Multi-View and Stereo 3D

Official Documentation:
http://www.blender.org/manual/render/workflows/multiview.html

Implemented Features
====================
Builtin Stereo Camera
* Convergence Mode
* Interocular Distance
* Convergence Distance
* Pivot Mode

Viewport
* Cameras
* Plane
* Volume

Compositor
* View Switch Node
* Image Node Multi-View OpenEXR support

Sequencer
* Image/Movie Strips 'Use Multiview'

UV/Image Editor
* Option to see Multi-View images in Stereo-3D or its individual images
* Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images

I/O
* Save/Open Multi-View (OpenEXR, Stereo3D, individual views) images

Scene Render Views
* Ability to have an arbitrary number of views in the scene

Missing Bits
============
First rule of Multi-View bug report: If something is not working as it should *when Views is off* this is a severe bug, do mention this in the report.

Second rule is, if something works *when Views is off* but doesn't (or crashes) when *Views is on*, this is a important bug. Do mention this in the report.

Everything else is likely small todos, and may wait until we are sure none of the above is happening.

Apart from that there are those known issues:
* Compositor Image Node poorly working for Multi-View OpenEXR
(this was working prefectly before the 'Use Multi-View' functionality)
* Selecting camera from Multi-View when looking from camera is problematic
* Animation Playback (ctrl+F11) doesn't support stereo formats
* Wrong filepath when trying to play back animated scene
* Viewport Rendering doesn't support Multi-View
* Overscan Rendering
* Fullscreen display modes need to warn the user
* Object copy should be aware of views suffix

Acknowledgments
===============
* Francesco Siddi for the help with the original feature specs and design
* Brecht Van Lommel for the original review of the code and design early on
* Blender Foundation for the Development Fund to support the project wrap up

Final patch reviewers:
* Antony Riakiotakis (psy-fi)
* Campbell Barton (ideasman42)
* Julian Eisel (Severin)
* Sergey Sharybin (nazgul)
* Thomas Dinged (dingto)

Code contributors of the original branch in github:
* Alexey Akishin
* Gabriel Caraballo

