What is the difference in having a tile have sub-regions versus splitting that tile up? Is one better than the other?
Also, does anyone have any tips when using jigsaw single-frame rendering and assembling over a completed image and how to best deal with render elements? Right now, I can only get it to assemble correctly over a chosen diffuse pass.
Thanks!
From Deadline’s point of view, there is no difference. Each sub-region, or full tile, ends up being rendered as a region.
The reason for the sub-tiles existence is user friendliness - if you define a region, and you want to split it MxN times, but the size of the whole might need adjustment after the fact, having a region procedurally split into sub-tiles lets you adjust just the parent, and all sub-regions will be updated dynamically.
If the render elements are output as layers into the same EXR file as opposed to individual files, the Draft Tile Assembler should handle that and retain all render passes in the file when assembling.
With separate Render Element files, there is currently no way to do that. I will have to think about it…