unsigned char *src, unsigned char *dst, int pitch, int forward);

+  void Copyeven(unsigned char *src, unsigned char *dst, int pitch, int forward);

+public:

+  FILE		*VF_File;

+  int		VF_FrameRate;

+  DWORD		VF_FrameLimit;

+  DWORD		VF_FrameBound;

+  DWORD		VF_GOPLimit;

+  DWORD		VF_GOPNow;

+  DWORD		VF_GOPSize;

+  int		VF_FrameSize;

+  DWORD		VF_OldFrame;

+  DWORD		VF_OldRef;

+

+  enum DstFormat {

+	RGB24, YUY2

+  };

+  DstFormat m_dstFormat;

+

+  CMPEG2Dec();

+  ~CMPEG2Dec() {Close();}

+  int Open(LPCTSTR path, DstFormat);

+  void Close();

+  void Decode(unsigned char *dst, DWORD frame, int pitch);

+  bool dstRGB24() const { return m_dstFormat == RGB24; }

+  bool dstYUY2() const { return m_dstFormat == YUY2; }

+};

diff --git a/src/filters/source/D2VSource/d2vsource.rc b/src/filters/source/D2VSource/d2vsource.rc
new file mode 100644
index 000000000..6df83bdd8
--- /dev/null
+++ b/src/filters/source/D2VSource/d2vsource.rc
@@ -0,0 +1,117 @@
+// Microsoft Visual C++ generated resource script.

+//

+#include "resource.h"

+#include "..\..\..\..\include\Version.h"

+

+#define APSTUDIO_READONLY_SYMBOLS

+/////////////////////////////////////////////////////////////////////////////

+//

+// Generated from the TEXTINCLUDE 2 resource.

+//

+#include "afxres.h"

+

+/////////////////////////////////////////////////////////////////////////////

+#undef APSTUDIO_READONLY_SYMBOLS

+

+/////////////////////////////////////////////////////////////////////////////

+// Hungarian resources

+

+#if !defined(AFX_RESOURCE_DLL) || defined(AFX_TARG_HUN)

+#ifdef _WIN32

+LANGUAGE LANG_HUNGARIAN, SUBLANG_DEFAULT

+#pragma code_page(1250)

+#endif //_WIN32

+

+/////////////////////////////////////////////////////////////////////////////

+//

+// Version

+//

+

+VS_VERSION_INFO VERSIONINFO

+ FILEVERSION VERSION_MAJOR,VERSION_MINOR,VERSION_REV,VERSION_PATCH

+ PRODUCTVERSION VERSION_MAJOR,VERSION_MINOR,VERSION_REV,VERSION_PATCH

+ FILEFLAGSMASK 0x17L

+#ifdef _DEBUG

+ FILEFLAGS 0x1L

+#else

+ FILEFLAGS 0x0L

+#endif

+ FILEOS 0x4L

+ FILETYPE 0x2L

+ FILESUBTYPE 0x0L

+BEGIN

+    BLOCK "StringFileInfo"

+    BEGIN

+        BLOCK "040e04b0"

+        BEGIN

+            VALUE "Comments", "http://sourceforge.net/projects/mpc-hc/"

+            VALUE "CompanyName", "MPC-HC Team"

+            VALUE "FileDescription", "D2V Source Filter"

+            VALUE "FileVersion", "1, 1, 0, 0"

+            VALUE "InternalName", "D2V Source Filter"

+            VALUE "LegalCopyright", "Copyright (C) 2002-2010 see AUTHORS file"

+            VALUE "OriginalFilename", "D2VSource.ax"

+            VALUE "ProductName", "D2V Source Filter"

+            VALUE "ProductVersion", "1, 1, 0, 0"

+        END

+    END

+    BLOCK "VarFileInfo"

+    BEGIN

+        VALUE "Translation", 0x40e, 1200

+    END

+END

+

+#endif    // Hungarian resources

+/////////////////////////////////////////////////////////////////////////////

+

+

+/////////////////////////////////////////////////////////////////////////////

+// English (U.S.) resources

+

+#if !defined(AFX_RESOURCE_DLL) || defined(AFX_TARG_ENU)

+#ifdef _WIN32

+LANGUAGE LANG_ENGLISH, SUBLANG_ENGLISH_US

+#pragma code_page(1252)

+#endif //_WIN32

+

+#ifdef APSTUDIO_INVOKED

+/////////////////////////////////////////////////////////////////////////////

+//

+// TEXTINCLUDE

+//

+

+1 TEXTINCLUDE 

+BEGIN

+    "resource.h\0"

+END

+

+2 TEXTINCLUDE 

+BEGIN

+    "#include ""afxres.h""\r\n"

+    "\0"

+END

+

+3 TEXTINCLUDE 

+BEGIN

+    "\r\n"

+    "\0"

+END

+

+#endif    // APSTUDIO_INVOKED

+

+#endif    // English (U.S.) resources

+/////////////////////////////////////////////////////////////////////////////

+

+

+

+#ifndef APSTUDIO_INVOKED

+/////////////////////////////////////////////////////////////////////////////

+//

+// Generated from the TEXTINCLUDE 3 resource.

+//

+

+

+/////////////////////////////////////////////////////////////////////////////

+#endif    // not APSTUDIO_INVOKED

+

+

diff --git a/src/filters/source/D2VSource/d2vsource.vcproj b/src/filters/source/D2VSource/d2vsource.vcproj
new file mode 100644
index 000000000..0fcded0c4
--- /dev/null
+++ b/src/filters/source/D2VSource/d2vsource.vcproj
@@ -0,0 +1,962 @@
+<?xml version="1.0" encoding="windows-1250"?>

+<VisualStudioProject

+	ProjectType="Visual C++"

+	Version="9,00"

+	Name="d2vsource"

+	ProjectGUID="{83CC6B88-A112-4192-BD5A-F2A249AF2277}"

+	RootNamespace="d2vsource"

+	Keyword="Win32Proj"

+	TargetFrameworkVersion="131072"

+	>

+	<Platforms>

+		<Platform

+			Name="Win32"

+		/>

+		<Platform

+			Name="x64"

+		/>

+	</Platforms>

+	<ToolFiles>

+	</ToolFiles>

+	<Configurations>

+		<Configuration

+			Name="Debug Unicode|Win32"

+			ConfigurationType="2"

+			InheritedPropertySheets="..\..\..\common.vsprops;..\..\..\debug.vsprops"

+			UseOfMFC="1"

+			CharacterSet="1"

+			>

+			<Tool

+				Name="VCPreBuildEventTool"

+			/>

+			<Tool

+				Name="VCCustomBuildTool"

+			/>

+			<Tool

+				Name="VCXMLDataGeneratorTool"

+			/>

+			<Tool

+				Name="VCWebServiceProxyGeneratorTool"

+			/>

+			<Tool

+				Name="VCMIDLTool"

+			/>

+			<Tool

+				Name="VCCLCompilerTool"

+				AdditionalIncludeDirectories="..\..\..\..\include;..\..\BaseClasses"

+				PreprocessorDefinitions="REGISTER_FILTER;WIN32;_DEBUG;_USRDLL"

+			/>

+			<Tool

+				Name="VCManagedResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCPreLinkEventTool"

+			/>

+			<Tool

+				Name="VCLinkerTool"

+				RegisterOutput="true"

+				AdditionalDependencies="strmbaseDU.lib basesourceDU.lib Winmm.lib"

+				OutputFile="$(OutDir)\$(ProjectName).ax"

+				AdditionalLibraryDirectories="..\..\..\..\lib"

+				ModuleDefinitionFile="D2VSource.def"

+				SubSystem="2"

+				RandomizedBaseAddress="1"

+				DataExecutionPrevention="0"

+				TargetMachine="1"

+			/>

+			<Tool

+				Name="VCALinkTool"

+			/>

+			<Tool

+				Name="VCManifestTool"

+			/>

+			<Tool

+				Name="VCXDCMakeTool"

+			/>

+			<Tool

+				Name="VCBscMakeTool"

+			/>

+			<Tool

+				Name="VCFxCopTool"

+			/>

+			<Tool

+				Name="VCAppVerifierTool"

+			/>

+			<Tool

+				Name="VCPostBuildEventTool"

+			/>

+		</Configuration>

+		<Configuration

+			Name="Debug Unicode|x64"

+			OutputDirectory="$(PlatformName)\$(ConfigurationName)"

+			IntermediateDirectory="$(PlatformName)\$(ConfigurationName)"

+			ConfigurationType="2"

+			InheritedPropertySheets="..\..\..\common.vsprops;..\..\..\debug.vsprops"

+			UseOfMFC="1"

+			CharacterSet="1"

+			>

+			<Tool

+				Name="VCPreBuildEventTool"

+			/>

+			<Tool

+				Name="VCCustomBuildTool"

+			/>

+			<Tool

+				Name="VCXMLDataGeneratorTool"

+			/>

+			<Tool

+				Name="VCWebServiceProxyGeneratorTool"

+			/>

+			<Tool

+				Name="VCMIDLTool"

+				TargetEnvironment="3"

+			/>

+			<Tool

+				Name="VCCLCompilerTool"

+				AdditionalIncludeDirectories="..\..\..\..\include;..\..\BaseClasses"

+				PreprocessorDefinitions="REGISTER_FILTER;WIN32;_DEBUG;_USRDLL"

+				DebugInformationFormat="3"

+			/>

+			<Tool

+				Name="VCManagedResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCPreLinkEventTool"

+			/>

+			<Tool

+				Name="VCLinkerTool"

+				AdditionalDependencies="strmbaseDU.lib basesourceDU.lib Winmm.lib"

+				OutputFile="$(OutDir)\$(ProjectName).ax"

+				AdditionalLibraryDirectories="..\..\..\..\lib64"

+				ModuleDefinitionFile="D2VSource.def"

+				SubSystem="2"

+				RandomizedBaseAddress="1"

+				DataExecutionPrevention="0"

+				TargetMachine="17"

+			/>

+			<Tool

+				Name="VCALinkTool"

+			/>

+			<Tool

+				Name="VCManifestTool"

+			/>

+			<Tool

+				Name="VCXDCMakeTool"

+			/>

+			<Tool

+				Name="VCBscMakeTool"

+			/>

+			<Tool

+				Name="VCFxCopTool"

+			/>

+			<Tool

+				Name="VCAppVerifierTool"

+			/>

+			<Tool

+				Name="VCPostBuildEventTool"

+			/>

+		</Configuration>

+		<Configuration

+			Name="Release Unicode|Win32"

+			ConfigurationType="2"

+			InheritedPropertySheets="..\..\..\common.vsprops;..\..\..\release.vsprops"

+			UseOfMFC="1"

+			CharacterSet="1"

+			>

+			<Tool

+				Name="VCPreBuildEventTool"

+			/>

+			<Tool

+				Name="VCCustomBuildTool"

+			/>

+			<Tool

+				Name="VCXMLDataGeneratorTool"

+			/>

+			<Tool

+				Name="VCWebServiceProxyGeneratorTool"

+			/>

+			<Tool

+				Name="VCMIDLTool"

+			/>

+			<Tool

+				Name="VCCLCompilerTool"

+				AdditionalIncludeDirectories="..\..\..\..\include;..\..\BaseClasses"

+				PreprocessorDefinitions="REGISTER_FILTER;WIN32;NDEBUG;_USRDLL"

+				BufferSecurityCheck="true"

+				EnableEnhancedInstructionSet="1"

+			/>

+			<Tool

+				Name="VCManagedResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCPreLinkEventTool"

+			/>

+			<Tool

+				Name="VCLinkerTool"

+				RegisterOutput="true"

+				AdditionalDependencies="strmbaseRU.lib basesourceRU.lib Winmm.lib"

+				OutputFile="..\..\..\..\bin\x86\$(ProjectName).ax"

+				AdditionalLibraryDirectories="..\..\..\..\lib"

+				ModuleDefinitionFile="D2VSource.def"

+				GenerateDebugInformation="true"

+				SubSystem="2"

+				LargeAddressAware="2"

+				RandomizedBaseAddress="2"

+				DataExecutionPrevention="2"

+				TargetMachine="1"

+			/>

+			<Tool

+				Name="VCALinkTool"

+			/>

+			<Tool

+				Name="VCManifestTool"