1 files changed, 1292 insertions, 0 deletions

diff --git a/source/blender/imbuf/intern/stereoimbuf.c b/source/blender/imbuf/intern/stereoimbuf.c
new file mode 100644
index 00000000000..cf9c6b388ff
--- /dev/null
+++ b/source/blender/imbuf/intern/stereoimbuf.c
@@ -0,0 +1,1292 @@
+/*
+ * ***** BEGIN GPL LICENSE BLOCK *****
+ *
+ * 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
+ * of the License, 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 this program; if not, write to the Free Software Foundation,
+ * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ *
+ * The Original Code is Copyright (C) 2015 by Blender Foundation
+ * All rights reserved.
+ *
+ * The Original Code is: all of this file.
+ *
+ * Contributor(s): Dalai Felinto
+ *
+ * ***** END GPL LICENSE BLOCK *****
+ */
+
+/** \file blender/imbuf/intern/stereoimbuf.c
+ *  \ingroup imbuf
+ */
+
+#include <stddef.h>
+
+#include "IMB_imbuf.h"
+#include "IMB_imbuf_types.h"
+
+#include "IMB_allocimbuf.h"
+#include "IMB_filetype.h"
+#include "IMB_metadata.h"
+#include "IMB_colormanagement_intern.h"
+
+#include "imbuf.h"
+
+#include "MEM_guardedalloc.h"
+
+#include "BLI_threads.h"
+#include "BLI_utildefines.h"
+
+#include "BLI_math.h"
+
+#include "DNA_userdef_types.h"
+#include "DNA_scene_types.h"
+
+/* prototypes */
+struct Stereo3DData;
+static void imb_stereo3d_write_doit(struct Stereo3DData *s3d_data, struct Stereo3dFormat *s3d);
+static void imb_stereo3d_read_doit(struct Stereo3DData *s3d_data, struct Stereo3dFormat *s3d);
+
+typedef struct Stereo3DData {
+	struct { float *left, *right, *stereo; } rectf;
+	struct { uchar *left, *right, *stereo; } rect;
+	size_t x, y, channels;
+	bool is_float;
+} Stereo3DData;
+
+static void imb_stereo3d_write_anaglyph(Stereo3DData *s3d, enum eStereo3dAnaglyphType mode)
+{
+	int x, y;
+	size_t width = s3d->x;
+	size_t height = s3d->y;
+	const size_t channels = s3d->channels;
+
+	const int stride_from = width;
+	const int stride_to = width;
+
+	int anaglyph_encoding[3][3] = {
+	    {0, 1, 1},
+	    {1, 0, 1},
+	    {0, 0, 1},
+	};
+
+	int r, g, b;
+
+	r = anaglyph_encoding[mode][0];
+	g = anaglyph_encoding[mode][1];
+	b = anaglyph_encoding[mode][2];
+
+	if (s3d->is_float) {
+		float *rect_left  = s3d->rectf.left;
+		float *rect_right = s3d->rectf.right;
+		float *rect_to    = s3d->rectf.stereo;
+
+		if (channels == 3) {
+			for (y = 0; y < height; y++) {
+				float *to = rect_to + stride_to * y * 3;
+				float *from[2] = {
+				        rect_left  + stride_from * y * 3,
+				        rect_right + stride_from * y * 3,
+				};
+
+				for (x = 0; x < width; x++, from[0] += 3, from[1] += 3, to += 3) {
+					to[0] = from[r][0];
+					to[1] = from[g][1];
+					to[2] = from[b][2];
+				}
+			}
+		}
+		else if (channels == 4) {
+			for (y = 0; y < height; y++) {
+				float *to = rect_to + stride_to * y * 4;
+				float *from[2] = {
+				        rect_left  + stride_from * y * 4,
+				        rect_right + stride_from * y * 4,
+				};
+
+				for (x = 0; x < width; x++, from[0] += 4, from[1] += 4, to += 4) {
+					to[0] = from[r][0];
+					to[1] = from[g][1];
+					to[2] = from[b][2];
+					to[3] = MAX2(from[0][2], from[0][2]);
+				}
+			}
+		}
+	}
+	else {
+		uchar *rect_left  = s3d->rect.left;
+		uchar *rect_right = s3d->rect.right;
+		uchar *rect_to    = s3d->rect.stereo;
+
+		if (channels == 3) {
+			for (y = 0; y < height; y++) {
+				uchar *to = rect_to + stride_to * y * 3;
+				uchar *from[2] = {
+				        rect_left  + stride_from * y * 3,
+				        rect_right + stride_from * y * 3,
+				};
+
+				for (x = 0; x < width; x++, from[0] += 3, from[1] += 3, to += 3) {
+					to[0] = from[r][0];
+					to[1] = from[g][1];
+					to[2] = from[b][2];
+				}
+			}
+		}
+		else if (channels == 4) {
+			for (y = 0; y < height; y++) {
+				uchar *to = rect_to + stride_to * y * 4;
+				uchar *from[2] = {
+				        rect_left  + stride_from * y * 4,
+				        rect_right + stride_from * y * 4,
+				};
+
+				for (x = 0; x < width; x++, from[0] += 4, from[1] += 4, to += 4) {
+					to[0] = from[r][0];
+					to[1] = from[g][1];
+					to[2] = from[b][2];
+					to[3] = MAX2(from[0][2], from[0][2]);
+				}
+			}
+		}
+	}
+}
+
+static void imb_stereo3d_write_interlace(Stereo3DData *s3d, enum eStereo3dInterlaceType mode, const bool swap)
+{
+	int x, y;
+	size_t width = s3d->x;
+	size_t height = s3d->y;
+	const size_t channels = s3d->channels;
+
+	const int stride_from = width;
+	const int stride_to = width;
+
+	if (s3d->is_float) {
+		const float *rect_left  = s3d->rectf.left;
+		const float *rect_right = s3d->rectf.right;
+		float       *rect_to    = s3d->rectf.stereo;
+
+		switch (mode) {
+			case S3D_INTERLACE_ROW:
+			{
+				char i = (char) swap;
+				for (y = 0; y < height; y++) {
+					float *to = rect_to + stride_to * y * channels;
+					const float *from[2] = {
+					      rect_left  + stride_from * y * channels,
+					      rect_right + stride_from * y * channels,
+					};
+					memcpy(to, from[i], sizeof(float) * channels * stride_from);
+					i = !i;
+				}
+				break;
+			}
+			case S3D_INTERLACE_COLUMN:
+			{
+				if (channels == 1) {
+					for (y = 0; y < height; y++) {
+						float *to = rect_to + stride_to * y;
