Byte bit depth support

[elaye]

Hello, I'm interested in using the FFmpegWriter plugin with a byte bit depth but discovered that it wasn't supported. By digging in the code I saw that it seemed to be because of colorspace conversion with OpenColorIO. Is it a limitation of OpenColorIO or a use case you didn't want to support?

As a proof of concept I forked the repo added flags to disable OCIO and replaced the bit depth with bytes instead of floats everywhere. Here is the relevant commit: elaye/openfx-io@1e12925329e7839f2195766572195430fb2e18cb Ideally for my use case openfx-io could support both bit depths. What would be your preferred approach to do that?

I'm happy to work on this but it would be good to have your direction.

[MrKepzie]

Hi, we actually never bothered to do something else than float because in Natron everything is 32bit floating point: we don't have any use case to have the Reader output directly a 8-bit image. If you remove the OCIO colourspace conversion as well as the premult/unpremult that may occur in some situations, you can then quiet easily support 8-bit rendering.