+			/>

+			<Tool

+				Name="VCXDCMakeTool"

+			/>

+			<Tool

+				Name="VCBscMakeTool"

+			/>

+			<Tool

+				Name="VCFxCopTool"

+			/>

+			<Tool

+				Name="VCAppVerifierTool"

+			/>

+			<Tool

+				Name="VCPostBuildEventTool"

+			/>

+		</Configuration>

+		<Configuration

+			Name="Release Unicode|x64"

+			OutputDirectory="$(PlatformName)\$(ConfigurationName)"

+			IntermediateDirectory="$(PlatformName)\$(ConfigurationName)"

+			ConfigurationType="2"

+			InheritedPropertySheets="..\..\..\common.vsprops;..\..\..\release.vsprops"

+			UseOfMFC="1"

+			CharacterSet="1"

+			>

+			<Tool

+				Name="VCPreBuildEventTool"

+			/>

+			<Tool

+				Name="VCCustomBuildTool"

+			/>

+			<Tool

+				Name="VCXMLDataGeneratorTool"

+			/>

+			<Tool

+				Name="VCWebServiceProxyGeneratorTool"

+			/>

+			<Tool

+				Name="VCMIDLTool"

+				TargetEnvironment="3"

+			/>

+			<Tool

+				Name="VCCLCompilerTool"

+				AdditionalIncludeDirectories="..\..\..\..\include;..\..\BaseClasses"

+				PreprocessorDefinitions="REGISTER_FILTER;WIN32;NDEBUG;_USRDLL"

+				BufferSecurityCheck="true"

+			/>

+			<Tool

+				Name="VCManagedResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCPreLinkEventTool"

+			/>

+			<Tool

+				Name="VCLinkerTool"

+				AdditionalDependencies="strmbaseRU.lib basesourceRU.lib Winmm.lib"

+				OutputFile="..\..\..\..\bin\x64\$(ProjectName).ax"

+				AdditionalLibraryDirectories="..\..\..\..\lib64"

+				ModuleDefinitionFile="D2VSource.def"

+				GenerateDebugInformation="true"

+				SubSystem="2"

+				LargeAddressAware="2"

+				RandomizedBaseAddress="2"

+				DataExecutionPrevention="2"

+				TargetMachine="17"

+			/>

+			<Tool

+				Name="VCALinkTool"

+			/>

+			<Tool

+				Name="VCManifestTool"

+			/>

+			<Tool

+				Name="VCXDCMakeTool"

+			/>

+			<Tool

+				Name="VCBscMakeTool"

+			/>

+			<Tool

+				Name="VCFxCopTool"

+			/>

+			<Tool

+				Name="VCAppVerifierTool"

+			/>

+			<Tool

+				Name="VCPostBuildEventTool"

+			/>

+		</Configuration>

+		<Configuration

+			Name="Debug Unicode lib|Win32"

+			ConfigurationType="4"

+			InheritedPropertySheets="..\..\..\common.vsprops;..\..\..\debug.vsprops"

+			UseOfMFC="1"

+			CharacterSet="1"

+			>

+			<Tool

+				Name="VCPreBuildEventTool"

+			/>

+			<Tool

+				Name="VCCustomBuildTool"

+			/>

+			<Tool

+				Name="VCXMLDataGeneratorTool"

+			/>

+			<Tool

+				Name="VCWebServiceProxyGeneratorTool"

+			/>

+			<Tool

+				Name="VCMIDLTool"

+			/>

+			<Tool

+				Name="VCCLCompilerTool"

+				AdditionalOptions="/MP"

+				AdditionalIncludeDirectories="..\..\..\..\include;..\..\BaseClasses"

+				PreprocessorDefinitions="WIN32;_DEBUG"

+			/>

+			<Tool

+				Name="VCManagedResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCPreLinkEventTool"

+			/>

+			<Tool

+				Name="VCLibrarianTool"

+				OutputFile="..\..\..\..\lib\$(ProjectName)DU.lib"

+			/>

+			<Tool

+				Name="VCALinkTool"

+			/>

+			<Tool

+				Name="VCXDCMakeTool"

+			/>

+			<Tool

+				Name="VCBscMakeTool"

+			/>

+			<Tool

+				Name="VCFxCopTool"

+			/>

+			<Tool

+				Name="VCPostBuildEventTool"

+			/>

+		</Configuration>

+		<Configuration

+			Name="Debug Unicode lib|x64"

+			OutputDirectory="$(PlatformName)\$(ConfigurationName)"

+			IntermediateDirectory="$(PlatformName)\$(ConfigurationName)"

+			ConfigurationType="4"

+			InheritedPropertySheets="..\..\..\common.vsprops;..\..\..\debug.vsprops"

+			UseOfMFC="1"

+			CharacterSet="1"

+			>

+			<Tool

+				Name="VCPreBuildEventTool"

+			/>

+			<Tool

+				Name="VCCustomBuildTool"

+			/>

+			<Tool

+				Name="VCXMLDataGeneratorTool"

+			/>

+			<Tool

+				Name="VCWebServiceProxyGeneratorTool"

+			/>

+			<Tool

+				Name="VCMIDLTool"

+				TargetEnvironment="3"

+			/>

+			<Tool

+				Name="VCCLCompilerTool"

+				AdditionalOptions="/MP"

+				AdditionalIncludeDirectories="..\..\..\..\include;..\..\BaseClasses"

+				PreprocessorDefinitions="_WIN64;_DEBUG"

+				DebugInformationFormat="3"

+			/>

+			<Tool

+				Name="VCManagedResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCPreLinkEventTool"

+			/>

+			<Tool

+				Name="VCLibrarianTool"

+				OutputFile="..\..\..\..\lib64\$(ProjectName)DU.lib"

+			/>

+			<Tool

+				Name="VCALinkTool"

+			/>

+			<Tool

+				Name="VCXDCMakeTool"

+			/>

+			<Tool

+				Name="VCBscMakeTool"

+			/>

+			<Tool

+				Name="VCFxCopTool"

+			/>

+			<Tool

+				Name="VCPostBuildEventTool"

+			/>

+		</Configuration>

+		<Configuration

+			Name="Release Unicode lib|Win32"

+			ConfigurationType="4"

+			InheritedPropertySheets="..\..\..\common.vsprops;..\..\..\release.vsprops"

+			UseOfMFC="1"

+			CharacterSet="1"

+			>

+			<Tool

+				Name="VCPreBuildEventTool"

+			/>

+			<Tool

+				Name="VCCustomBuildTool"

+			/>

+			<Tool

+				Name="VCXMLDataGeneratorTool"

+			/>

+			<Tool

+				Name="VCWebServiceProxyGeneratorTool"

+			/>

+			<Tool

+				Name="VCMIDLTool"

+			/>

+			<Tool

+				Name="VCCLCompilerTool"

+				AdditionalOptions="/MP"

+				AdditionalIncludeDirectories="..\..\..\..\include;..\..\BaseClasses"

+				PreprocessorDefinitions="WIN32;NDEBUG"

+				BufferSecurityCheck="true"

+				EnableEnhancedInstructionSet="1"

+				DisableSpecificWarnings="4244;4799;4731;"

+			/>

+			<Tool

+				Name="VCManagedResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCPreLinkEventTool"

+			/>

+			<Tool

+				Name="VCLibrarianTool"

+				OutputFile="..\..\..\..\lib\$(ProjectName)RU.lib"

+			/>

+			<Tool

+				Name="VCALinkTool"

+			/>

+			<Tool

+				Name="VCXDCMakeTool"

+			/>

+			<Tool

+				Name="VCBscMakeTool"

+			/>

+			<Tool

+				Name="VCFxCopTool"

+			/>

+			<Tool

+				Name="VCPostBuildEventTool"

+			/>

+		</Configuration>

+		<Configuration

+			Name="Release Unicode lib|x64"

+			OutputDirectory="$(PlatformName)\$(ConfigurationName)"

+			IntermediateDirectory="$(PlatformName)\$(ConfigurationName)"

+			ConfigurationType="4"

+			InheritedPropertySheets="..\..\..\common.vsprops;..\..\..\release.vsprops"

+			UseOfMFC="1"

+			CharacterSet="1"

+			>

+			<Tool

+				Name="VCPreBuildEventTool"

+			/>

+			<Tool

+				Name="VCCustomBuildTool"

+			/>

+			<Tool

+				Name="VCXMLDataGeneratorTool"

+			/>

+			<Tool

+				Name="VCWebServiceProxyGeneratorTool"

+			/>

+			<Tool

+				Name="VCMIDLTool"

+				TargetEnvironment="3"

+			/>

+			<Tool

+				Name="VCCLCompilerTool"

+				AdditionalOptions="/MP"

+				AdditionalIncludeDirectories="..\..\..\..\include;..\..\BaseClasses"

+				PreprocessorDefinitions="_WIN64;NDEBUG"

+				BufferSecurityCheck="true"

+				EnableEnhancedInstructionSet="0"

+			/>

+			<Tool

+				Name="VCManagedResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCResourceCompilerTool"

+			/>

+			<Tool

+				Name="VCPreLinkEventTool"

+			/>

+			<Tool

+				Name="VCLibrarianTool"

+				OutputFile="..\..\..\..\lib64\$(ProjectName)RU.lib"

+			/>

+			<Tool

+				Name="VCALinkTool"

+			/>

+			<Tool

+				Name="VCXDCMakeTool"

+			/>

+			<Tool

+				Name="VCBscMakeTool"

+			/>

+			<Tool

+				Name="VCFxCopTool"

+			/>

+			<Tool

+				Name="VCPostBuildEventTool"

+			/>

+		</Configuration>

+	</Configurations>

+	<References>

+	</References>

+	<Files>

+		<Filter

+			Name="Source Files"

+			Filter="cpp;c;cxx;def;odl;idl;hpj;bat;asm"

+			>

+			<File

+				RelativePath="D2VSource.cpp"

+				>

+				<FileConfiguration

+					Name="Debug Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+			</File>

+			<File

+				RelativePath="D2VSource.def"

+				>

+			</File>

+			<File

+				RelativePath="idctfpu.cpp"

+				>

+				<FileConfiguration

+					Name="Debug Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+			</File>

+			<File

+				RelativePath="idctmmx.asm"

+				>

+				<FileConfiguration

+					Name="Debug Unicode|Win32"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+						CommandLine="ml /c /coff /Cx /nologo /Fo&quot;$(OutDir)\$(InputName).obj&quot; &quot;$(InputPath)&#x0D;&#x0A;"

+						Outputs="$(OutDir)\$(InputName).obj"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+						CommandLine="ml /c /coff /Cx /nologo /Fo&quot;$(OutDir)\$(InputName).obj&quot; &quot;$(InputPath)&#x0D;&#x0A;"

+						Outputs="$(OutDir)\$(InputName).obj"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode|Win32"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+						CommandLine="ml /c /coff /Cx /nologo /Fo&quot;$(OutDir)\$(InputName).obj&quot; &quot;$(InputPath)&#x0D;&#x0A;"

+						Outputs="$(OutDir)\$(InputName).obj"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+						CommandLine="ml /c /coff /Cx /nologo /Fo&quot;$(OutDir)\$(InputName).obj&quot; &quot;$(InputPath)&#x0D;&#x0A;"

+						Outputs="$(OutDir)\$(InputName).obj"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode lib|Win32"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+						CommandLine="ml /c /coff /Cx /nologo /Fo&quot;$(OutDir)\$(InputName).obj&quot; &quot;$(InputPath)&#x0D;&#x0A;"

+						Outputs="$(OutDir)\$(InputName).obj"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+						CommandLine="ml /c /coff /Cx /nologo /Fo&quot;$(OutDir)\$(InputName).obj&quot; &quot;$(InputPath)&#x0D;&#x0A;"

+						Outputs="$(OutDir)\$(InputName).obj"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|Win32"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+						CommandLine="ml /c /coff /Cx /nologo /Fo&quot;$(OutDir)\$(InputName).obj&quot; &quot;$(InputPath)&#x0D;&#x0A;"

+						Outputs="$(OutDir)\$(InputName).obj"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+						CommandLine="ml /c /coff /Cx /nologo /Fo&quot;$(OutDir)\$(InputName).obj&quot; &quot;$(InputPath)&#x0D;&#x0A;"