+						const float *from[2] = {
+						      rect_left  + stride_from * y,
+						      rect_right + stride_from * y,
+						};
+
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from[0] += 1, from[1] += 1, to += 1) {
+							to[0] = from[i][0];
+							i = !i;
+						}
+					}
+				}
+				else if (channels == 3) {
+					for (y = 0; y < height; y++) {
+						float *to = rect_to + stride_to * y * 3;
+						const float *from[2] = {
+						      rect_left  + stride_from * y * 3,
+						      rect_right + stride_from * y * 3,
+						};
+
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from[0] += 3, from[1] += 3, to += 3) {
+							copy_v3_v3(to, from[i]);
+							i = !i;
+						}
+					}
+				}
+				else if (channels == 4) {
+					for (y = 0; y < height; y++) {
+						float *to = rect_to + stride_to * y * channels;
+						const float *from[2] = {
+						      rect_left  + stride_from * y * channels,
+						      rect_right + stride_from * y * channels,
+						};
+
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from[0] += 4, from[1] += 4, to += 4) {
+							copy_v4_v4(to, from[i]);
+							i = !i;
+						}
+					}
+				}
+				break;
+			}
+			case S3D_INTERLACE_CHECKERBOARD:
+			{
+				if (channels == 1) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						float *to = rect_to + stride_to * y;
+						const float *from[2] = {
+						        rect_left  + stride_from * y,
+						        rect_right + stride_from * y,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from[0] += 1, from[1] += 1, to += 1) {
+							to[0] = from[j][0];
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				else if (channels == 3) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						float *to = rect_to + stride_to * y * 3;
+						const float *from[2] = {
+						        rect_left  + stride_from * y * 3,
+						        rect_right + stride_from * y * 3,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from[0] += 3, from[1] += 3, to += 3) {
+							copy_v3_v3(to, from[j]);
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				else if (channels == 4) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						float *to = rect_to + stride_to * y * 4;
+						const float *from[2] = {
+						        rect_left  + stride_from * y * 4,
+						        rect_right + stride_from * y * 4,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from[0] += 4, from[1] += 4, to += 4) {
+							copy_v4_v4(to, from[j]);
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				break;
+			}
+			default:
+			{
+				break;
+			}
+		}
+	}
+	else {
+		const uchar *rect_left  = s3d->rect.left;
+		const uchar *rect_right = s3d->rect.right;
+		uchar       *rect_to    = s3d->rect.stereo;
+
+		switch (mode) {
+			case S3D_INTERLACE_ROW:
+			{
+				char i = (char) swap;
+				for (y = 0; y < height; y++) {
+					uchar *to = rect_to + stride_to * y * channels;
+					const uchar *from[2] = {
+					      rect_left  + stride_from * y * channels,
+					      rect_right + stride_from * y * channels,
+					};
+					memcpy(to, from[i], sizeof(uchar) * channels * stride_from);
+					i = !i;
+				}
+				break;
+			}
+			case S3D_INTERLACE_COLUMN:
+			{
+				if (channels == 1) {
+					for (y = 0; y < height; y++) {
+						uchar *to = rect_to + stride_to * y;
+						const uchar *from[2] = {
+						        rect_left  + stride_from * y,
+						        rect_right + stride_from * y,
+						};
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from[0] += 1, from[1] += 1, to += 1) {
+							to[0] = from[i][0];
+							i = !i;
+						}
+					}
+				}
+				else if (channels == 3) {
+					for (y = 0; y < height; y++) {
+						uchar *to = rect_to + stride_to * y * 3;
+						const uchar *from[2] = {
+						        rect_left  + stride_from * y * 3,
+						        rect_right + stride_from * y * 3,
+						};
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from[0] += 3, from[1] += 3, to += 3) {
+							copy_v3_v3_char((char *)to, (char *)from[i]);
+							i = !i;
+						}
+					}
+				}
+				else if (channels == 4) {
+					for (y = 0; y < height; y++) {
+						uchar *to = rect_to + stride_to * y * 4;
+						const uchar *from[2] = {
+						        rect_left  + stride_from * y * 4,
+						        rect_right + stride_from * y * 4,
+						};
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from[0] += 4, from[1] += 4, to += 4) {
+							copy_v4_v4_char((char *)to, (char *)from[i]);
+							i = !i;
+						}
+					}
+				}
+				break;
+			}
+			case S3D_INTERLACE_CHECKERBOARD:
+			{
+				if (channels == 1) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						uchar *to = rect_to + stride_to * y;
+						const uchar *from[2] = {
+						        rect_left  + stride_from * y,
+						        rect_right + stride_from * y,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from[0] += 1, from[1] += 1, to += 1) {
+							to[0] = from[j][0];
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				else if (channels == 3) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						uchar *to = rect_to + stride_to * y * 3;
+						const uchar *from[2] = {
+						        rect_left  + stride_from * y * 3,
+						        rect_right + stride_from * y * 3,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from[0] += 3, from[1] += 3, to += 3) {
