| Age | Commit message (Collapse) | Author |
|---|
|
old issue, the formulas here were never quite right, should all work ok now
with byte and float images.
Some differences:
* Colors with zero alpha from the background will never have an influence, so
you don't get alpha fringes when painting over such areas. This does give
hard edges when looking at the RGB channels alone, but there's no way to
avoid that and fringes at the same time, same behavior as other painting apps.
* Add/Subtract/Multiply/Lighten/Darken now leave the alpha channel unchanged
and work only the RGB channels, again same behavior as many other apps.
* Erase/Add alpha now compensates for premultiplied float images to keep the
straight RGB colors the same.
Next: fix projection painting.
|
|
|
|
Additional changes:
- Made mipmapping operate with unsigned short instead of char
which allowed to eliminate extra division by 255, so prevision
should be a bit better now.
- Actually, this is not real unsigned short range, but it's a
range of 255*255 which is more convenient for mipmapping, so
made conversion functions private for scaling.c
Not sure it worth making this functions operate in 65535
range, for now current behavior seems to be just fine.
|
|
Regression was caused by alpha premul cleanup commit and the reason
of slowdown was uchar <-> float conversion which is slow.
Replaced with uchar <-> int conversion which seeps to be accurate
enough and mostly eliminates slowdown.
Slowdown was easy to notice when movie clip is used for 3d vierport
background and undistortion is enabled. In this case every frame
will re-calculate mipmaps.
It's still a nit slower than mipmap generation before cleanup
commit, but couldn't think about extra boost here atm.
|
|
This assumptions are now made:
- Internally float buffers are always linear alpha-premul colors
- Readers should worry about delivering float buffers with that
assumptions.
- There's an input image setting to say whether it's stored with
straight/premul alpha on the disk.
- Byte buffers are now assumed have straight alpha, readers should
deliver straight alpha.
Some implementation details:
- Removed scene's color unpremultiply setting, which was very
much confusing and was wrong for default settings.
Now all renderers assumes to deliver premultiplied alpha.
- IMB_buffer_byte_from_float will now linearize alpha when
converting from buffer.
- Sequencer's effects were changed to assume bytes have got
straight alpha. Most of effects will work with bytes still,
however for glow it was more tricky to avoid data loss, so
there's a commented out glow implementation which converts
byte buffer to floats first, operates on floats and returns
bytes back. It's slower and not sure if it should actually
be used -- who're using glow on alpha anyway?
- Sequencer modifiers should also be working nice with straight
bytes now.
- GLSL preview will predivide float textures to make nice shading,
shading with byte textures worked nice (GLSL was assuming straight
alpha).
- Blender Internal will set alpha=1 to the whole sky. The same
happens in Cycles and there's no way to avoid this -- sky is
neither straight nor premul and doesn't fit color pipeline well.
- Straight alpha mode for render result was also eliminated.
- Conversion to correct alpha need to be done before linearizing
float buffer.
- TIFF will now load and save files with proper alpha mode setting
in file meta data header.
- Remove Use Alpha from texture mapping and replaced with image
datablock setting.
Behaves much more predictable and clear from code point of view
and solves possible regressions when non-premultiplied images were
used as textures with ignoring alpha channel.
|
|
* MEM_CacheLimitier - Size type to int conversion, should be safe for now (doing my best Bill Gates 640k impression)
* OpenNL CMakeLists.txt - MSVC and GCC have slightly different ways to remove definitions (DEBUG) without the compiler complaining
* BLI_math inlines - The include guard name and inline option macro name should be different. Suppressed warning about not exporting any symbols from inline math library
* BLI string / utf8 - Fixed some inconsistencies between declarations and definitions
* nodes - node_composite_util is apparently not used unless you enable the legacy compositor, so it should not be compiled in that case.
