Is theree any way - with integrated renderers or third party renderers - to render out a 64-bit floating point depth buffer? If so, with what renderer and what output format?
64-bit per channel or total? If it’s the second, then yeah, just use 16-bit EXR, if it’s the first, then I don’t believe so.
dds rpf rla rgb can save as 64B image , exr channel map can save as 64B , but exr itself is 128B image
for render , it is float , even use scanline
What sense would it make to have 4 copies of 16-bit depth? of course one (or 4) 64-bit (double precision) floating point channels. The OpenEXR format does support 64-bit floating point - but how to render it? of course it only makes sense if the renderer computes everything in double precision too
Your wording was not specific, was it “(64-bit floating point depth) buffer” or “64-bit floating point (depth buffer)”. Not sure where you got that OpenEXR supports double precision, it’s not one of the supported formats according to the official introduction doc:
Regardless, the only way I can think of is calculating it with double variables and exporting it yourself.