Hi
I'm playing around with PLE and am amazed.
However, I was wondering how the render bucket sizes are calculated and if they can be controlled, like in some other renderers.
If I'm rendering an image with a small area that has really demanding shading (to reduce noise), Clarisse might assign just one bucket to that area and can spend 15-30 minutes chewing on it with just one core, while my other 35 cores on my Dual Xeon workstation are doing absolutely nothing, because they have finished the rest of the image.
Then also, each new refinement in display quality cannot be started until the previous has finished, again causing several periods during the render, where the CPU is doing almost nothing.
I use the setting of ''Always Compute Full Quality Image' to avoid interruption in the render.
Clarisse can easily render a difficult small area with all cores, when using region render, so it is technically possible.
Overall, due to the sequential nature of the bucket rendering, unrelated parts are simply waiting for each other to finish, and it can dramatically increase render time.
Will Angie fix any of this?