+						Outputs="$(OutDir)\$(InputName).obj"

+					/>

+				</FileConfiguration>

+			</File>

+			<File

+				RelativePath="idctref.cpp"

+				>

+				<FileConfiguration

+					Name="Debug Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+			</File>

+			<File

+				RelativePath="MPEG2Dec.cpp"

+				>

+				<FileConfiguration

+					Name="Debug Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+					/>

+				</FileConfiguration>

+			</File>

+			<File

+				RelativePath="stdafx.cpp"

+				>

+				<FileConfiguration

+					Name="Debug Unicode|Win32"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+						UsePrecompiledHeader="1"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+						UsePrecompiledHeader="1"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode|Win32"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+						UsePrecompiledHeader="1"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+						UsePrecompiledHeader="1"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode lib|Win32"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+						UsePrecompiledHeader="1"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode lib|x64"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+						UsePrecompiledHeader="1"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|Win32"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+						UsePrecompiledHeader="1"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|x64"

+					>

+					<Tool

+						Name="VCCLCompilerTool"

+						UsePrecompiledHeader="1"

+					/>

+				</FileConfiguration>

+			</File>

+		</Filter>

+		<Filter

+			Name="Header Files"

+			Filter="h;hpp;hxx;hm;inl;inc"

+			>

+			<File

+				RelativePath="D2VSource.h"

+				>

+			</File>

+			<File

+				RelativePath="MPEG2Dec.h"

+				>

+			</File>

+			<File

+				RelativePath=".\resource.h"

+				>

+				<FileConfiguration

+					Name="Debug Unicode lib|Win32"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|Win32"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCCustomBuildTool"

+					/>

+				</FileConfiguration>

+			</File>

+			<File

+				RelativePath="stdafx.h"

+				>

+			</File>

+		</Filter>

+		<Filter

+			Name="Resource Files"

+			Filter="rc;ico;cur;bmp;dlg;rc2;rct;bin;rgs;gif;jpg;jpeg;jpe"

+			>

+			<File

+				RelativePath=".\d2vsource.rc"

+				>

+				<FileConfiguration

+					Name="Debug Unicode lib|Win32"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCResourceCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Debug Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCResourceCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|Win32"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCResourceCompilerTool"

+					/>

+				</FileConfiguration>

+				<FileConfiguration

+					Name="Release Unicode lib|x64"

+					ExcludedFromBuild="true"

+					>

+					<Tool

+						Name="VCResourceCompilerTool"

+					/>

+				</FileConfiguration>

+			</File>

+		</Filter>

+	</Files>

+	<Globals>

+		<Global

+			Name="DevPartner_IsInstrumented"

+			Value="0"

+		/>

+	</Globals>

+</VisualStudioProject>

diff --git a/src/filters/source/D2VSource/idctfpu.cpp b/src/filters/source/D2VSource/idctfpu.cpp
new file mode 100644
index 000000000..92fcd3954
--- /dev/null
+++ b/src/filters/source/D2VSource/idctfpu.cpp
@@ -0,0 +1,456 @@
+#include "stdafx.h"

+

+/* idct.c, inverse fast discrete cosine transform                           */

+

+

+/*************************************************************/

+/* inverse two dimensional DCT, Chen-Wang algorithm          */

+/* (cf. IEEE ASSP-32, pp. 803-816, Aug. 1984)                */

+/*                                                           */

+/* floating point conversion by Miha Peternel                */

+/* x87 hand-optimized assembly by Miha Peternel              */

+/*                                    27.11. - 11.12.2000    */

+/*                                                           */

+/* You are free to use this code in your project if:         */

+/* - no changes are made to this message                     */

+/* - any changes to this code are publicly available         */

+/* - your project documentation contains the following text: */

+/*   "This software contains fast high-quality IDCT decoder  */

+/*    by Miha Peternel."                                     */

+/*                                                           */

+/*************************************************************/

+

+

+/////////////////////////////////////////////////////

+//

+// TODO:

+// - loops can be easily vectorized for SIMD

+//

+/////////////////////////////////////////////////////

+

+#include <math.h>

+#  define PI 3.1415926535897932384626433832795

+

+#define FLOAT double

+

+const static double RC = 1.0*1024*1024*1024*1024*256*16 + 1024; // magic + clip center

+

+static FLOAT W1; // /* sqrt(2)*cos(1*pi/16) */

+static FLOAT W2; // /* sqrt(2)*cos(2*pi/16) */

+static FLOAT W5; // /* sqrt(2)*cos(5*pi/16) */

+

+static FLOAT W1_8;

+static FLOAT W2_8;

+static FLOAT W5_8;

+

+static FLOAT W7; // /* sqrt(2)*cos(7*pi/16) */

+static FLOAT W1mW7; // W1-W7

+static FLOAT W1pW7; // W1+W7

+

+static FLOAT W3; // /* sqrt(2)*cos(3*pi/16) */

+static FLOAT W3mW5; // W3-W5

+static FLOAT W3pW5; // W3+W5

+

+static FLOAT W6; // /* sqrt(2)*cos(6*pi/16) */

+static FLOAT W2mW6; // W2-W6

+static FLOAT W2pW6; // W2+W6

+

+static FLOAT S2; // 1/sqrt(2)

+static FLOAT D8 = 1.0/8;

+

+static FLOAT W7_8;

+static FLOAT W1mW7_8;

+static FLOAT W1pW7_8;

+

+static FLOAT W3_8;

+static FLOAT W3mW5_8;

+static FLOAT W3pW5_8;

+

+static FLOAT W6_8;

+static FLOAT W2mW6_8;

+static FLOAT W2pW6_8;

+

+/* global declarations */

+

+/* private data */

+static short iclip[1024+1024]; /* clipping table */

+static short *iclp;

+

+void Initialize_FPU_IDCT()

+{

+  int i;

+

+  S2 = sqrt(0.5); // 1.0/sqrt(2);

+

+  W1 = sqrt(2.0)*cos(PI*(1.0/16)); 

+	W1_8 = W1/8;

+  W2 = sqrt(2.0)*cos(PI*(2.0/16)); 

+	W2_8 = W2/8;

+  W3 = sqrt(2.0)*cos(PI*(3.0/16)); 

+	W3_8 = W3/8;

+  W5 = sqrt(2.0)*cos(PI*(5.0/16)); 

+	W5_8 = W5/8;

+  W6 = sqrt(2.0)*cos(PI*(6.0/16)); 

+	W6_8 = W6/8;

+  W7 = sqrt(2.0)*cos(PI*(7.0/16));

+	W7_8 = W7/8;

+

+  W1mW7 = W1-W7;  W1mW7_8 = W1mW7/8;

+  W1pW7 = W1+W7;  W1pW7_8 = W1pW7/8;

+  W3mW5 = W3-W5;  W3mW5_8 = W3mW5/8;

+  W3pW5 = W3+W5;  W3pW5_8 = W3pW5/8;

+  W2mW6 = W2-W6;  W2mW6_8 = W2mW6/8;

+  W2pW6 = W2+W6;  W2pW6_8 = W2pW6/8;

+

+  iclp = iclip+1024;

+  for (i= -1024; i<1024; i++)

+    iclp[i] = (i<-256) ? -256 : ((i>255) ? 255 : i);

+}

+

+void FPU_IDCT(short *block)

+{

+	int *b = (int *) block;

+  if( b[0]==0 && (b[31]==0x10000 || b[31]==0) )

+	{

+	  if( b[ 1]|b[ 2]|b[ 3]|b[ 4]|b[ 5] )

+		  goto normal;

+	  if( b[ 6]|b[ 7]|b[ 8]|b[ 9]|b[10] )

+		  goto normal;

+	  if( b[11]|b[12]|b[13]|b[14]|b[15] )

+		  goto normal;

+	  if( b[16]|b[17]|b[18]|b[19]|b[20] )

+		  goto normal;

+	  if( b[21]|b[22]|b[23]|b[24]|b[25] )

+		  goto normal;

+	  if( b[26]|b[27]|b[28]|b[29]|b[30] )

+		  goto normal;

+		b[31]=0;

+		////empty++;

+		return;

+	}

+normal:

+

+#define tmp  ebx

+#define tmp1 ebx-1*8

+#define tmp2 ebx-2*8

+#define tmp3 ebx-3*8

+#define int0 ebx-3*8-1*4

+#define int1 ebx-3*8-2*4

+#define int2 ebx-3*8-3*4

+#define int3 ebx-3*8-4*4

+#define int4 ebx-3*8-5*4

+#define int5 ebx-3*8-6*4

+#define int6 ebx-3*8-7*4

+#define int7 ebx-3*8-8*4

+#define SIZE 8*8*8+3*8+8*4+16 // locals + 16-byte alignment area

+	__asm

+	{

+	  lea ebx,[esp-8*8*8]

+		sub esp,SIZE

+		and ebx,-16 // force 16-byte alignment of locals

+

+// rows

+		mov esi,[block]

+		lea edi,[tmp]

+		mov ecx,8

+

+		align 16

+Lrows:

+    movsx eax,word ptr [esi+2]

+		or    eax,         [esi+4]

+		or    eax,         [esi+8]

+		or    eax,         [esi+12]

+		jnz L1

+

+		fild word ptr [esi+0*2]

+		fst  qword ptr [edi+7*8]

+		fst  qword ptr [edi+6*8]

+		fst  qword ptr [edi+5*8]

+		fst  qword ptr [edi+4*8]

+		fst  qword ptr [edi+3*8]

+		fst  qword ptr [edi+2*8]

+		fst  qword ptr [edi+1*8]

+		fstp qword ptr [edi+0*8]

+		jmp L2

+

+		align 16

+	L1:

+

+		fild word ptr [esi+7*2]

+		fld st(0)

+		fild word ptr [esi+1*2]

+		fadd st(1),st(0)

+		fld qword ptr [W7]

+		fxch st(1)

+		fmul qword ptr [W1mW7]

+		fxch st(1)

+		fmulp st(2),st(0)

+		fadd st(0),st(1)

+		fstp qword ptr [tmp1]

+		fild word ptr [esi+3*2]

+		fld st(0)

+		fxch st(3)

+		fmul qword ptr [W1pW7]

+		fild word ptr [esi+5*2]

+		fadd st(4),st(0)

+		fmul qword ptr [W3mW5]

+		fxch st(1)

+		fsubp st(3),st(0)//fsubrp

+		fld qword ptr [W3]

+		fmulp st(4),st(0)

+		fsubr st(0),st(3)

+		fstp qword ptr [tmp2]

+		fmul qword ptr [W3pW5]

+		fsubp st(2),st(0)//fsubrp

+		fxch st(1)

+		fstp qword ptr [tmp3]

+		fild word ptr [esi+0*2]

+		fild word ptr [esi+4*2]

+		fild word ptr [esi+2*2]

+		fld st(0)

+		fmul qword ptr [W2mW6]

+		fld st(3)

+		fild word ptr [esi+6*2]

+		fxch st(5)

+		fsub st(0),st(4)

+		fxch st(3)

+		fadd st(0),st(5)

+		fxch st(1)

+		faddp st(4),st(0)

+		fld qword ptr [W6]

+		fmulp st(1),st(0)

+		fxch st(4)

+		fmul qword ptr [W2pW6]

+		fld qword ptr [tmp1]

+		fsub qword ptr [tmp2]

+		fld st(5)

+		fxch st(3)

+		faddp st(6),st(0)

+		fld qword ptr [tmp1]

+		fxch st(1)

+		fstp qword ptr [tmp1]

+		fld st(6)

+		fadd qword ptr [tmp3]

+		fxch st(1)

+		fadd qword ptr [tmp2]

+		fxch st(7)

+		fsub qword ptr [tmp3]

+		fxch st(1)

+		fstp qword ptr [tmp2]

+		fld st(4)

+		fxch st(3)

+		fsubrp st(2),st(0)//fsubp

+		fxch st(4)

+		fsub st(0),st(5)

+		fxch st(2)

+		faddp st(5),st(0)

+		fld st(2)

+		fsub st(0),st(1)

+		fxch st(5)

+		fstp qword ptr [tmp3]

+		fld qword ptr [tmp1]

+		fld qword ptr [S2]

+		fxch st(4)

+		faddp st(2),st(0)

+		fld st(3)

+		fxch st(1)

+		fadd st(0),st(5)

