I think what Romano mentioned is close to what happens indeed.
Just FYI... by default - I use "standard color matrix" in input profile
and gamut clipping is "off". This produced quite strong blue color (but
it so happen that it works for me - I like it). I expected when CBS
module is on - it would continue pushing the colors in certain direction
(but only if I change values).
Then when johannes hanika mentioned
as soon as you enable any module that makes reasonable assumptions
about Lab (like, L >= 0), you'd clip the values to the gamut of Lab,
I realized that most likely this is affecting the results - the
assumption by the module.
Further I tested with what Tobias mentioned
Try enabling the
gamut clipping in the input color profile (2nd dropdown there). Set it to
"linear Rec2020 RGB" for a start.
and it brought the color to be on the bluish side. Not only that - it
also resulted in turning on CBS module with 0,0,0 to not affect the
image (so it worked as I expected).
Then I changed the clipping from Rec2020 to sRGB and the result was
purple (but again - CBS would behave well - it would not change the
image when with 0,0,0)
And this leads me back to what Johannes pointed - that some modules make
reasonable assumption.
I guess - for a better control - I would have to enforce gamut clipping
on the input color profile module. The thing is - it can differ from one
image to another. So - likely I will have to pay more attention to it
(and to the out of gamut colors).
My final image is with the suggestion from Tobias.
Regards,
B
On 2017-05-18 01:20 PM, Patrick Rudin wrote:
johannes hanika wrote:
these out-of-gamut/imaginary stimuli result in negative L values in
the pipeline.
Could you explain a bit further how exactly this negative L-value is
created? I always thought the camera-specific profile is a kind of LUT,
but obviously it is a bit more complicated :)
I know that all extreme values in a or b-channel are out of gamut in
sRGB, but how can a clipping in the blue camera-channel lead to a
negative L-value?
Regards
Patrick
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to [email protected]
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to [email protected]
