About using CAM16-UCS as working space for the picker.

[KelSolaar]

Hi @OlivierVicario,

Have you thought about using CAM16-UCS colourspace for your picker? It should have better uniformity than CIE Lab.

Cheers,

Thomas

[OlivierVicario]

Hi Thomas,

I had not heard about developments after CIELab when I did this program. I've been talking about it since then and I'm looking for an implementation in java because I do not practice Python.
A contact says to me that the system that seems at the moment the most interesting for designers is the Jazbz of Ming Ronier Luo (AIC 2017 medal for color research, Director of Division 1 at CIE) which has just come out, see the equations here : https://www.osapublishing.org/oe/abstract.cfm?URI=oe-25-13-15131#articleEquations .
There is an Python implementation at https://github.com/nschloe/colorio .
I have tryed to make a java implementation from the article equations but it does not yet work ! : /

Thanks for your comment. I looked at your site and I found your characters super ! ;-)

Cheers,

Olivier

[KelSolaar]

Thanks for the kind words!

We have an implementation of CAM16 here: https://github.com/colour-science/colour/blob/develop/colour/models/cam16_ucs.py

Jazbz is planned on our end but depending the dataset you use to measure performance, ICtCp from Dolby might work better:
https://www.dolby.com/us/en/technologies/dolby-vision/ICtCp-white-paper.pdf It has the same goals anyway.

Cheers,

Thomas

[OlivierVicario]

Thanks for the link. ICtCp looks great ! Cheers, Olivier 2018-02-20 9:16 GMT+01:00 Thomas Mansencal <[email protected]>:

…

Thanks for the kind words! We have an implementation of CAM16 here: https://github.com/colour- science/colour/blob/develop/colour/models/cam16_ucs.py *Jazbz* is planned on our end <colour-science/colour#378> but depending the dataset you use to measure performance, ICtCp from Dolby might work better: https://www.dolby.com/us/en/technologies/dolby-vision/ ICtCp-white-paper.pdf It has the same goals anyway. Cheers, Thomas — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#1 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/APgwNTi6SxOKd4shfuNBSJSmUL0FuEWOks5tWn9CgaJpZM4SLdlQ> .

[OlivierVicario]

Hi Thomas, I tried to write an implementation in java of ICtCp but I have difficulties to go from LMS to L'M'S '. Could you send me this step in an explicit algebraic form please ? That would be very nice of you. I am more of an artist than a scientist : / Cheers, Olivier Le mar. 20 févr. 2018 à 17:51, olivier vicario <[email protected]> a écrit :

…

Thanks for the link. ICtCp looks great ! Cheers, Olivier 2018-02-20 9:16 GMT+01:00 Thomas Mansencal ***@***.***>: > Thanks for the kind words! > > We have an implementation of CAM16 here: > https://github.com/colour-science/colour/blob/develop/colour/models/cam16_ucs.py > > *Jazbz* is planned on our end > <colour-science/colour#378> but depending the > dataset you use to measure performance, ICtCp from Dolby might work better: > > https://www.dolby.com/us/en/technologies/dolby-vision/ICtCp-white-paper.pdf > It has the same goals anyway. > > Cheers, > > Thomas > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <#1 (comment)>, > or mute the thread > <https://github.com/notifications/unsubscribe-auth/APgwNTi6SxOKd4shfuNBSJSmUL0FuEWOks5tWn9CgaJpZM4SLdlQ> > . >

[KelSolaar]

Oh the conversion from LMS to L'M'S' is just about encoding using the Dolby Perceptual Quantizer: https://github.com/colour-science/colour/blob/develop/colour/models/rgb/transfer_functions/st_2084.py#L59

Encoding

Decoding

[OlivierVicario]

Thank you for the answer. I understand that Color value abbreviated as :math:`N`, normalized to the range [0, 1], but what do we do after to transform from LMS to L'M'S' ? Is it : L' = N * L M' = N * M S' = n * S ? 2018-02-27 20:43 GMT+01:00 Thomas Mansencal <[email protected]>:

…

Oh the conversion from *LMS* to *L'M'S'* is just about encoding using the Dolby Perceptual Quantizer: https://github.com/colour- science/colour/blob/develop/colour/models/rgb/transfer_ functions/st_2084.py#L59 *Encoding* [image: image] <https://user-images.githubusercontent.com/99779/36750668-018a714e-1c63-11e8-8fed-f4edf0f88626.png> *Decoding* [image: image] <https://user-images.githubusercontent.com/99779/36750767-383aae02-1c63-11e8-91c3-0f0d4b81c32c.png> — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#1 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/APgwNfveOtRgurEIipijc7WM71ID30g4ks5tZFq3gaJpZM4SLdlQ> .

[KelSolaar]

Basically you need to implement the equation from the first image for encoding and pass each separate L, M, S channel through it. They are the F_D variable in the image.

[OlivierVicario]

Okay. It's clear. Thank you very much ! Have a good day. Olivier 2018-02-28 9:58 GMT+01:00 Thomas Mansencal <[email protected]>:

…

Basically you need to implement the equation from the first image for encoding and pass each separate *L*, *M*, *S* channel through it. They are the *F_D* variable in the image. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#1 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/APgwNX5xxBaCmL3cIyH_wI0-hoTvzzoDks5tZRUhgaJpZM4SLdlQ> .

[KelSolaar]

You are welcome!