Leaving out changes to BLI_fileops for now, need to do more testing.
|
|
not working at all.
|
|
C with gcc.
helps for finding unused functions and making functions static, also did some minor code cleanup.
|
|
|
|
|
|
|
|
|
|
|
|
without the underscores these clogged up the namespace for autocompleation which was annoying.
|
|
many, also don't check uchar range after casting to int, this can still cause
overflow with large float values.
|
|
also add rgba_float_to_uchar, rgba_uchar_to_float
|
|
rgb_uchar_to_float and swap args (math functions mostly have dest arg first like strcpy).
|
|
use rgb_to_grayscale in more places.
|
|
brushes were doing curve lookups within this macro for example.
|
|
* Accelerated sRGB <=> linear conversion using lookup table, this can speed up
loading of images in the compositor and simple renders quite a bit.
* Dithering now uses the Floyd-Steinberg algorithm. Previously it would simply
randomize each pixel slightly, adding noise, now that should be reduced.
Patch #29309 by David M.
|
|
byte => float, float => float, byte => byte conversions with profile, dither
and predivide. Previously code for this was spread out too much.
There should be no functional changes, this is so the predivide/table/dither
patches can work correctly.
|
|
http://markmail.org/message/fp7ozcywxum3ar7n
|
|
convert to grayscale when saving renders rather then only writing the red channel.
|
|
|
|
|
|
- made theme colors for mesh edge len & face angle/area display.
- use %g rather then %f for float display, trims unneeded zeros.
- store cached 2d and 3d text color as bytes rather then floats, compare when drawing to avoid setting the context.
- use unsigned char for more color functions, avoids casting to glColorubv().
|
|
|
|
functions
By Luca Bonavita (mindrones)
From detailed description: This patch doesnt change functionality, but uses the existing color math functions from math_color.c into
sequencer_draw.c.
|
|
the render engine assumes the RenderResult's rectf is not in linear color space when color management is disabled so the sequencer and opengl render need to follow this else it results in dark images.
|
|
YCbCr conversion function according to ITU-R BT.601/709 and JFIF
Clarifies color space usage in rgb->ycc conversion, doesn't change existing results.
|
|
|
|
* inline some more functions, from math_base and math_vector
* also made some changes to the way inline is done so it can
work for more than one file
* reflect_v3_v3v3 requires input vectors to be normalized now.
* added rgb_to_grayscale
* added zero_v4, copy_v4_v4, swap_v4_v4, is_one_v3
* added box_clip_bounds_m4 to clip a bounding box against a
projection matrix
|
|
similar to sequence editor.
--> http://mke3.net/blender/devel/2.5/color_balance_node.jpg
Also added 0 key (zero key) shortcut when mouse is over a button, to reset it to its default value.
Same as the RMB menu ->Reset to Default, except for color wheels, it only resets the hue/sat/value
components that that widget affects.
Peter/Xavier: The existing color balance code can generate NaNs (fractional power of a negative),
which causes havoc along the image pipeline. I added a check in the node code to prevent this.
Still plenty of potential for lots of better colour correction tools in the compositor, just needs time...
|
|
A solid color backdrop is now drawn behind the mute/protect toggles and sliders, reducing the visual clutter with long names still appearing behind the UI widgets.
|
|
After testing and feedback, I've decided to slightly modify the way color
management works internally. While the previous method worked well for
rendering, was a smaller transition and had some advantages over this
new method, it was a bit more ambiguous, and was making things difficult
for other areas such as compositing.
This implementation now considers all color data (with only a couple of
exceptions such as brush colors) to be stored in linear RGB color space,
rather than sRGB as previously. This brings it in line with Nuke, which also
operates this way, quite successfully. Color swatches, pickers, color ramp
display are now gamma corrected to display gamma so you can see what
you're doing, but the numbers themselves are considered linear. This
makes understanding blending modes more clear (a 0.5 value on overlay
will not change the result now) as well as making color swatches act more
predictably in the compositor, however bringing over color values from
applications like photoshop or gimp, that operate in a gamma space,
will give identical results.
This commit will convert over existing files saved by earlier 2.5 versions to
work generally the same, though there may be some slight differences with
things like textures. Now that we're set on changing other areas of shading,
this won't be too disruptive overall.
I've made a diagram explaining the pipeline here:
http://mke3.net/blender/devel/2.5/25_linear_workflow_pipeline.png
and some docs here:
http://www.blender.org/development/release-logs/blender-250/color-management/
|
|
* New header and source files.
* Still need a few tweaks before switching code to use them.
|
