The part of jpeg-xl I'm most excited for is when the 3d community starts standardizing the extra channels for depth maps, bump maps, roughness etc in the extra channels. You can use a single jpeg-xl as a full material system with progressive decoding for LOD and all the rest.
Does jxl support specifying what each channel contains rather than just having a channel number with a convention? I guess you could always add it as additional metadata if such a field doesn't exist already.