+		fmulp st(1),st(0)

+

+		fld qword ptr [tmp3]

+		fadd st(0),st(7)

+		fxch st(5)

+		fsubr qword ptr [tmp1]

+		fxch st(5)

+		fstp qword ptr [edi+0*8]

+		fxch st(6)

+		fsubr qword ptr [tmp3]

+		fld st(2)

+		fxch st(1)

+		fstp qword ptr [edi+7*8]

+		fadd qword ptr [tmp2]

+		fxch st(3)

+		fmulp st(4),st(0)

+		fxch st(2)

+		fstp qword ptr [edi+3*8]

+		fld st(1)

+		fadd st(0),st(5)

+		fxch st(1)

+		fsub qword ptr [tmp2]

+		fxch st(2)

+		fsubrp st(5),st(0)//fsubp

+		fstp qword ptr [edi+1*8]

+		fld st(2)

+		fxch st(1)

+		fstp qword ptr [edi+4*8]

+		fxch st(2)

+		fsub st(0),st(1)

+		fxch st(2)

+		faddp st(1),st(0)

+		fxch st(2)

+		fstp qword ptr [edi+6*8]

+		fstp qword ptr [edi+5*8]

+		fstp qword ptr [edi+2*8]

+	L2:

+	  add esi,8*2

+		add edi,8*8

+		dec ecx

+		jnz Lrows

+

+// columns

+    lea esi,[tmp]

+		mov edi,[block]

+		lea edx,[iclip+1024*2]

+		mov ecx,8

+

+    align 16

+Lcols:

+		fld qword ptr [esi+7*8*8]

+		fld st(0)

+		fld qword ptr [esi+1*8*8]

+		fadd st(1),st(0)

+		fld qword ptr [W7_8]

+		fxch st(1)

+		fmul qword ptr [W1mW7_8]

+		fxch st(1)

+		fmulp st(2),st(0)

+		fadd st(0),st(1)

+		fstp qword ptr [tmp2]

+		fld qword ptr [esi+3*8*8]

+		fld st(0)

+		fxch st(3)

+		fmul qword ptr [W1pW7_8]

+		fld qword ptr [esi+5*8*8]

+		fadd st(4),st(0)

+		fmul qword ptr [W3mW5_8]

+		fxch st(1)

+		fsubp st(3),st(0)//fsubrp

+		fld qword ptr [W3_8]

+		fmulp st(4),st(0)

+		fsubr st(0),st(3)

+		fstp qword ptr [tmp3]

+		fld qword ptr [D8]

+		fld qword ptr [esi+0*8*8]

+		fmul st(0),st(1)

+		fxch st(2)

+		fmul qword ptr [W3pW5_8]

+		fld qword ptr [esi+4*8*8]

+		fmulp st(2),st(0)

+		fld qword ptr [esi+6*8*8]

+		fld st(3)

+		fxch st(6)

+		fsubrp st(2),st(0)//fsubp

+		fld qword ptr [esi+2*8*8]

+		fld st(0)

+		fxch st(5)

+		fsub st(0),st(4)

+		fxch st(7)

+		faddp st(4),st(0)

+		fxch st(4)

+		fadd st(0),st(1)

+		fld qword ptr [W6_8]

+		fxch st(2)

+		fmul qword ptr [W2pW6_8]

+		fxch st(2)

+		fmulp st(1),st(0)

+		fxch st(4)

+		fmul qword ptr [W2mW6_8]

+		fld qword ptr [tmp2]

+		fsub qword ptr [tmp3]

+		fxch st(2)

+		fsubr st(0),st(5)

+		fxch st(1)

+		faddp st(5),st(0)

+		fld qword ptr [tmp2]

+		fxch st(2)

+		fstp qword ptr [tmp2]

+		fld st(5)

+		fxch st(2)

+		fadd qword ptr [tmp3]

+		fxch st(6)

+		fsub st(0),st(3)

+		fxch st(2)

+		faddp st(3),st(0)

+		fld st(3)

+		fsub st(0),st(5)

+		fxch st(3)

+		fstp qword ptr [tmp3]

+		fxch st(3)

+		faddp st(4),st(0)

+		fld st(5)

+		fld qword ptr [tmp2]

+		fxch st(7)

+		fsub st(0),st(4)

+		fxch st(7)

+		fadd st(0),st(2)

+		fxch st(1)

+		faddp st(4),st(0)

+		fld qword ptr [S2]

+		fmul st(1),st(0)

+		fxch st(1)

+		fstp qword ptr [tmp1]

+		fld st(4)

+		fadd st(0),st(6)

+		fxch st(2)

+		fsubr qword ptr [tmp2]

+		fxch st(5)

+		fsubrp st(6),st(0)//fsubp

+		fxch st(1)

+		fistp dword ptr [int0]

+		fxch st(4)

+		mov eax,[int0]

+		movsx eax,word ptr [edx+2*eax]

+		mov [edi+0*8*2],ax

+		fistp dword ptr [int7]

+		mov eax,[int7]

+		fld st(0)

+		movsx eax,word ptr [edx+2*eax]

+		mov [edi+7*8*2],ax

+		fadd qword ptr [tmp3]

+		fistp dword ptr [int3]

+		mov eax,[int3]

+		movsx eax,word ptr [edx+2*eax]

+		mov [edi+3*8*2],ax

+		fsub qword ptr [tmp3]

+		fld st(1)

+		fxch st(1)

+		fistp dword ptr [int4]

+		mov eax,[int4]

+		movsx eax,word ptr [edx+2*eax]

+		mov [edi+4*8*2],ax

+		fadd qword ptr [tmp1]

+		fxch st(3)

+		fmulp st(2),st(0)

+		fxch st(2)

+		fistp dword ptr [int1]

+		fxch st(1)

+		mov eax,[int1]

+		movsx eax,word ptr [edx+2*eax]

+		mov [edi+1*8*2],ax

+		fsub qword ptr [tmp1]

+		fld st(2)

+		fsub st(0),st(2)

+		fxch st(1)

+		fistp dword ptr [int6]

+		fxch st(2)

+		mov eax,[int6]

+		faddp st(1),st(0)

+		movsx eax,word ptr [edx+2*eax]

+		mov [edi+6*8*2],ax

+		fistp dword ptr [int2]

+		mov eax,[int2]

+		movsx eax,word ptr [edx+2*eax]

+		mov [edi+2*8*2],ax

+		fistp dword ptr [int5]

+		mov eax,[int5]

+		movsx eax,word ptr [edx+2*eax]

+		mov [edi+5*8*2],ax

+

+	  add esi,8

+		add edi,2

+		dec ecx

+		jnz Lcols

+

+		add esp,SIZE

+  }

+}

diff --git a/src/filters/source/D2VSource/idctmmx.asm b/src/filters/source/D2VSource/idctmmx.asm
new file mode 100644
index 000000000..7ebe4d7e3
--- /dev/null
+++ b/src/filters/source/D2VSource/idctmmx.asm
@@ -0,0 +1,738 @@
+;

+; idct8x8_xmm.asm

+;

+; Originally provided by Intel at AP-922

+; http://developer.intel.com/vtune/cbts/strmsimd/922down.htm

+; (See more app notes at http://developer.intel.com/vtune/cbts/strmsimd/appnotes.htm)

+; but in a limited edition.

+; New macro implements a column part for precise iDCT

+; The routine precision now satisfies IEEE standard 1180-1990. 

+;

+; Copyright (c) 2000-2001 Peter Gubanov <peter@elecard.net.ru>

+; Rounding trick Copyright (c) 2000 Michel Lespinasse <walken@zoy.org>

+;

+; http://www.elecard.com/peter/idct.html

+; http://www.linuxvideo.org/mpeg2dec/

+;

+;=============================================================================

+;

+; These examples contain code fragments for first stage iDCT 8x8

+; (for rows) and first stage DCT 8x8 (for columns)

+;

+;=============================================================================

+mword 	typedef qword

+mptr 	equ mword ptr

+

+BITS_INV_ACC = 5 			; 4 or 5 for IEEE

+SHIFT_INV_ROW = 16 - BITS_INV_ACC

+SHIFT_INV_COL = 1 + BITS_INV_ACC

+RND_INV_ROW = 1024 * (6 - BITS_INV_ACC) ; 1 << (SHIFT_INV_ROW-1)

+RND_INV_COL = 16 * (BITS_INV_ACC - 3) 	; 1 << (SHIFT_INV_COL-1)

+RND_INV_CORR = RND_INV_COL - 1 		; correction -1.0 and round

+

+BITS_FRW_ACC = 3 			; 2 or 3 for accuracy

+SHIFT_FRW_COL = BITS_FRW_ACC

+SHIFT_FRW_ROW = BITS_FRW_ACC + 17

+RND_FRW_ROW = 262144 * (BITS_FRW_ACC - 1)	; 1 << (SHIFT_FRW_ROW-1)

+

+_MMX = 1

+

+.nolist

+

+.586

+

+if @version GE 612

+.mmx

+;mmword	TEXTEQU	<QWORD>

+else

+include IAMMX.INC

+endif

+

+if @version GE 614

+.xmm

+;mm2word	TEXTEQU	<QWORD>			; needed for Streaming SIMD Extensions macros

+else

+include iaxmm.inc				; Streaming SIMD Extensions Emulator Macros

+endif

+

+	.list

+	.model flat

+

+_DATA SEGMENT PARA PUBLIC USE32 'DATA'

+

+one_corr 	sword            1,            1,            1,            1

+round_inv_row	dword  RND_INV_ROW,  RND_INV_ROW

+round_inv_col	sword  RND_INV_COL,  RND_INV_COL,  RND_INV_COL, RND_INV_COL

+round_inv_corr	sword RND_INV_CORR, RND_INV_CORR, RND_INV_CORR, RND_INV_CORR

+round_frw_row	dword  RND_FRW_ROW,  RND_FRW_ROW

+  tg_1_16	sword  13036,  13036,  13036,  13036	; tg * (2<<16) + 0.5

+  tg_2_16	sword  27146,  27146,  27146,  27146 	; tg * (2<<16) + 0.5

+  tg_3_16	sword -21746, -21746, -21746, -21746 	; tg * (2<<16) + 0.5

+ cos_4_16	sword -19195, -19195, -19195, -19195 	; cos * (2<<16) + 0.5

+ocos_4_16	sword  23170,  23170,  23170,  23170 	; cos * (2<<15) + 0.5

+

+ otg_3_16	sword  21895, 21895, 21895, 21895 	; tg * (2<<16) + 0.5

+

+; assume SHIFT_INV_ROW == 12

+;rounder_0	dword  65536, 65536

+;rounder_4       dword      0,     0

+;rounder_1       dword   7195,  7195

+;rounder_7	dword   1024,  1024

+;rounder_2	dword   4520,  4520

+;rounder_6	dword   1024,  1024

+;rounder_3	dword   2407,  2407

+;rounder_5	dword    240,   240

+

+; assume SHIFT_INV_ROW == 11

+rounder_0	dword  65536, 65536

+rounder_4       dword      0,     0

+rounder_1       dword   3597,  3597

+rounder_7	dword    512,   512

+rounder_2	dword   2260,  2260

+rounder_6	dword    512,   512

+rounder_3	dword   1203,  1203

+rounder_5	dword    120,   120

+

+;=============================================================================

+;

+; The first stage iDCT 8x8 - inverse DCTs of rows

+;

+;-----------------------------------------------------------------------------

+; The 8-point inverse DCT direct algorithm

+;-----------------------------------------------------------------------------

+;

+; static const short w[32] = {

+;	FIX(cos_4_16),  FIX(cos_2_16),  FIX(cos_4_16),  FIX(cos_6_16),

+;	FIX(cos_4_16),  FIX(cos_6_16), -FIX(cos_4_16), -FIX(cos_2_16),

+;	FIX(cos_4_16), -FIX(cos_6_16), -FIX(cos_4_16),  FIX(cos_2_16),

+;	FIX(cos_4_16), -FIX(cos_2_16),  FIX(cos_4_16), -FIX(cos_6_16),

+;	FIX(cos_1_16),  FIX(cos_3_16),  FIX(cos_5_16),  FIX(cos_7_16),

+;	FIX(cos_3_16), -FIX(cos_7_16), -FIX(cos_1_16), -FIX(cos_5_16),

+;	FIX(cos_5_16), -FIX(cos_1_16),  FIX(cos_7_16),  FIX(cos_3_16),

+;	FIX(cos_7_16), -FIX(cos_5_16),  FIX(cos_3_16), -FIX(cos_1_16) };

