- Allow ocio to use aesthetic transforms(looks)
- Make luts for this. Examine following code:
https://github.com/nvkelso/color-oracle-java/blob/master/src/ika/colororacle/ColorOracle.java
http://vision.psychol.cam.ac.uk/jdmollon/papers/colourmaps.pdf
https://github.com/nvkelso/color-oracle-java/blob/master/src/ika/colororacle/ColorOracle.java
http://vision.psychol.cam.ac.uk/jdmollon/papers/colourmaps.pdf
Status | Assigned | Task | ||
---|---|---|---|---|
Restricted Maniphest Task | ||||
Open | None | T2395 Colorblindness display filters and ocio look transforms |
Ok, after some research:
There's two ways:
Either by converting our rgb space with a matrix to LMS cone response space thingy-mabob. Then doing something in LMS space, and then converting back.
Or the color matrices from the last wayback machine link(which seems to be based on http://colorfilter.wickline.org/):
'Normal':[1,0,0,0,0, 0,1,0,0,0, 0,0,1,0,0, 0,0,0,1,0, 0,0,0,0,1], //identity matrix 'Protanopia':[0.567,0.433,0,0,0, 0.558,0.442,0,0,0, 0,0.242,0.758,0,0, 0,0,0,1,0, 0,0,0,0,1], 'Protanomaly':[0.817,0.183,0,0,0, 0.333,0.667,0,0,0, 0,0.125,0.875,0,0, 0,0,0,1,0, 0,0,0,0,1], 'Deuteranopia':[0.625,0.375,0,0,0, 0.7,0.3,0,0,0, 0,0.3,0.7,0,0, 0,0,0,1,0, 0,0,0,0,1], 'Deuteranomaly':[0.8,0.2,0,0,0, 0.258,0.742,0,0,0, 0,0.142,0.858,0,0, 0,0,0,1,0, 0,0,0,0,1], 'Tritanopia':[0.95,0.05,0,0,0, 0,0.433,0.567,0,0, 0,0.475,0.525,0,0, 0,0,0,1,0, 0,0,0,0,1], 'Tritanomaly':[0.967,0.033,0,0,0, 0,0.733,0.267,0,0, 0,0.183,0.817,0,0, 0,0,0,1,0, 0,0,0,0,1], 'Achromatopsia':[0.299,0.587,0.114,0,0, 0.299,0.587,0.114,0,0, 0.299,0.587,0.114,0,0, 0,0,0,1,0, 0,0,0,0,1], 'Achromatomaly':[0.618,0.320,0.062,0,0, 0.163,0.775,0.062,0,0, 0.163,0.320,0.516,0,0,0,0,0,1,0,0,0,0,0]}[v]);
The later should be possible to be made into LUTs.
The one used in the gimp plugin is pretty sohpisticated and even allows for multiple severerities, but might be UI wise too sophisticated for a mere OCIO Look LUT.
After testing it out with Gimp's channel mixer, it seems that the javascript ones are complete nonsense luminosity wise. The gimp registery plugin's interpretation of the 2009 paper has the best results in regards to differences between luminosity.