We definitely should, although probably from a completely different perspective than ‘94: High end phones can capture in 3D, so we should be able to store and view 3D images properly.
I’m starting with the idea of PNGs where the channels represent RGBZ instead of RGBA, but obviously we need more than 8 bits of depth information, and we are starting to get and care about high dynamic range for RGB data as well… would 16R:16G:16B:16Z work?
There’s also the point that depth information will have sharp boundaries, and it will probably be much worse for RGB information to blur over such a boundary as compared with the equivalent parts of a flat JPEG.
I’m starting with the idea of PNGs where the channels represent RGBZ instead of RGBA, but obviously we need more than 8 bits of depth information, and we are starting to get and care about high dynamic range for RGB data as well… would 16R:16G:16B:16Z work?
There’s also the point that depth information will have sharp boundaries, and it will probably be much worse for RGB information to blur over such a boundary as compared with the equivalent parts of a flat JPEG.