+							copy_v3_v3_char((char *)to, (char *)from[j]);
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				else if (channels == 4) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						uchar *to = rect_to + stride_to * y * 4;
+						const uchar *from[2] = {
+						        rect_left  + stride_from * y * 4,
+						        rect_right + stride_from * y * 4,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from[0] += 4, from[1] += 4, to += 4) {
+							copy_v4_v4_char((char *)to, (char *)from[j]);
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				break;
+			}
+			default:
+			{
+				break;
+			}
+		}
+	}
+}
+
+/* stereo3d output (s3d->rectf.stereo) is always unsqueezed */
+static void imb_stereo3d_write_sidebyside(Stereo3DData *s3d, const bool crosseyed)
+{
+	int y;
+	size_t width = s3d->x;
+	size_t height = s3d->y;
+	const size_t channels = s3d->channels;
+
+	const int stride_from = width;
+	const int stride_to = width * 2;
+
+	const int l =  (int) crosseyed;
+	const int r = !l;
+
+	if (s3d->is_float) {
+		const float *rect_left  = s3d->rectf.left;
+		const float *rect_right = s3d->rectf.right;
+		float       *rect_to    = s3d->rectf.stereo;
+
+		for (y = 0; y < height; y++) {
+			float *to = rect_to + stride_to * y * channels;
+			const float *from[2] = {
+			    rect_left  + stride_from * y * channels,
+			    rect_right + stride_from * y * channels,
+			};
+
+			memcpy(to, from[l], sizeof(float) * channels * stride_from);
+			memcpy(to + channels * stride_from, from[r], sizeof(float) * channels * stride_from);
+		}
+	}
+	else {
+		const uchar *rect_left  = s3d->rect.left;
+		const uchar *rect_right = s3d->rect.right;
+		uchar       *rect_to    = s3d->rect.stereo;
+
+		for (y = 0; y < height; y++) {
+			uchar *to = rect_to + stride_to * y * channels;
+			const uchar *from[2] = {
+			    rect_left  + stride_from * y * channels,
+			    rect_right + stride_from * y * channels,
+			};
+
+			memcpy(to, from[l], sizeof(uchar) * channels * stride_from);
+			memcpy(to + channels * stride_from, from[r], sizeof(uchar) * channels * stride_from);
+		}
+	}
+}
+
+/* stereo3d output (s3d->rectf.stereo) is always unsqueezed */
+static void imb_stereo3d_write_topbottom(Stereo3DData *s3d)
+{
+	int y;
+	size_t width = s3d->x;
+	size_t height = s3d->y;
+	const size_t channels = s3d->channels;
+
+	const int stride_from = width;
+	const int stride_to = width;
+
+	if (s3d->is_float) {
+		const float *rect_left  = s3d->rectf.left;
+		const float *rect_right = s3d->rectf.right;
+		float       *rect_to    = s3d->rectf.stereo;
+
+		for (y = 0; y < height; y++) {
+			float *to = rect_to + stride_to * y * channels;
+			const float *from[2] = {
+			    rect_left  + stride_from * y * channels,
+			    rect_right + stride_from * y * channels,
+			};
+
+			memcpy(to, from[1], sizeof(float) * channels * stride_from);
+			memcpy(to + channels * height * stride_from, from[0], sizeof(float) * channels * stride_from);
+		}
+	}
+	else {
+		const uchar *rect_left  = s3d->rect.left;
+		const uchar *rect_right = s3d->rect.right;
+		uchar       *rect_to    = s3d->rect.stereo;
+
+		for (y = 0; y < height; y++) {
+			uchar *to = rect_to + stride_to * y * channels;
+			const uchar *from[2] = {
+			    rect_left  + stride_from * y * channels,
+				rect_right + stride_from * y * channels,
+			};
+
+			memcpy(to, from[1], sizeof(uchar) * channels * stride_from);
+			memcpy(to + channels * height * stride_from, from[0], sizeof(uchar) * channels * stride_from);
+		}
+	}
+}
+
+/**************************** dimension utils ****************************************/
+
+void IMB_stereo3d_write_dimensions(
+        const char mode, const bool is_squeezed, const size_t width, const size_t height,
+        size_t *r_width, size_t *r_height)
+{
+	switch (mode) {
+		case S3D_DISPLAY_SIDEBYSIDE:
+		{
+			*r_width = is_squeezed ? width : width * 2;
+			*r_height = height;
+			break;
+		}
+		case S3D_DISPLAY_TOPBOTTOM:
+		{
+			*r_width = width;
+			*r_height = is_squeezed ? height : height * 2;
+			break;
+		}
+		case S3D_DISPLAY_ANAGLYPH:
+		case S3D_DISPLAY_INTERLACE:
+		default:
+		{
+			*r_width = width;
+			*r_height = height;
+			break;
+		}
+	}
+}
+
+void IMB_stereo3d_read_dimensions(
+        const char mode, const bool is_squeezed, const size_t width, const size_t height,
+        size_t *r_width, size_t *r_height)
+{
+	switch (mode) {
+		case S3D_DISPLAY_SIDEBYSIDE:
+		{
+			*r_width = is_squeezed ? width / 2 : width;
+			*r_height = height;
+			break;
+		}
+		case S3D_DISPLAY_TOPBOTTOM:
+		{
+			*r_width = width;
+			*r_height = is_squeezed ? height / 2 : height;
+			break;
+		}
+		case S3D_DISPLAY_ANAGLYPH:
+		case S3D_DISPLAY_INTERLACE:
+		default:
+		{
+			*r_width = width;
+			*r_height = height;
+			break;
+		}
+	}
+}
+
+/**************************** un/squeeze frame ****************************************/
+
+static void imb_stereo3d_squeeze_ImBuf(ImBuf *ibuf, Stereo3dFormat *s3d, const size_t x, const size_t y)
+{
+	if (ELEM(s3d->display_mode, S3D_DISPLAY_SIDEBYSIDE, S3D_DISPLAY_TOPBOTTOM) == false)
+		return;
+
+	if ((s3d->flag & S3D_SQUEEZED_FRAME) == 0)
+		return;
+
+	IMB_scaleImBuf_threaded(ibuf, x, y);
