search

meshula / Nanocolor

Nanocolor, a lightweight color transformation library
Apache License 2.0
61 stars 5 forks source link

Plain Text Specification? #1

Closed sobotka closed 8 months ago

sobotka commented 8 months ago

Apologies for logging this as an “issue” and not a GitHub discussion, but it seems Discussions aren’t yet available on this repository.

This is terrific work folks.

Is there any chance of a text based specification being developed?

I had authored something that is awfully close to what seems to be guiding ideas for NanoColor back when OpenTimelineIO was pondering colourimetric based ideas.

Not sure it is of use, but I am curious if there’s any plans here?

meshula commented 8 months ago

Hi Troy, since you pointed it out, I've activated discussions. Note that there is also a #nanocolor channel in the OCIO Slack (I just invited you to the channel). Your OTIO proposal does indeed it look very similar in spirit. I think what you're proposing there encompasses what nanocolor is meant to solve, but may go further into a larger OCIO-lite project.

I think a text based specification is very much in line with the project's intent.

sobotka commented 8 months ago

Thanks Nick. I posted it because it is an extension of a hack patch I did up a long time ago for metadata within OpenColorIO forever.

Getting a text specification “right” seems crucially important. It seems important to be able to accommodate the entities at a granular level (EG: An EOTF, OOTF, OETF, primaries, achromatic centroid, etc.) while also supplying a low-labour descriptor format for higher order constructs.

I am curious your thoughts on the ability to derive “origin” encodings in such a discussion? EG: Being able to get to the “lowest” post FPGA quantal catch array from some specific colourimetric specification?

You can convert an Issue to the Discussion, as an aside. Might be good to port this one over to the Discussion.