+;

+; #define DCT_8_INV_ROW(x, y)

+; {

+; 	int a0, a1, a2, a3, b0, b1, b2, b3;

+;

+; 	a0 =x[0]*w[0]+x[2]*w[1]+x[4]*w[2]+x[6]*w[3];

+; 	a1 =x[0]*w[4]+x[2]*w[5]+x[4]*w[6]+x[6]*w[7];

+; 	a2 = x[0] * w[ 8] + x[2] * w[ 9] + x[4] * w[10] + x[6] * w[11];

+; 	a3 = x[0] * w[12] + x[2] * w[13] + x[4] * w[14] + x[6] * w[15];

+; 	b0 = x[1] * w[16] + x[3] * w[17] + x[5] * w[18] + x[7] * w[19];

+; 	b1 = x[1] * w[20] + x[3] * w[21] + x[5] * w[22] + x[7] * w[23];

+; 	b2 = x[1] * w[24] + x[3] * w[25] + x[5] * w[26] + x[7] * w[27];

+; 	b3 = x[1] * w[28] + x[3] * w[29] + x[5] * w[30] + x[7] * w[31];

+;

+; 	y[0] = SHIFT_ROUND ( a0 + b0 );

+; 	y[1] = SHIFT_ROUND ( a1 + b1 );

+; 	y[2] = SHIFT_ROUND ( a2 + b2 );

+; 	y[3] = SHIFT_ROUND ( a3 + b3 );

+; 	y[4] = SHIFT_ROUND ( a3 - b3 );

+; 	y[5] = SHIFT_ROUND ( a2 - b2 );

+; 	y[6] = SHIFT_ROUND ( a1 - b1 );

+; 	y[7] = SHIFT_ROUND ( a0 - b0 );

+; }

+;

+;-----------------------------------------------------------------------------

+;

+; In this implementation the outputs of the iDCT-1D are multiplied

+; 	for rows 0,4 - by cos_4_16,

+; 	for rows 1,7 - by cos_1_16,

+; 	for rows 2,6 - by cos_2_16,

+; 	for rows 3,5 - by cos_3_16

+; and are shifted to the left for better accuracy

+;

+; For the constants used,

+; 	FIX(float_const) = (short) (float_const * (1<<15) + 0.5)

+;

+;=============================================================================

+

+;=============================================================================

+; MMX code

+;=============================================================================

+

+; Table for rows 0,4 - constants are multiplied by cos_4_16

+

+tab_i_04 	sword  16384,  16384,  16384, -16384	; movq-> w06 w04 w02 w00

+		sword  21407,   8867,   8867, -21407	; w07 w05 w03 w01

+		sword  16384, -16384,  16384,  16384	; w14 w12 w10 w08

+		sword  -8867,  21407, -21407,  -8867	; w15 w13 w11 w09

+		sword  22725,  12873,  19266, -22725	; w22 w20 w18 w16

+		sword  19266,   4520,  -4520, -12873	; w23 w21 w19 w17

+		sword  12873,   4520,   4520,  19266	; w30 w28 w26 w24

+		sword -22725,  19266, -12873, -22725	; w31 w29 w27 w25

+

+; Table for rows 1,7 - constants are multiplied by cos_1_16

+

+tab_i_17	sword  22725,  22725,  22725, -22725	; movq-> w06 w04 w02 w00

+		sword  29692,  12299,  12299, -29692	; w07 w05 w03 w01

+		sword  22725, -22725,  22725,  22725	; w14 w12 w10 w08

+		sword -12299,  29692, -29692, -12299	; w15 w13 w11 w09

+		sword  31521,  17855,  26722, -31521	; w22 w20 w18 w16

+		sword  26722,   6270,  -6270, -17855	; w23 w21 w19 w17

+		sword  17855,   6270,   6270,  26722	; w30 w28 w26 w24

+		sword -31521,  26722, -17855, -31521	; w31 w29 w27 w25

+

+; Table for rows 2,6 - constants are multiplied by cos_2_16

+

+tab_i_26	sword  21407,  21407,  21407, -21407	; movq-> w06 w04 w02 w00

+		sword  27969,  11585,  11585, -27969	; w07 w05 w03 w01

+		sword  21407, -21407,  21407,  21407	; w14 w12 w10 w08

+		sword -11585,  27969, -27969, -11585	; w15 w13 w11 w09

+		sword  29692,  16819,  25172, -29692	; w22 w20 w18 w16

+		sword  25172,   5906,  -5906, -16819	; w23 w21 w19 w17

+		sword  16819,   5906,   5906,  25172	; w30 w28 w26 w24

+		sword -29692,  25172, -16819, -29692	; w31 w29 w27 w25

+

+; Table for rows 3,5 - constants are multiplied by cos_3_16

+

+tab_i_35	sword  19266,  19266,  19266, -19266	; movq-> w06 w04 w02 w00

+		sword  25172,  10426,  10426, -25172	; w07 w05 w03 w01

+		sword  19266, -19266,  19266,  19266	; w14 w12 w10 w08

+		sword -10426,  25172, -25172, -10426	; w15 w13 w11 w09

+		sword  26722,  15137,  22654, -26722	; w22 w20 w18 w16

+		sword  22654,   5315,  -5315, -15137	; w23 w21 w19 w17

+		sword  15137,   5315,   5315,  22654	; w30 w28 w26 w24

+		sword -26722,  22654, -15137, -26722	; w31 w29 w27 w25

+

+;-----------------------------------------------------------------------------

+

+DCT_8_INV_ROW_1 MACRO INP:REQ, OUT:REQ, TABLE:REQ, ROUNDER:REQ

+

+	movq mm0, mptr [INP] 		; 0	; x3 x2 x1 x0

+

+	movq mm1, mptr [INP+8]		; 1	; x7 x6 x5 x4

+	movq mm2, mm0 			; 2	; x3 x2 x1 x0

+

+	movq mm3, mptr [TABLE]		; 3	; w06 w04 w02 w00

+	punpcklwd mm0, mm1 			; x5 x1 x4 x0

+

+	movq mm5, mm0 			; 5	; x5 x1 x4 x0

+	punpckldq mm0, mm0 			; x4 x0 x4 x0

+

+	movq mm4, mptr [TABLE+8] 	; 4	; w07 w05 w03 w01

+	punpckhwd mm2, mm1		; 1	; x7 x3 x6 x2

+

+	pmaddwd mm3, mm0 			; x4*w06+x0*w04 x4*w02+x0*w00

+	movq mm6, mm2 			; 6 	; x7 x3 x6 x2

+

+	movq mm1, mptr [TABLE+32] 	; 1 	; w22 w20 w18 w16

+	punpckldq mm2, mm2 			; x6 x2 x6 x2

+

+	pmaddwd mm4, mm2 			; x6*w07+x2*w05 x6*w03+x2*w01

+	punpckhdq mm5, mm5 			; x5 x1 x5 x1

+

+	pmaddwd mm0, mptr [TABLE+16] 		; x4*w14+x0*w12 x4*w10+x0*w08

+	punpckhdq mm6, mm6 			; x7 x3 x7 x3

+

+	movq mm7, mptr [TABLE+40] 	; 7 	; w23 w21 w19 w17

+	pmaddwd mm1, mm5 			; x5*w22+x1*w20 x5*w18+x1*w16

+

+	paddd mm3, mptr [ROUNDER] 		; +rounder

+	pmaddwd mm7, mm6 			; x7*w23+x3*w21 x7*w19+x3*w17

+

+	pmaddwd mm2, mptr [TABLE+24] 		; x6*w15+x2*w13 x6*w11+x2*w09

+	paddd mm3, mm4 			; 4 	; a1=sum(even1) a0=sum(even0)

+

+	pmaddwd mm5, mptr [TABLE+48] 		; x5*w30+x1*w28 x5*w26+x1*w24

+	movq mm4, mm3 			; 4 	; a1 a0

+

+	pmaddwd mm6, mptr [TABLE+56] 		; x7*w31+x3*w29 x7*w27+x3*w25

+	paddd mm1, mm7 			; 7 	; b1=sum(odd1) b0=sum(odd0)

+

+	paddd mm0, mptr [ROUNDER]		; +rounder

+	psubd mm3, mm1 				; a1-b1 a0-b0

+

+	psrad mm3, SHIFT_INV_ROW 		; y6=a1-b1 y7=a0-b0

+	paddd mm1, mm4 			; 4 	; a1+b1 a0+b0

+

+	paddd mm0, mm2 			; 2 	; a3=sum(even3) a2=sum(even2)

+	psrad mm1, SHIFT_INV_ROW 		; y1=a1+b1 y0=a0+b0

+

+	paddd mm5, mm6 			; 6 	; b3=sum(odd3) b2=sum(odd2)