+}
+
+static void imb_stereo3d_unsqueeze_ImBuf(ImBuf *ibuf, Stereo3dFormat *s3d, const size_t x, const size_t y)
+{
+	if (ELEM(s3d->display_mode, S3D_DISPLAY_SIDEBYSIDE, S3D_DISPLAY_TOPBOTTOM) == false)
+		return;
+
+	if ((s3d->flag & S3D_SQUEEZED_FRAME) == 0)
+		return;
+
+	IMB_scaleImBuf_threaded(ibuf, x, y);
+}
+
+static void imb_stereo3d_squeeze_rectf(float *rectf, Stereo3dFormat *s3d, const size_t x, const size_t y, const size_t channels)
+{
+	ImBuf *ibuf;
+	size_t width, height;
+
+	if (ELEM(s3d->display_mode, S3D_DISPLAY_SIDEBYSIDE, S3D_DISPLAY_TOPBOTTOM) == false)
+		return;
+
+	if ((s3d->flag & S3D_SQUEEZED_FRAME) == 0)
+		return;
+
+	/* creates temporary imbuf to store the rectf */
+	IMB_stereo3d_write_dimensions(s3d->display_mode, false, x, y, &width, &height);
+	ibuf = IMB_allocImBuf(width, height, channels, IB_rectfloat);
+
+	IMB_buffer_float_from_float(
+	        ibuf->rect_float, rectf, channels,
+	        IB_PROFILE_LINEAR_RGB, IB_PROFILE_LINEAR_RGB, false,
+	        width, height, width, width);
+
+	IMB_scaleImBuf_threaded(ibuf, x, y);
+	rectf = MEM_dupallocN(ibuf->rect_float);
+	IMB_freeImBuf(ibuf);
+}
+
+static void imb_stereo3d_squeeze_rect(int *rect, Stereo3dFormat *s3d, const size_t x, const size_t y, const size_t channels)
+{
+	ImBuf *ibuf;
+	size_t width, height;
+
+	if (ELEM(s3d->display_mode, S3D_DISPLAY_SIDEBYSIDE, S3D_DISPLAY_TOPBOTTOM) == false)
+		return;
+
+	if ((s3d->flag & S3D_SQUEEZED_FRAME) == 0)
+		return;
+
+	/* creates temporary imbuf to store the rectf */
+	IMB_stereo3d_write_dimensions(s3d->display_mode, false, x, y, &width, &height);
+	ibuf = IMB_allocImBuf(width, height, channels, IB_rect);
+
+	IMB_buffer_byte_from_byte(
+	        (unsigned char *)ibuf->rect, (unsigned char *)rect,
+	        IB_PROFILE_SRGB, IB_PROFILE_SRGB, false,
+	        width, height, width, width);
+
+	IMB_scaleImBuf_threaded(ibuf, x, y);
+	rect = MEM_dupallocN(ibuf->rect);
+	IMB_freeImBuf(ibuf);
+}
+
+
+/*************************** preparing to call the write functions **************************/
+
+static void imb_stereo3d_data_initialize(
+        Stereo3DData *s3d_data, const bool is_float,
+        const size_t x, const size_t y, const size_t channels,
+        int *rect_left, int *rect_right, int *rect_stereo,
+        float *rectf_left, float *rectf_right, float *rectf_stereo)
+{
+	s3d_data->is_float = is_float;
+	s3d_data->x = x;
+	s3d_data->y = y;
+	s3d_data->channels = channels;
+	s3d_data->rect.left    = (uchar *)rect_left;
+	s3d_data->rect.right   = (uchar *)rect_right;
+	s3d_data->rect.stereo  = (uchar *)rect_stereo;
+	s3d_data->rectf.left   = rectf_left;
+	s3d_data->rectf.right  = rectf_right;
+	s3d_data->rectf.stereo = rectf_stereo;
+}
+
+int *IMB_stereo3d_from_rect(
+        ImageFormatData *im_format, const size_t x, const size_t y, const size_t channels,
+        int *rect_left, int *rect_right)
+{
+	int *r_rect;
+	Stereo3DData s3d_data = {{NULL}};
+	size_t width, height;
+	const bool is_float = im_format->depth > 8;
+
+	IMB_stereo3d_write_dimensions(im_format->stereo3d_format.display_mode, false, x, y, &width, &height);
+	r_rect = MEM_mallocN(channels * sizeof(int) * width * height, __func__);
+
+	imb_stereo3d_data_initialize(&s3d_data, is_float, x, y, channels, rect_left, rect_right, r_rect, NULL, NULL, NULL);
+	imb_stereo3d_write_doit(&s3d_data, &im_format->stereo3d_format);
+	imb_stereo3d_squeeze_rect(r_rect, &im_format->stereo3d_format, x, y, channels);
+
+	return r_rect;
+}
+
+float *IMB_stereo3d_from_rectf(
+        ImageFormatData *im_format, const size_t x, const size_t y, const size_t channels,
+        float *rectf_left, float *rectf_right)
+{
+	float *r_rectf;
+	Stereo3DData s3d_data = {{NULL}};
+	size_t width, height;
+	const bool is_float = im_format->depth > 8;
+
+	IMB_stereo3d_write_dimensions(im_format->stereo3d_format.display_mode, false, x, y, &width, &height);
+	r_rectf = MEM_mallocN(channels * sizeof(float) * width * height, __func__);
+
+	imb_stereo3d_data_initialize(&s3d_data, is_float, x, y, channels, NULL, NULL, NULL, rectf_left, rectf_right, r_rectf);
+	imb_stereo3d_write_doit(&s3d_data, &im_format->stereo3d_format);
+	imb_stereo3d_squeeze_rectf(r_rectf, &im_format->stereo3d_format, x, y, channels);
+
+	return r_rectf;
+}
+
+/* left/right are always float */
+ImBuf *IMB_stereo3d_ImBuf(ImageFormatData *im_format, ImBuf *ibuf_left, ImBuf *ibuf_right)
+{
+	ImBuf *ibuf_stereo = NULL;
+	Stereo3DData s3d_data = {{NULL}};
+	size_t width, height;
+	const bool is_float = im_format->depth > 8;
+
+	IMB_stereo3d_write_dimensions(im_format->stereo3d_format.display_mode, false, ibuf_left->x, ibuf_left->y, &width, &height);
+	ibuf_stereo = IMB_allocImBuf(width, height, ibuf_left->planes, (is_float ? IB_rectfloat : IB_rect));
+
+	/* copy flags for IB_fields and other settings */
+	ibuf_stereo->flags = ibuf_left->flags;
+
+	imb_stereo3d_data_initialize(
+	        &s3d_data, is_float, ibuf_left->x, ibuf_left->y, 4,
+	        (int *)ibuf_left->rect, (int *)ibuf_right->rect, (int *)ibuf_stereo->rect,
+	        ibuf_left->rect_float, ibuf_right->rect_float, ibuf_stereo->rect_float);
+
+	imb_stereo3d_write_doit(&s3d_data, &im_format->stereo3d_format);
