I know that you can save scanline exr’s out of max directly in current versions of the software but I am under the belief that tiled exr’s are more effecient for rendering in terms of memory usage.
I’d still like to save exr’s directly out of the vray framebuffer and post process them instead of using the vrimg format just because it seems kind of slow. I know there is a method of doing it with a python script and nuke but I don’t have nuke available to me.
It would be great if the vrimg2exr utility would allow you to input exr’s to do this or adding some other simple utility that could be run as a post render script.
V-Ray 2.0 adds compression to .vrimg files so they are comparable to OpenEXR files in size (for 32-bit OpenEXR files, as .vrimg files are always full floating-point).
When vray renders things it does it in buckets and saves the output of each bucket as a tile within the image. A lot of compositing programs render their image from the top down like the scanline renderer (hence the two ways of saving exr are called tile and scanline) so if you give it a tiled exr to work with, it doesn’t fit quite as well as a scanline one. It won’t change quality or your image in any way, it just makes the comping program render faster.