Uh, I don't think adaptive bucket splitting is doing its job. I appreciate the effort, but there needs to be an option to manually set the bucket size. I'm in a situation where CPU usage keeps going down, super inefficient. All of the complexity in the image is focused in about 10-20% of image area. It runs for a minute or two, renders all the easy bits. Then hits a wall. I have 16 cores but the image is pretty small, less than 1K. Those last 8 to 10 buckets are taking forever, seriously it's no fun to see the CPU usage go down, and down, and down. Can't use GPU because features aren't supported. And Progressive CPU is even less efficient, it's only using 35% of CPU time. Huh?
If I had the option to reduce the bucket size I could reduce the inefficiency. This is dumb. Please bring back a manual bucket size. New features are great until they're not. Old features shouldn't be removed until we're 100% certain the new functionality is better than the old.
If I had the option to reduce the bucket size I could reduce the inefficiency. This is dumb. Please bring back a manual bucket size. New features are great until they're not. Old features shouldn't be removed until we're 100% certain the new functionality is better than the old.
Comment