+	imb_stereo3d_squeeze_ImBuf(ibuf_stereo, &im_format->stereo3d_format, ibuf_left->x, ibuf_left->y);
+
+	return ibuf_stereo;
+}
+
+static void imb_stereo3d_write_doit(Stereo3DData *s3d_data, Stereo3dFormat *s3d)
+{
+	switch (s3d->display_mode) {
+		case S3D_DISPLAY_ANAGLYPH:
+			imb_stereo3d_write_anaglyph(s3d_data, s3d->anaglyph_type);
+			break;
+		case S3D_DISPLAY_INTERLACE:
+			imb_stereo3d_write_interlace(s3d_data, s3d->interlace_type, (s3d->flag & S3D_INTERLACE_SWAP) != 0);
+			break;
+		case S3D_DISPLAY_SIDEBYSIDE:
+			imb_stereo3d_write_sidebyside(s3d_data, (s3d->flag & S3D_SIDEBYSIDE_CROSSEYED) != 0);
+			break;
+		case S3D_DISPLAY_TOPBOTTOM:
+			imb_stereo3d_write_topbottom(s3d_data);
+			break;
+		default:
+			break;
+	}
+}
+
+/******************************** reading stereo imbufs **********************/
+
+static void imb_stereo3d_read_anaglyph(Stereo3DData *s3d, enum eStereo3dAnaglyphType mode)
+{
+	int x, y;
+	size_t width = s3d->x;
+	size_t height = s3d->y;
+	const size_t channels = s3d->channels;
+
+	const int stride_from = width;
+	const int stride_to = width;
+
+	int anaglyph_encoding[3][3] = {
+	    {0, 1, 1},
+	    {1, 0, 1},
+	    {0, 0, 1},
+	};
+
+	int r, g, b;
+
+	r = anaglyph_encoding[mode][0];
+	g = anaglyph_encoding[mode][1];
+	b = anaglyph_encoding[mode][2];
+
+	if (s3d->is_float) {
+		float *rect_left  = s3d->rectf.left;
+		float *rect_right = s3d->rectf.right;
+		float *rect_from  = s3d->rectf.stereo;
+
+		if (channels == 3) {
+			for (y = 0; y < height; y++) {
+				float *from = rect_from + stride_from * y * 3;
+				float *to[2] = {
+					rect_left  + stride_to * y * 3,
+					rect_right + stride_to * y * 3,
+				};
+
+				for (x = 0; x < width; x++, from += 3, to[0] += 3, to[1] += 3) {
+					to[r][0] = from[0];
+					to[g][1] = from[1];
+					to[b][2] = from[2];
+				}
+			}
+		}
+		else if (channels == 4) {
+			for (y = 0; y < height; y++) {
+				float *from = rect_from + stride_from * y * 4;
+				float *to[2] = {
+					rect_left  + stride_to * y * 4,
+					rect_right + stride_to * y * 4,
+				};
+
+				for (x = 0; x < width; x++, from += 4, to[0] += 4, to[1] += 4) {
+					to[r][0] = from[0];
+					to[g][1] = from[1];
+					to[b][2] = from[2];
+					to[0][3] = to[1][3] = from[3];
+				}
+			}
+		}
+	}
+	else {
+		uchar *rect_left  = s3d->rect.left;
+		uchar *rect_right = s3d->rect.right;
+		uchar *rect_from  = s3d->rect.stereo;
+
+		if (channels == 3) {
+			for (y = 0; y < height; y++) {
+				uchar *from = rect_from + stride_from * y * 3;
+				uchar *to[2] = {
+					rect_left  + stride_to * y * 3,
+					rect_right + stride_to * y * 3,
+				};
+
+				for (x = 0; x < width; x++, from += 3, to[0] += 3, to[1] += 3) {
+					to[r][0] = from[0];
+					to[g][1] = from[1];
+					to[b][2] = from[2];
+				}
+			}
+		}
+		else if (channels == 4) {
+			for (y = 0; y < height; y++) {
+				uchar *from = rect_from + stride_from * y * 4;
+				uchar *to[2] = {
+					rect_left  + stride_to * y * 4,
+					rect_right + stride_to * y * 4,
+				};
+
+				for (x = 0; x < width; x++, from += 4, to[0] += 4, to[1] += 4) {
+					to[r][0] = from[0];
+					to[g][1] = from[1];
+					to[b][2] = from[2];
+					to[0][3] = to[1][3] = from[3];
+				}
+			}
+		}
+	}
+}
+
+static void imb_stereo3d_read_interlace(Stereo3DData *s3d, enum eStereo3dInterlaceType mode, const bool swap)
+{
+	int x, y;
+	size_t width = s3d->x;
+	size_t height = s3d->y;
+	const size_t channels = s3d->channels;
+
+	const int stride_from = width;
+	const int stride_to = width;
+
+	if (s3d->is_float) {
+		float *rect_left  = s3d->rectf.left;
+		float *rect_right = s3d->rectf.right;
+		const float *rect_from = s3d->rectf.stereo;
+
+		switch (mode) {
+			case S3D_INTERLACE_ROW:
+			{
+				char i = (char) swap;
+				for (y = 0; y < height; y++) {
+					const float *from = rect_from + stride_from * y * channels;
+					float *to[2] = {
+						rect_left  + stride_to * y * channels,
+						rect_right + stride_to * y * channels,
+					};
+					memcpy(to[i], from, sizeof(float) * channels * stride_to);
+					i = !i;
+				}
+				break;
+			}
+			case S3D_INTERLACE_COLUMN:
+			{
+				if (channels == 1) {
+					for (y = 0; y < height; y++) {
+						const float *from = rect_from + stride_from * y;
+						float *to[2] = {
+							rect_left  + stride_to * y,
+							rect_right + stride_to * y,
+						};
+
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from += 1, to[0] += 1, to[1] += 1) {
+							to[i][0] = from[0];
+							i = !i;
+						}
+					}
+				}
+				else if (channels == 3) {
+					for (y = 0; y < height; y++) {
+						const float *from = rect_from + stride_from * y * 3;
+						float *to[2] = {
+							rect_left  + stride_to * y * 3,
+							rect_right + stride_to * y * 3,
+						};
+
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from += 3, to[0] += 3, to[1] += 3) {
+							copy_v3_v3(to[i], from);
+							i = !i;
+						}
+					}
+				}
+				else if (channels == 4) {
+					for (y = 0; y < height; y++) {
+						const float *from = rect_from + stride_from * y * channels;
+						float *to[2] = {
+							rect_left  + stride_to * y * channels,
+							rect_right + stride_to * y * channels,
+						};
+
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from += 4, to[0] += 4, to[1] += 4) {
