fix primaries used for display p3 gamut #331
Conversation
ram-mohan
force-pushed
the
p3primaries
branch
from
80bad93 to
9f25e09
Compare
ram-mohan
changed the title
ram-mohan
force-pushed
the
p3primaries
branch
from
9f25e09 to
dd56e5c
Compare
ram-mohan
force-pushed
the
p3primaries
branch
2 times, most recently
from
94ada70 to
c44bd39
Compare
- luminance function of p3 gamut is using weights of dci p3 instead of display p3. this is corrected. accordingly the gamut conversions involving display p3 are updated. - increased precision of some constants used in csc Test: ./ultrahdr_unit_test Change-Id: Iec77b3d24adb7ad887a8f53d805eecc0c0160f28
ram-mohan
force-pushed
the
p3primaries
branch
from
c44bd39 to
ee9e793
Compare
| return {{{1.34369f * e.r + -0.28223f * e.g + -0.06135f * e.b, | ||
| -0.06533f * e.r + 1.07580f * e.g + -0.01051f * e.b, | ||
| 0.00283f * e.r + -0.01957f * e.g + 1.01679f * e.b}}}; | ||
| return {{{1.343578f * e.r + -0.282179f * e.g + -0.061399f * e.b, |
There was a problem hiding this comment.
Could you double check if this matrix is correct?
I use
[2.4934969119,−0.9313836179,−0.4027107845; −0.8294889696,1.7626640603,0.0236246858; 0.0358458302,−0.0761723893,0.9568845240;] x [0.636958,0.144617,0.168881; 0.262700,0.677998,0.059302; 0.000000,0.028073,1.060985;]
and got this: (maybe I am wrong)
[ 1.33736487 -0.33736476 -7.08830856e-07 ] [ 0.02128283 0.97871719 -1.14552781e-07 ] [ 0.00418868 0.03418871 0.99581128 ]
There was a problem hiding this comment.
the matrix multiplication seems to be wrong. The individual matrices are same as the ones i am using.
There was a problem hiding this comment.
OK I re-did the math with gemini and this time the result matches with yours. My previous result with chatGPT was wrong. Sorry for the confusion.
| return {{{1.34369f * e.r + -0.28223f * e.g + -0.06135f * e.b, | ||
| -0.06533f * e.r + 1.07580f * e.g + -0.01051f * e.b, | ||
| 0.00283f * e.r + -0.01957f * e.g + 1.01679f * e.b}}}; | ||
| return {{{1.343578f * e.r + -0.282179f * e.g + -0.061399f * e.b, |
There was a problem hiding this comment.
OK I re-did the math with gemini and this time the result matches with yours. My previous result with chatGPT was wrong. Sorry for the confusion.
2 participants
Test: ./ultrahdr_unit_test