Improvements to CIECAM02.

[nschloe]

I recently implemented CIECAM02 (and CAM16) and found some improvements. colour could also profit from it. Check out https://arxiv.org/pdf/1802.06067.pdf?fname=cm&font=TypeI, in particular page 3, or its implementation in colorio https://github.com/nschloe/colorio/blob/master/colorio/ciecam02.py.

[KelSolaar]

That looks great @nschloe! I think it would be worth contacting the authors about it. We are usually not integrating anything that is not coming from official literature though but will definitely keep an eye on it.

[nschloe]

Yeah well I am the author. Just a quick write-up right there.

[KelSolaar]

Oh I was referring to the CAM16 authors :)

[nschloe]

Ah yes. Okay, let's try to get officially endorsed. :)

[KelSolaar]

That would be great!

[KelSolaar]

@nschloe : Any news on this front?

[nschloe]

Uh... What exactly do you have in mind? I wrote the authors and got a replay back from Changjun Li (in Feb 2018) saying that they'll have a look. That's fair enough, right? It doesn't really matter if the authors look at it, right? The note is more for anyone who wants to implement CIECAM02/CAM16.

[KelSolaar]

Fair enough! Your last comment was:

Ah yes. Okay, let's try to get officially endorsed. :)

So I was just inquiring if there was any progress.

[nschloe]

Yeah, I mean, it'd be nice if this was used by upstream since the standard is a bit flawed here and there, but I cannot make anyone use it. At least the ideas are out there for everyone to implement.

And besides, I'm working on a color model that will perhaps obsolete all of CAM ones anyway. 😄

coloria-dev/coloria@eedda66

[KelSolaar]

Ah! :)

Thanks, I will close this one to tidy up our stack a bit!

[nschloe]

I wouldn't close this before having made the respective changes, e.g., here.