+	movq mm4, mm0 			; 4 	; a3 a2

+

+	paddd mm0, mm5 				; a3+b3 a2+b2

+	psubd mm4, mm5 			; 5 	; a3-b3 a2-b2

+

+	psrad mm0, SHIFT_INV_ROW 		; y3=a3+b3 y2=a2+b2

+	psrad mm4, SHIFT_INV_ROW 		; y4=a3-b3 y5=a2-b2

+

+	packssdw mm1, mm0 		; 0 	; y3 y2 y1 y0

+	packssdw mm4, mm3 		; 3 	; y6 y7 y4 y5

+

+	movq mm7, mm4 			; 7 	; y6 y7 y4 y5

+	psrld mm4, 16 				; 0 y6 0 y4

+

+	pslld mm7, 16 				; y7 0 y5 0

+	movq mptr [OUT], mm1 		; 1 	; save y3 y2 y1 y0

+                             	

+	por mm7, mm4 			; 4 	; y7 y6 y5 y4

+	movq mptr [OUT+8], mm7 		; 7 	; save y7 y6 y5 y4

+ENDM

+

+;=============================================================================

+; code for Pentium III

+;=============================================================================

+

+; Table for rows 0,4 - constants are multiplied by cos_4_16

+

+tab_i_04_s 	sword 16384, 21407, 16384, 8867 ; movq-> w05 w04 w01 w00

+		sword 16384, 8867, -16384, -21407 ; w07 w06 w03 w02

+		sword 16384, -8867, 16384, -21407 ; w13 w12 w09 w08

+		sword -16384, 21407, 16384, -8867 ; w15 w14 w11 w10

+		sword 22725, 19266, 19266, -4520 ; w21 w20 w17 w16

+		sword 12873, 4520, -22725, -12873 ; w23 w22 w19 w18

+		sword 12873, -22725, 4520, -12873 ; w29 w28 w25 w24

+		sword 4520, 19266, 19266, -22725 ; w31 w30 w27 w26

+

+; Table for rows 1,7 - constants are multiplied by cos_1_16

+

+tab_i_17_s	sword 22725, 29692, 22725, 12299 ; movq-> w05 w04 w01 w00

+		sword 22725, 12299, -22725, -29692 ; w07 w06 w03 w02

+		sword 22725, -12299, 22725, -29692 ; w13 w12 w09 w08

+		sword -22725, 29692, 22725, -12299 ; w15 w14 w11 w10

+		sword 31521, 26722, 26722, -6270 ; w21 w20 w17 w16

+		sword 17855, 6270, -31521, -17855 ; w23 w22 w19 w18

+		sword 17855, -31521, 6270, -17855 ; w29 w28 w25 w24

+		sword 6270, 26722, 26722, -31521 ; w31 w30 w27 w26

+

+; Table for rows 2,6 - constants are multiplied by cos_2_16

+

+tab_i_26_s	sword 21407, 27969, 21407, 11585 ; movq-> w05 w04 w01 w00

+		sword 21407, 11585, -21407, -27969 ; w07 w06 w03 w02

+		sword 21407, -11585, 21407, -27969 ; w13 w12 w09 w08

+		sword -21407, 27969, 21407, -11585 ; w15 w14 w11 w10

+		sword 29692, 25172, 25172, -5906 ; w21 w20 w17 w16

+		sword 16819, 5906, -29692, -16819 ; w23 w22 w19 w18

+		sword 16819, -29692, 5906, -16819 ; w29 w28 w25 w24

+		sword 5906, 25172, 25172, -29692 ; w31 w30 w27 w26

+

+; Table for rows 3,5 - constants are multiplied by cos_3_16

+

+tab_i_35_s	sword 19266, 25172, 19266, 10426 ; movq-> w05 w04 w01 w00

+		sword 19266, 10426, -19266, -25172 ; w07 w06 w03 w02

+		sword 19266, -10426, 19266, -25172 ; w13 w12 w09 w08

+		sword -19266, 25172, 19266, -10426 ; w15 w14 w11 w10

+		sword 26722, 22654, 22654, -5315 ; w21 w20 w17 w16

+		sword 15137, 5315, -26722, -15137 ; w23 w22 w19 w18

+		sword 15137, -26722, 5315, -15137 ; w29 w28 w25 w24

+		sword 5315, 22654, 22654, -26722 ; w31 w30 w27 w26

+

+;-----------------------------------------------------------------------------

+

+DCT_8_INV_ROW_1_s MACRO INP:REQ, OUT:REQ, TABLE:REQ, ROUNDER:REQ

+

+	movq 	mm0, mptr [INP] 	; 0 	; x3 x2 x1 x0

+

+	movq 	mm1, mptr [INP+8]	; 1 	; x7 x6 x5 x4

+	movq 	mm2, mm0 		; 2 	; x3 x2 x1 x0

+

+	movq 	mm3, mptr [TABLE] 	; 3 	; w05 w04 w01 w00

+	pshufw	mm0, mm0, 10001000b 	; x2 x0 x2 x0

+

+	movq 	mm4, mptr [TABLE+8] 	; 4 	; w07 w06 w03 w02

+	movq 	mm5, mm1		; 5 	; x7 x6 x5 x4

+	pmaddwd mm3, mm0 		; x2*w05+x0*w04 x2*w01+x0*w00

+

+	movq 	mm6, mptr [TABLE+32] 	; 6 	; w21 w20 w17 w16

+	pshufw 	mm1, mm1, 10001000b 		; x6 x4 x6 x4

+	pmaddwd mm4, mm1 			; x6*w07+x4*w06 x6*w03+x4*w02

+

+	movq 	mm7, mptr [TABLE+40] 	; 7 	; w23 w22 w19 w18

+	pshufw 	mm2, mm2, 11011101b 		; x3 x1 x3 x1

+	pmaddwd mm6, mm2 			; x3*w21+x1*w20 x3*w17+x1*w16

+

+	pshufw 	mm5, mm5, 11011101b 		; x7 x5 x7 x5

+	pmaddwd mm7, mm5 			; x7*w23+x5*w22 x7*w19+x5*w18

+

+	paddd 	mm3, mptr [ROUNDER] 		; +rounder

+

+	pmaddwd mm0, mptr [TABLE+16] 		; x2*w13+x0*w12 x2*w09+x0*w08

+	paddd 	mm3, mm4 		; 4 	; a1=sum(even1) a0=sum(even0)

+

+	pmaddwd mm1, mptr [TABLE+24] 		; x6*w15+x4*w14 x6*w11+x4*w10

+	movq 	mm4, mm3 		; 4 	; a1 a0

+

+	pmaddwd mm2, mptr [TABLE+48] 		; x3*w29+x1*w28 x3*w25+x1*w24

+	paddd 	mm6, mm7 		; 7 	; b1=sum(odd1) b0=sum(odd0)

+

+	pmaddwd mm5, mptr [TABLE+56] 		; x7*w31+x5*w30 x7*w27+x5*w26

+	paddd mm3, mm6 				; a1+b1 a0+b0

+

+	paddd mm0, mptr [ROUNDER] 		; +rounder

+	psrad mm3, SHIFT_INV_ROW 		; y1=a1+b1 y0=a0+b0

+

+	paddd mm0, mm1 			; 1 	; a3=sum(even3) a2=sum(even2)

+	psubd mm4, mm6 			; 6 	; a1-b1 a0-b0

+

+	movq mm7, mm0 			; 7 	; a3 a2

+	paddd mm2, mm5 			; 5 	; b3=sum(odd3) b2=sum(odd2)

+

+	paddd mm0, mm2 				; a3+b3 a2+b2

+	psrad mm4, SHIFT_INV_ROW 		; y6=a1-b1 y7=a0-b0

+

+	psubd mm7, mm2 			; 2 	; a3-b3 a2-b2

+	psrad mm0, SHIFT_INV_ROW 		; y3=a3+b3 y2=a2+b2

+

+	psrad mm7, SHIFT_INV_ROW 		; y4=a3-b3 y5=a2-b2

+

+	packssdw mm3, mm0 		; 0 	; y3 y2 y1 y0

+

+	packssdw mm7, mm4 		; 4 	; y6 y7 y4 y5

+

+	movq mptr [OUT], mm3 		; 3 	; save y3 y2 y1 y0

+	pshufw mm7, mm7, 10110001b 		; y7 y6 y5 y4

+

+	movq mptr [OUT+8], mm7 		; 7 	; save y7 y6 y5 y4

+ENDM

+

+;=============================================================================

+;

+;=============================================================================

+

+;=============================================================================

+;

+; The first stage DCT 8x8 - forward DCTs of columns

+;

+; The outputs are multiplied

+; for rows 0,4 - on cos_4_16,

+; for rows 1,7 - on cos_1_16,

+; for rows 2,6 - on cos_2_16,

+; for rows 3,5 - on cos_3_16

+; and are shifted to the left for rise of accuracy

+;

+;-----------------------------------------------------------------------------

+;

+; The 8-point scaled forward DCT algorithm (26a8m)

+;

+;-----------------------------------------------------------------------------

+;

+; #define DCT_8_FRW_COL(x, y)

+;{

+; short t0, t1, t2, t3, t4, t5, t6, t7;

+; short tp03, tm03, tp12, tm12, tp65, tm65;

+; short tp465, tm465, tp765, tm765;

+;

+; t0 = LEFT_SHIFT ( x[0] + x[7] );

+; t1 = LEFT_SHIFT ( x[1] + x[6] );

+; t2 = LEFT_SHIFT ( x[2] + x[5] );

+; t3 = LEFT_SHIFT ( x[3] + x[4] );

+; t4 = LEFT_SHIFT ( x[3] - x[4] );

+; t5 = LEFT_SHIFT ( x[2] - x[5] );

+; t6 = LEFT_SHIFT ( x[1] - x[6] );

+; t7 = LEFT_SHIFT ( x[0] - x[7] );

+;

+; tp03 = t0 + t3;

+; tm03 = t0 - t3;

+; tp12 = t1 + t2;

+; tm12 = t1 - t2;

+;

+; y[0] = tp03 + tp12;

+; y[4] = tp03 - tp12;

+;

+; y[2] = tm03 + tm12 * tg_2_16;

+; y[6] = tm03 * tg_2_16 - tm12;

+;

+; tp65 =(t6 +t5 )*cos_4_16;

+; tm65 =(t6 -t5 )*cos_4_16;

+;

+; tp765 = t7 + tp65;

+; tm765 = t7 - tp65;

+; tp465 = t4 + tm65;

+; tm465 = t4 - tm65;

+;

+; y[1] = tp765 + tp465 * tg_1_16;

+; y[7] = tp765 * tg_1_16 - tp465;

+; y[5] = tm765 * tg_3_16 + tm465;

+; y[3] = tm765 - tm465 * tg_3_16;

+;}

+;

+;=============================================================================

+DCT_8_FRW_COL_4 MACRO INP:REQ, OUT:REQ

+LOCAL x0, x1, x2, x3, x4, x5, x6, x7

+LOCAL y0, y1, y2, y3, y4, y5, y6, y7

+x0 equ [INP + 0*16]

+x1 equ [INP + 1*16]

+x2 equ [INP + 2*16]

+x3 equ [INP + 3*16]

+x4 equ [INP + 4*16]

+x5 equ [INP + 5*16]

+x6 equ [INP + 6*16]

+x7 equ [INP + 7*16]

+y0 equ [OUT + 0*16]

+y1 equ [OUT + 1*16]

+y2 equ [OUT + 2*16]

+y3 equ [OUT + 3*16]

+y4 equ [OUT + 4*16]

+y5 equ [OUT + 5*16]

+y6 equ [OUT + 6*16]

+y7 equ [OUT + 7*16]

+movq mm0, x1 ; 0 ; x1

+movq mm1, x6 ; 1 ; x6

+movq mm2, mm0 ; 2 ; x1

+movq mm3, x2 ; 3 ; x2

+paddsw mm0, mm1 ; t1 = x[1] + x[6]

+movq mm4, x5 ; 4 ; x5

+psllw mm0, SHIFT_FRW_COL ; t1

+movq mm5, x0 ; 5 ; x0

+paddsw mm4, mm3 ; t2 = x[2] + x[5]

+paddsw mm5, x7 ; t0 = x[0] + x[7]

+psllw mm4, SHIFT_FRW_COL ; t2

+movq mm6, mm0 ; 6 ; t1

+psubsw mm2, mm1 ; 1 ; t6 = x[1] - x[6]

+movq mm1, mptr tg_2_16 ; 1 ; tg_2_16

+psubsw mm0, mm4 ; tm12 = t1 - t2

+movq mm7, x3 ; 7 ; x3

+pmulhw mm1, mm0 ; tm12*tg_2_16

+paddsw mm7, x4 ; t3 = x[3] + x[4]

+psllw mm5, SHIFT_FRW_COL ; t0

+paddsw mm6, mm4 ; 4 ; tp12 = t1 + t2

+psllw mm7, SHIFT_FRW_COL ; t3

+movq mm4, mm5 ; 4 ; t0

+psubsw mm5, mm7 ; tm03 = t0 - t3

+paddsw mm1, mm5 ; y2 = tm03 + tm12*tg_2_16

+paddsw mm4, mm7 ; 7 ; tp03 = t0 + t3

+por mm1, mptr one_corr ; correction y2 +0.5

+psllw mm2, SHIFT_FRW_COL+1 ; t6

+pmulhw mm5, mptr tg_2_16 ; tm03*tg_2_16

+movq mm7, mm4 ; 7 ; tp03

+psubsw mm3, x5 ; t5 = x[2] - x[5]

+psubsw mm4, mm6 ; y4 = tp03 - tp12

+movq y2, mm1 ; 1 ; save y2

+paddsw mm7, mm6 ; 6 ; y0 = tp03 + tp12

+movq mm1, x3 ; 1 ; x3

+psllw mm3, SHIFT_FRW_COL+1 ; t5

+psubsw mm1, x4 ; t4 = x[3] - x[4]

+movq mm6, mm2 ; 6 ; t6

+movq y4, mm4 ; 4 ; save y4

+paddsw mm2, mm3 ; t6 + t5

+pmulhw mm2, mptr ocos_4_16 ; tp65 = (t6 + t5)*cos_4_16

+psubsw mm6, mm3 ; 3 ; t6 - t5

+pmulhw mm6, mptr ocos_4_16 ; tm65 = (t6 - t5)*cos_4_16

+psubsw mm5, mm0 ; 0 ; y6 = tm03*tg_2_16 - tm12

+por mm5, mptr one_corr ; correction y6 +0.5

+psllw mm1, SHIFT_FRW_COL ; t4

+por mm2, mptr one_corr ; correction tp65 +0.5

+movq mm4, mm1 ; 4 ; t4

+movq mm3, x0 ; 3 ; x0

+paddsw mm1, mm6 ; tp465 = t4 + tm65

+psubsw mm3, x7 ; t7 = x[0] - x[7]