+							copy_v4_v4(to[i], from);
+							i = !i;
+						}
+					}
+				}
+				break;
+			}
+			case S3D_INTERLACE_CHECKERBOARD:
+			{
+				if (channels == 1) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						const float *from = rect_from + stride_from * y;
+						float *to[2] = {
+							rect_left  + stride_to * y,
+							rect_right + stride_to * y,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from += 1, to[0] += 1, to[1] += 1) {
+							to[j][0] = from[0];
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				else if (channels == 3) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						const float *from = rect_from + stride_from * y * 3;
+						float *to[2] = {
+							rect_left  + stride_to * y * 3,
+							rect_right + stride_to * y * 3,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from += 3, to[0] += 3, to[1] += 3) {
+							copy_v3_v3(to[j], from);
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				else if (channels == 4) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						const float *from = rect_from + stride_from * y * 4;
+						float *to[2] = {
+							rect_left  + stride_to * y * 4,
+							rect_right + stride_to * y * 4,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from += 4, to[0] += 4, to[1] += 4) {
+							copy_v4_v4(to[j], from);
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				break;
+			}
+			default:
+			{
+				break;
+			}
+		}
+	}
+	else {
+		uchar       *rect_left  = s3d->rect.right;
+		uchar       *rect_right = s3d->rect.left;
+		const uchar *rect_from  = s3d->rect.stereo;
+
+		switch (mode) {
+			case S3D_INTERLACE_ROW:
+			{
+				char i = (char) swap;
+				for (y = 0; y < height; y++) {
+					const uchar *from = rect_from + stride_from * y * channels;
+					uchar *to[2] = {
+						rect_left  + stride_to * y * channels,
+						rect_right + stride_to * y * channels,
+					};
+					memcpy(to[i], from, sizeof(uchar) * channels * stride_to);
+					i = !i;
+				}
+				break;
+			}
+			case S3D_INTERLACE_COLUMN:
+			{
+				if (channels == 1) {
+					for (y = 0; y < height; y++) {
+						const uchar *from = rect_from + stride_from * y;
+						uchar *to[2] = {
+							rect_left  + stride_to * y,
+							rect_right + stride_to * y,
+						};
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from += 1, to[0] += 1, to[1] += 1) {
+							to[i][0] = from[0];
+							i = !i;
+						}
+					}
+				}
+				else if (channels == 3) {
+					for (y = 0; y < height; y++) {
+						const uchar *from = rect_from + stride_from * y * 3;
+						uchar *to[2] = {
+							rect_left  + stride_to * y * 3,
+							rect_right + stride_to * y * 3,
+						};
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from += 3, to[0] += 3, to[1] += 3) {
+							copy_v3_v3_char((char *)to[i], (char *)from);
+							i = !i;
+						}
+					}
+				}
+				else if (channels == 4) {
+					for (y = 0; y < height; y++) {
+						const uchar *from = rect_from + stride_from * y * 4;
+						uchar *to[2] = {
+							rect_left  + stride_to * y * 4,
+							rect_right + stride_to * y * 4,
+						};
+						char i = (char) swap;
+						for (x = 0; x < width; x++, from += 4, to[0] += 4, to[1] += 4) {
+							copy_v4_v4_char((char *)to[i], (char *)from);
+							i = !i;
+						}
+					}
+				}
+				break;
+			}
+			case S3D_INTERLACE_CHECKERBOARD:
+			{
+				if (channels == 1) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						const uchar *from = rect_from + stride_from * y;
+						uchar *to[2] = {
+							rect_left  + stride_to * y,
+							rect_right + stride_to * y,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from += 1, to[0] += 1, to[1] += 1) {
+							to[j][0] = from[0];
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				else if (channels == 3) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						const uchar *from = rect_from + stride_from * y * 3;
+						uchar *to[2] = {
+							rect_left  + stride_to * y * 3,
+							rect_right + stride_to * y * 3,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from += 3, to[0] += 3, to[1] += 3) {
+							copy_v3_v3_char((char *)to[j], (char *)from);
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				else if (channels == 4) {
+					char i = (char) swap;
+					for (y = 0; y < height; y++) {
+						const uchar *from = rect_from + stride_from * y * 4;
+						uchar *to[2] = {
+							rect_left  + stride_to * y * 4,
+							rect_right + stride_to * y * 4,
+						};
+						char j = i;
+						for (x = 0; x < width; x++, from += 4, to[0] += 4, to[1] += 4) {
+							copy_v4_v4_char((char *)to[j], (char *)from);
+							j = !j;
+						}
+						i = !i;
+					}
+				}
+				break;
+			}
+			default:
+			{
+				break;
+			}
+		}
+	}
+}
+
+/* stereo input (s3d->rectf.stereo) is always unsqueezed */
+static void imb_stereo3d_read_sidebyside(Stereo3DData *s3d, const bool crosseyed)
+{
+	int y;
+	size_t width = s3d->x;
+	size_t height = s3d->y;
+	const size_t channels = s3d->channels;
+
+	const int stride_from = width * 2;
+	const int stride_to = width;
+
+	const int l =  (int) crosseyed;
+	const int r = !l;
+
+	if (s3d->is_float) {
+		float *rect_left = s3d->rectf.left;
+		float *rect_right = s3d->rectf.right;