+psubsw mm4, mm6 ; 6 ; tm465 = t4 - tm65

+movq mm0, mptr tg_1_16 ; 0 ; tg_1_16

+psllw mm3, SHIFT_FRW_COL ; t7

+movq mm6, mptr tg_3_16 ; 6 ; tg_3_16

+pmulhw mm0, mm1 ; tp465*tg_1_16

+movq y0, mm7 ; 7 ; save y0

+pmulhw mm6, mm4 ; tm465*tg_3_16

+movq y6, mm5 ; 5 ; save y6

+movq mm7, mm3 ; 7 ; t7

+movq mm5, mptr tg_3_16 ; 5 ; tg_3_16

+psubsw mm7, mm2 ; tm765 = t7 - tp65

+paddsw mm3, mm2 ; 2 ; tp765 = t7 + tp65

+pmulhw mm5, mm7 ; tm765*tg_3_16

+paddsw mm0, mm3 ; y1 = tp765 + tp465*tg_1_16

+paddsw mm6, mm4 ; tm465*tg_3_16

+pmulhw mm3, mptr tg_1_16 ; tp765*tg_1_16

+por mm0, mptr one_corr ; correction y1 +0.5

+paddsw mm5, mm7 ; tm765*tg_3_16

+psubsw mm7, mm6 ; 6 ; y3 = tm765 - tm465*tg_3_16

+movq y1, mm0 ; 0 ; save y1

+paddsw mm5, mm4 ; 4 ; y5 = tm765*tg_3_16 + tm465

+movq y3, mm7 ; 7 ; save y3

+psubsw mm3, mm1 ; 1 ; y7 = tp765*tg_1_16 - tp465

+movq y5, mm5 ; 5 ; save y5

+movq y7, mm3 ; 3 ; save y7

+ENDM

+

+DCT_8_INV_COL_4 MACRO INP:REQ, OUT:REQ

+	movq	mm0, qword ptr tg_3_16

+

+	movq	mm3, qword ptr [INP+16*3]

+	movq	mm1, mm0			; tg_3_16

+

+	movq	mm5, qword ptr [INP+16*5]

+	pmulhw	mm0, mm3			; x3*(tg_3_16-1)

+

+	movq	mm4, qword ptr tg_1_16

+	pmulhw	mm1, mm5			; x5*(tg_3_16-1)

+

+	movq	mm7, qword ptr [INP+16*7]

+	movq	mm2, mm4			; tg_1_16

+

+	movq	mm6, qword ptr [INP+16*1]

+	pmulhw	mm4, mm7			; x7*tg_1_16

+

+	paddsw	mm0, mm3			; x3*tg_3_16

+	pmulhw	mm2, mm6			; x1*tg_1_16

+

+	paddsw	mm1, mm3			; x3+x5*(tg_3_16-1)

+	psubsw	mm0, mm5			; x3*tg_3_16-x5 = tm35

+

+	movq	mm3, qword ptr ocos_4_16

+	paddsw	mm1, mm5			; x3+x5*tg_3_16 = tp35

+

+	paddsw	mm4, mm6			; x1+tg_1_16*x7 = tp17

+	psubsw	mm2, mm7			; x1*tg_1_16-x7 = tm17

+

+	movq	mm5, mm4			; tp17

+	movq	mm6, mm2			; tm17

+

+	paddsw	mm5, mm1			; tp17+tp35 = b0

+	psubsw	mm6, mm0			; tm17-tm35 = b3

+

+	psubsw	mm4, mm1			; tp17-tp35 = t1

+	paddsw	mm2, mm0			; tm17+tm35 = t2

+

+	movq	mm7, qword ptr tg_2_16

+	movq	mm1, mm4			; t1

+

+;	movq	qword ptr [SCRATCH+0], mm5	; save b0

+	movq	qword ptr [OUT+3*16], mm5	; save b0

+	paddsw	mm1, mm2			; t1+t2

+

+;	movq	qword ptr [SCRATCH+8], mm6	; save b3

+	movq	qword ptr [OUT+5*16], mm6	; save b3

+	psubsw	mm4, mm2			; t1-t2

+

+	movq	mm5, qword ptr [INP+2*16]

+	movq	mm0, mm7			; tg_2_16

+

+	movq	mm6, qword ptr [INP+6*16]

+	pmulhw	mm0, mm5			; x2*tg_2_16

+

+	pmulhw	mm7, mm6			; x6*tg_2_16

+; slot

+	pmulhw	mm1, mm3			; ocos_4_16*(t1+t2) = b1/2

+; slot

+	movq	mm2, qword ptr [INP+0*16]

+	pmulhw	mm4, mm3			; ocos_4_16*(t1-t2) = b2/2

+

+	psubsw	mm0, mm6			; t2*tg_2_16-x6 = tm26

+	movq	mm3, mm2			; x0

+

+	movq	mm6, qword ptr [INP+4*16]

+	paddsw	mm7, mm5			; x2+x6*tg_2_16 = tp26

+

+	paddsw	mm2, mm6			; x0+x4 = tp04

+	psubsw	mm3, mm6			; x0-x4 = tm04

+

+	movq	mm5, mm2			; tp04

+	movq	mm6, mm3			; tm04

+

+	psubsw	mm2, mm7			; tp04-tp26 = a3

+	paddsw	mm3, mm0			; tm04+tm26 = a1

+

+	paddsw mm1, mm1				; b1

+	paddsw mm4, mm4				; b2

+

+	paddsw	mm5, mm7			; tp04+tp26 = a0

+	psubsw	mm6, mm0			; tm04-tm26 = a2

+

+	movq	mm7, mm3			; a1

+	movq	mm0, mm6			; a2

+

+	paddsw	mm3, mm1			; a1+b1

+	paddsw	mm6, mm4			; a2+b2

+

+	psraw	mm3, SHIFT_INV_COL		; dst1

+	psubsw	mm7, mm1			; a1-b1

+

+	psraw	mm6, SHIFT_INV_COL		; dst2

+	psubsw	mm0, mm4			; a2-b2

+

+;	movq	mm1, qword ptr [SCRATCH+0]	; load b0

+	movq	mm1, qword ptr [OUT+3*16]	; load b0

+	psraw	mm7, SHIFT_INV_COL		; dst6

+

+	movq	mm4, mm5			; a0

+	psraw	mm0, SHIFT_INV_COL		; dst5

+

+	movq	qword ptr [OUT+1*16], mm3

+	paddsw	mm5, mm1			; a0+b0

+

+	movq	qword ptr [OUT+2*16], mm6

+	psubsw	mm4, mm1			; a0-b0

+

+;	movq	mm3, qword ptr [SCRATCH+8]	; load b3

+	movq	mm3, qword ptr [OUT+5*16]	; load b3

+	psraw	mm5, SHIFT_INV_COL		; dst0

+

+	movq	mm6, mm2			; a3

+	psraw	mm4, SHIFT_INV_COL		; dst7

+

+	movq	qword ptr [OUT+5*16], mm0

+	paddsw	mm2, mm3			; a3+b3

+

+	movq	qword ptr [OUT+6*16], mm7

+	psubsw	mm6, mm3			; a3-b3

+

+	movq	qword ptr [OUT+0*16], mm5

+	psraw	mm2, SHIFT_INV_COL		; dst3

+

+	movq	qword ptr [OUT+7*16], mm4

+	psraw	mm6, SHIFT_INV_COL		; dst4

+

+	movq	qword ptr [OUT+3*16], mm2

+

+	movq	qword ptr [OUT+4*16], mm6

+ENDM

+

+_TEXT SEGMENT PARA PUBLIC USE32 'CODE'

+

+;

+; extern "C" __fastcall void idct8x8_mmx (short *src_result);

+;

+public  @MMX_IDCT@4

+

+@MMX_IDCT@4 proc near

+	mov     eax, ecx          ; source

+

+	DCT_8_INV_ROW_1	[eax+0], [eax+0], tab_i_04, rounder_0

+	DCT_8_INV_ROW_1	[eax+16], [eax+16], tab_i_17, rounder_1

+	DCT_8_INV_ROW_1	[eax+32], [eax+32], tab_i_26, rounder_2

+	DCT_8_INV_ROW_1	[eax+48], [eax+48], tab_i_35, rounder_3

+	DCT_8_INV_ROW_1	[eax+64], [eax+64], tab_i_04, rounder_4

+	DCT_8_INV_ROW_1	[eax+80], [eax+80], tab_i_35, rounder_5

+	DCT_8_INV_ROW_1	[eax+96], [eax+96], tab_i_26, rounder_6

+	DCT_8_INV_ROW_1	[eax+112], [eax+112], tab_i_17, rounder_7

+

+	DCT_8_INV_COL_4 [eax+0],[eax+0]

+	DCT_8_INV_COL_4 [eax+8],[eax+8]

+

+	ret    

+

+@MMX_IDCT@4 ENDP

+

+_TEXT ENDS

+

+_TEXT SEGMENT PARA PUBLIC USE32 'CODE'

+

+;

+; extern "C" __fastcall void idct8x8_sse (short *src_result);

+;

+public  @SSEMMX_IDCT@4

+

+@SSEMMX_IDCT@4 proc near

+	mov     eax, ecx          ; source

+

+	DCT_8_INV_ROW_1_s [eax+0], [eax+0], tab_i_04_s, rounder_0

+	DCT_8_INV_ROW_1_s [eax+16], [eax+16], tab_i_17_s, rounder_1

+	DCT_8_INV_ROW_1_s [eax+32], [eax+32], tab_i_26_s, rounder_2

+	DCT_8_INV_ROW_1_s [eax+48], [eax+48], tab_i_35_s, rounder_3

+	DCT_8_INV_ROW_1_s [eax+64], [eax+64], tab_i_04_s, rounder_4

+	DCT_8_INV_ROW_1_s [eax+80], [eax+80], tab_i_35_s, rounder_5

+	DCT_8_INV_ROW_1_s [eax+96], [eax+96], tab_i_26_s, rounder_6

+	DCT_8_INV_ROW_1_s [eax+112], [eax+112], tab_i_17_s, rounder_7

+

+	DCT_8_INV_COL_4 [eax+0],[eax+0]

+	DCT_8_INV_COL_4 [eax+8],[eax+8]

+

+	ret

+

+@SSEMMX_IDCT@4 ENDP

+

+_TEXT ENDS

+

+END
\ No newline at end of file
diff --git a/src/filters/source/D2VSource/idctref.cpp b/src/filters/source/D2VSource/idctref.cpp
new file mode 100644
index 000000000..74b695ce1
--- /dev/null
+++ b/src/filters/source/D2VSource/idctref.cpp
@@ -0,0 +1,362 @@
+#include "stdafx.h"

+

+/* idctref_miha.c, Inverse Discrete Fourier Transform, double precision */

+

+/*************************************************************/

+/*                                                           */

+/* x87 hand-optimized assembly by Miha Peternel              */

+/*                                     27.11. - 20.1.2001    */

+/*                                                           */

+/* You are free to use this code in your project if:         */

+/* - no changes are made to this message                     */

+/* - any changes to this code are publicly available         */

+/* - your project documentation contains the following text: */

+/*   "This software contains fast high-quality IDCT decoder  */

+/*    by Miha Peternel."                                     */

+/*                                                           */

+/*************************************************************/

+

+/*  Perform IEEE 1180 reference (64-bit floating point, separable 8x1

+ *  direct matrix multiply) Inverse Discrete Cosine Transform

+*/

+

+#define ModelX 123 // enable C-level optimizations by Miha Peternel

+

+/* Here we use math.h to generate constants.  Compiler results may

+   vary a little */

+

+#include <math.h>

+

+#define M_PI	3.1415926535897932384626433832795

+const static double HALF = 0.5;

+

+/* private data */

+static short iclip[1024+1024]; /* clipping table */

+static short *iclp;

+

+/* cosine transform matrix for 8x1 IDCT */

+static double c[8][8];

+

+/* initialize DCT coefficient matrix */

+void Initialize_REF_IDCT()

+{

+  int freq, time, i;

+  double scale;

+

+  for (freq=0; freq < 8; freq++)

+  {

+    scale = (freq == 0) ? sqrt(0.125) : 0.5;

+    for (time=0; time<8; time++)

+      c[freq][time] = scale*cos((M_PI/8.0)*freq*(time + 0.5));

+  }

+

+#ifdef ModelX

+  iclp = iclip+1024;

+  for (i= -1024; i<1024; i++)

+    iclp[i] = (i<-256) ? -256 : ((i>255) ? 255 : i);

+#endif

+}

+

+void REF_IDCT(short *block)

+{

+  double tmp[64];

+	double rnd[64];

+	int int0, int1, int2, int3, int4, int5, int6, int7;

+	unsigned short fpold;

+	unsigned short fpnew;

+

+	int *b = (int *) block;

+

+  if( !(b[0]|(b[31]&~0x10000)) )

+	{

+	  if( b[ 1]|b[ 2]|b[ 3]|b[ 4]|b[ 5]|b[ 6] )

+		  goto normal;

+	  if( b[ 7]|b[ 8]|b[ 9]|b[10]|b[11]|b[12] )

+		  goto normal;

+	  if( b[13]|b[14]|b[15]|b[16]|b[17]|b[18] )

+		  goto normal;

+	  if( b[19]|b[20]|b[21]|b[22]|b[23]|b[24] )

+		  goto normal;

+	  if( b[25]|b[26]|b[27]|b[28]|b[29]|b[30] )

+		  goto normal;

+		b[31]=0;

+		return;

+	}

+normal:

+

+	__asm

+	{

+		// do the IDCT

+		mov esi,[block]

+		lea eax,[c]

+		lea edi,[tmp]

+		//mov ebx,8

+		mov ebx,8 // 0x77000000 // 8

+		align 16

+	__col1:

+	    movzx edx,[esi+1*2]

+			mov   ecx,[esi+2*2]

+			or    edx,[esi+4*2]

+			or    ecx,[esi+6*2]

+			or edx,ecx

+			//mov ecx,8

+			mov ecx,8/2 // 0x77000000 // 8

+

+			jnz __row1

+				fild  word ptr [esi+0*2]

+				fmul qword ptr [eax+0*8*8]

+				fst  qword ptr [edi+0*8]

+				fst  qword ptr [edi+1*8]

+				fst  qword ptr [edi+2*8]

+				fst  qword ptr [edi+3*8]

+				fst  qword ptr [edi+4*8]

+				fst  qword ptr [edi+5*8]

+				fst  qword ptr [edi+6*8]

+				fstp qword ptr [edi+7*8]

+				add edi,8*8

+				jmp __next1

+			align 16

+		__row1:

+				fild  word ptr [esi+0*2]

+				fmul qword ptr [eax+0*8*8]

+				fild  word ptr [esi+1*2]

+				fmul qword ptr [eax+1*8*8]

+				fadd

+				fild  word ptr [esi+2*2]

+				fmul qword ptr [eax+2*8*8]

+				fadd

+				fild  word ptr [esi+3*2]

+				fmul qword ptr [eax+3*8*8]

+				fadd

+				fild  word ptr [esi+4*2]

+				fmul qword ptr [eax+4*8*8]

+				fadd

+				fild  word ptr [esi+5*2]

+				fmul qword ptr [eax+5*8*8]

+				fadd

+				fild  word ptr [esi+6*2]

+				fmul qword ptr [eax+6*8*8]

+				fadd

+				fild  word ptr [esi+7*2]

+				fmul qword ptr [eax+7*8*8]

+				fadd

+

+				fild  word ptr [esi+0*2]

+				fmul qword ptr [eax+0*8*8+8]

+				fild  word ptr [esi+1*2]

+				fmul qword ptr [eax+1*8*8+8]

+				fadd

+				fild  word ptr [esi+2*2]

+				fmul qword ptr [eax+2*8*8+8]

+				fadd

+				fild  word ptr [esi+3*2]

+				fmul qword ptr [eax+3*8*8+8]

+				fadd

+				fild  word ptr [esi+4*2]

+				fmul qword ptr [eax+4*8*8+8]

+				fadd

+				fild  word ptr [esi+5*2]

+				fmul qword ptr [eax+5*8*8+8]

+				fadd

+				fild  word ptr [esi+6*2]

+				fmul qword ptr [eax+6*8*8+8]

+				fadd

+				fild  word ptr [esi+7*2]

+				fmul qword ptr [eax+7*8*8+8]

+				fadd

+				add eax,8*2

+				fxch st(1)

+				fstp qword ptr [edi]//

+				fstp qword ptr [edi+8]

+				add edi,8*2

+			dec ecx

+

+			jnz __row1

+			add eax,-8*8

+			  //align 16

+		__next1:

+			add esi,+8*2

+

+		sub ebx,0x80000001 // add ebx,ebx 

+		js  __col1

+			//align 16

+			test ebx,ebx // align jump &| redo flags

+		jnz __col1

+

+		lea esi,[tmp]

+		lea eax,[c]

+		lea edi,[rnd]

+		//mov edi,[block]

+    fld qword ptr [HALF]

+		mov ebx,8

+	__row2:

+			mov ecx,8/2

+			align 16

+			__col2:

+				fld  qword ptr [esi+0*8*8]

+				fmul qword ptr [eax+0*8*8]

+				fld  qword ptr [esi+1*8*8]

+				fmul qword ptr [eax+1*8*8]

+				fadd

+				fld  qword ptr [esi+2*8*8]

+				fmul qword ptr [eax+2*8*8]

+				fadd

+				fld  qword ptr [esi+3*8*8]

+				fmul qword ptr [eax+3*8*8]

+				fadd

+				fld  qword ptr [esi+4*8*8]

+				fmul qword ptr [eax+4*8*8]

+				fadd

+				fld  qword ptr [esi+5*8*8]

+				fmul qword ptr [eax+5*8*8]

+				fadd

+				fld  qword ptr [esi+6*8*8]

+				fmul qword ptr [eax+6*8*8]

+				fadd

+				fld  qword ptr [esi+7*8*8]

+				fmul qword ptr [eax+7*8*8]

+				fadd

+				fadd st(0),st(1)

+

+				fxch st(1)

+

+				fld  qword ptr [esi+0*8*8]

+				fmul qword ptr [eax+0*8*8+8]

+				fld  qword ptr [esi+1*8*8]

+				fmul qword ptr [eax+1*8*8+8]

+				fadd

+				fld  qword ptr [esi+2*8*8]

+				fmul qword ptr [eax+2*8*8+8]

+				fadd

+				fld  qword ptr [esi+3*8*8]

+				fmul qword ptr [eax+3*8*8+8]

+				fadd

+				fld  qword ptr [esi+4*8*8]

+				fmul qword ptr [eax+4*8*8+8]

+				fadd

+				fld  qword ptr [esi+5*8*8]

+				fmul qword ptr [eax+5*8*8+8]

+				fadd

+				fld  qword ptr [esi+6*8*8]

+				fmul qword ptr [eax+6*8*8+8]

+				fadd

+				fld  qword ptr [esi+7*8*8]

+				fmul qword ptr [eax+7*8*8+8]

+				fadd

+				fadd st(0),st(1)

+				add eax,8*2

+

+				fxch st(2)

+				fstp qword ptr [edi]

+				fxch st(1)

+				fstp qword ptr [edi+8*8]

+				add edi,8*8*2

+

+			dec ecx

+

+			jnz __col2

+			add eax,-8*8

+			add esi,+8

+			add edi,8-8*8*8

+

+		sub ebx,0x80000001

+		js  __row2

+			  //align 16

+				test ebx,ebx // align jump &| redo flags

+		jnz __row2

+		ffree st(0) // bye bye 0.5

+

+	  // set x87 to floor mode

+		fstcw [fpold]

+		movzx eax, [fpold]

+

+		or eax, 0x0400 // round down - floor

+		mov [fpnew], ax

+		fldcw [fpnew]

+

+		// now floor the damn array

+		lea esi, [rnd]

+		mov edi, [block]

+		mov ebx, -256 // clip min

+		mov edx, +255 // clip max

+		mov ecx, 8

+		align 16

+	__floor:

+		  fld   qword ptr [esi+0*8]

+			fistp dword ptr [int0]

+			  mov eax,[int0]

+				cmp   eax,ebx

+				cmovl eax,ebx

+				cmp   eax,edx

+				cmovg eax,edx

+		  fld   qword ptr [esi+1*8]

+			fistp dword ptr [int1]

+				mov word ptr [edi+0*2],ax

+			  mov eax,[int1]

+				cmp   eax,ebx

+				cmovl eax,ebx

+				cmp   eax,edx

+				cmovg eax,edx

+		  fld   qword ptr [esi+2*8]

+			fistp dword ptr [int2]

+				mov word ptr [edi+1*2],ax

+			  mov eax,[int2]

+				cmp   eax,ebx

+				cmovl eax,ebx

+				cmp   eax,edx

+				cmovg eax,edx

+		  fld   qword ptr [esi+3*8]

+			fistp dword ptr [int3]

+				mov word ptr [edi+2*2],ax

+			  mov eax,[int3]

+				cmp   eax,ebx

+				cmovl eax,ebx

+				cmp   eax,edx

+				cmovg eax,edx

+		  fld   qword ptr [esi+4*8]

+			fistp dword ptr [int4]

+				mov word ptr [edi+3*2],ax

+			  mov eax,[int4]

+				cmp   eax,ebx

+				cmovl eax,ebx

+				cmp   eax,edx

+				cmovg eax,edx

+		  fld   qword ptr [esi+5*8]

+			fistp dword ptr [int5]

+				mov word ptr [edi+4*2],ax

+			  mov eax,[int5]

+				cmp   eax,ebx

+				cmovl eax,ebx

+				cmp   eax,edx

+				cmovg eax,edx

+		  fld   qword ptr [esi+6*8]

+			fistp dword ptr [int6]

+				mov word ptr [edi+5*2],ax

+			  mov eax,[int6]

+				cmp   eax,ebx

+				cmovl eax,ebx

+				cmp   eax,edx

+				cmovg eax,edx

+		  fld   qword ptr [esi+7*8]

+			fistp dword ptr [int7]

+				mov word ptr [edi+6*2],ax

+			  mov eax,[int7]

+				cmp   eax,ebx

+				cmovl eax,ebx

+				cmp   eax,edx

+				cmovg eax,edx

+				mov word ptr [edi+7*2],ax

+

+			add esi, 8*8

+			add edi, 8*2

+

+		sub ecx,0x80000001

+		js  __floor

+			  //align 16

+				test ecx,ecx // align jump &| redo flags

+		jnz __floor

+

+		// set x87 to default mode

+		fldcw [fpold]

+	};

+}

diff --git a/src/filters/source/D2VSource/resource.h b/src/filters/source/D2VSource/resource.h
new file mode 100644
index 000000000..1d1658441
--- /dev/null
+++ b/src/filters/source/D2VSource/resource.h
@@ -0,0 +1,14 @@
+//{{NO_DEPENDENCIES}}

+// Microsoft Visual C++ generated include file.

+// Used by d2vsource.rc

+

+// Next default values for new objects

+// 

+#ifdef APSTUDIO_INVOKED

+#ifndef APSTUDIO_READONLY_SYMBOLS

+#define _APS_NEXT_RESOURCE_VALUE        101

+#define _APS_NEXT_COMMAND_VALUE         40001

+#define _APS_NEXT_CONTROL_VALUE         1001

+#define _APS_NEXT_SYMED_VALUE           101

+#endif

+#endif

diff --git a/src/filters/source/D2VSource/stdafx.cpp b/src/filters/source/D2VSource/stdafx.cpp
new file mode 100644
index 000000000..0d549c051
--- /dev/null
+++ b/src/filters/source/D2VSource/stdafx.cpp
@@ -0,0 +1,8 @@
+// stdafx.cpp : source file that includes just the standard includes

+// d2vsource.pch will be the pre-compiled header

+// stdafx.obj will contain the pre-compiled type information

+

+#include "stdafx.h"

+

+// TODO: reference any additional headers you need in STDAFX.H

+// and not in this file

diff --git a/src/filters/source/D2VSource/stdafx.h b/src/filters/source/D2VSource/stdafx.h
new file mode 100644
index 000000000..e596301ef
--- /dev/null
+++ b/src/filters/source/D2VSource/stdafx.h
@@ -0,0 +1,22 @@
+// stdafx.h : include file for standard system include files,

+// or project specific include files that are used frequently, but

+// are changed infrequently

+//

+

+#pragma once

+#include "../../../DSUtil/SharedInclude.h"

+

+#define WIN32_LEAN_AND_MEAN		// Exclude rarely-used stuff from Windows headers

+#define _ATL_CSTRING_EXPLICIT_CONSTRUCTORS	// some CString constructors will be explicit

+

+#ifndef VC_EXTRALEAN

+#define VC_EXTRALEAN		// Exclude rarely-used stuff from Windows headers

+#endif

+

+#include <afx.h>

+#include <afxwin.h>         // MFC core and standard components

+

+// TODO: reference additional headers your program requires here

+

+#include <streams.h>

+#include <dvdmedia.h>