+		const float *rect_from = s3d->rectf.stereo;
+
+		for (y = 0; y < height; y++) {
+			const float *from = rect_from + stride_from * y * channels;
+			float *to[2] = {
+			    rect_left  + stride_to * y * channels,
+			    rect_right + stride_to * y * channels,
+			};
+
+			memcpy(to[l], from, sizeof(float) * channels * stride_to);
+			memcpy(to[r], from + channels * stride_to, sizeof(float) * channels * stride_to);
+		}
+	}
+	else {
+		uchar       *rect_left  = s3d->rect.left;
+		uchar       *rect_right = s3d->rect.right;
+		const uchar *rect_from  = s3d->rect.stereo;
+
+		/* always RGBA input/output */
+		for (y = 0; y < height; y++) {
+			const uchar *from = rect_from + stride_from * y * channels;
+			uchar *to[2] = {
+			    rect_left  + stride_to * y * channels,
+			    rect_right + stride_to * y * channels,
+			};
+
+			memcpy(to[l], from, sizeof(uchar) * channels * stride_to);
+			memcpy(to[r], from + channels * stride_to, sizeof(uchar) * channels * stride_to);
+		}
+	}
+}
+
+/* stereo input (s3d->rectf.stereo) is always unsqueezed */
+static void imb_stereo3d_read_topbottom(Stereo3DData *s3d)
+{
+	int y;
+	size_t width = s3d->x;
+	size_t height = s3d->y;
+	const size_t channels = s3d->channels;
+
+	const int stride_from = width;
+	const int stride_to = width;
+
+	if (s3d->is_float) {
+		float *rect_left = s3d->rectf.left;
+		float *rect_right = s3d->rectf.right;
+		const float *rect_from = s3d->rectf.stereo;
+
+		for (y = 0; y < height; y++) {
+			const float *from = rect_from + stride_from * y * channels;
+			float *to[2] = {
+			    rect_left  + stride_to * y * channels,
+			    rect_right + stride_to * y * channels,
+			};
+
+			memcpy(to[1], from, sizeof(float) * channels * stride_to);
+			memcpy(to[0], from + channels * height * stride_to, sizeof(float) * channels * stride_to);
+		}
+	}
+	else {
+		uchar       *rect_left  = s3d->rect.left;
+		uchar       *rect_right = s3d->rect.right;
+		const uchar *rect_from  = s3d->rect.stereo;
+
+		for (y = 0; y < height; y++) {
+			const uchar *from = rect_from + stride_from * y * channels;
+			uchar *to[2] = {
+			    rect_left  + stride_to * y * channels,
+				rect_right + stride_to * y * channels,
+			};
+
+			memcpy(to[1], from, sizeof(uchar) * channels * stride_to);
+			memcpy(to[0], from + channels * height * stride_to, sizeof(uchar) * channels * stride_to);
+		}
+	}
+}
+
+
+/*************************** preparing to call the read functions **************************/
+
+/* reading a stereo encoded ibuf (*left) and generating two ibufs from it (*left and *right) */
+void IMB_ImBufFromStereo3d(
+        Stereo3dFormat *s3d, ImBuf *ibuf_stereo3d,
+        ImBuf **r_ibuf_left, ImBuf **r_ibuf_right)
+{
+	Stereo3DData s3d_data = {{NULL}};
+	ImBuf *ibuf_left, *ibuf_right;
+	size_t width, height;
+	const bool is_float = (ibuf_stereo3d->rect_float != NULL);
+
+	IMB_stereo3d_read_dimensions(
+	        s3d->display_mode, ((s3d->flag & S3D_SQUEEZED_FRAME) == 0), ibuf_stereo3d->x, ibuf_stereo3d->y,
+	        &width, &height);
+
+	ibuf_left = IMB_allocImBuf(width, height, ibuf_stereo3d->planes, (is_float ? IB_rectfloat : IB_rect));
+	ibuf_right = IMB_allocImBuf(width, height, ibuf_stereo3d->planes, (is_float ? IB_rectfloat : IB_rect));
+
+	/* copy flags for IB_fields and other settings */
+	ibuf_left->flags = ibuf_stereo3d->flags;
+	ibuf_right->flags = ibuf_stereo3d->flags;
+
+	/* we always work with unsqueezed formats */
+	IMB_stereo3d_write_dimensions(
+	        s3d->display_mode, ((s3d->flag & S3D_SQUEEZED_FRAME) == 0), ibuf_stereo3d->x, ibuf_stereo3d->y,
+	        &width, &height);
+	imb_stereo3d_unsqueeze_ImBuf(ibuf_stereo3d, s3d, width, height);
+
+	imb_stereo3d_data_initialize(
+	        &s3d_data, is_float, ibuf_left->x, ibuf_left->y, 4,
+	        (int *)ibuf_left->rect, (int *)ibuf_right->rect, (int *)ibuf_stereo3d->rect,
+	        ibuf_left->rect_float, ibuf_right->rect_float, ibuf_stereo3d->rect_float);
+
+	imb_stereo3d_read_doit(&s3d_data, s3d);
+
+	if (ibuf_stereo3d->flags & (IB_zbuf | IB_zbuffloat)) {
+		if (is_float) {
+			addzbuffloatImBuf(ibuf_left);
+			addzbuffloatImBuf(ibuf_right);
+		}
+		else {
+			addzbufImBuf(ibuf_left);
+			addzbufImBuf(ibuf_right);
+		}
+
+		imb_stereo3d_data_initialize(
+		        &s3d_data, is_float, ibuf_left->x, ibuf_left->y, 1,
+		        (int *)ibuf_left->zbuf, (int *)ibuf_right->zbuf, (int *)ibuf_stereo3d->zbuf,
+		        ibuf_left->zbuf_float, ibuf_right->zbuf_float, ibuf_stereo3d->zbuf_float);
+
+		imb_stereo3d_read_doit(&s3d_data, s3d);
+	}
+
+	IMB_freeImBuf(ibuf_stereo3d);
+
+	*r_ibuf_left = ibuf_left;
+	*r_ibuf_right = ibuf_right;
+}
+
+static void imb_stereo3d_read_doit(Stereo3DData *s3d_data, Stereo3dFormat *s3d)
+{
+	switch (s3d->display_mode) {
+		case S3D_DISPLAY_ANAGLYPH:
+			imb_stereo3d_read_anaglyph(s3d_data, s3d->anaglyph_type);
+			break;
+		case S3D_DISPLAY_INTERLACE:
+			imb_stereo3d_read_interlace(s3d_data, s3d->interlace_type, (s3d->flag & S3D_INTERLACE_SWAP) != 0);
+			break;
+		case S3D_DISPLAY_SIDEBYSIDE:
+			imb_stereo3d_read_sidebyside(s3d_data, (s3d->flag & S3D_SIDEBYSIDE_CROSSEYED) != 0);
+			break;
+		case S3D_DISPLAY_TOPBOTTOM:
+			imb_stereo3d_read_topbottom(s3d_data);
+			break;
+		default:
+			break;
+	}
+}