Those numbers will change as the scene gets more complex, that test represented a relatively small uncomplicated scene that only took a few minutes to render. A more realistic test would be an actual production quality scene with lots of maps rendering at higher settings and resolutions.
Announcement
Collapse
No announcement yet.
How to estimate size/benefit ratio for DR renderfarm
Collapse
X
-
if you mostly queue up a series of images like me, I find that not using DR is more time efficient and you have no downtime while things render out.
some people like the 20 minute test renders though i guess so they can browse youtube or play games etc :P. Dont get me wrong DR is great but i think its used by some people as an excuse to have lots of breaks throughout the day.
Queue the images up, let them all render on your renderfarm and while they are rendering on your farm via backburner, you can start on the next project . Thats what im getting at (no downtime). If every job you do requires 4-5 images then 4-5 rendering machines would be perfect. Sorry, I know this isnt what you were asking but from someone on the other side of the DR fence i thought it worth offering my perspective
Comment
-
@paulison: not at all - that's a very helpful perspective. I have never used BB before and am just now starting to really think about how it can free up machines, and the whole pipeline. I'm one of those people that have the render downtime, but usually would rather not, so will give it some serious thought and maybe finally give BB a try after all.
Anyway, having 4-5 machines would allow me to get the best of both worlds either way. At this point I am going to move ahead with 4 new machines as that is a pretty safe bet and when I get a chance I'll work out a test on a larger setup and see if things scale as expected.
@Devin: I agree it's worth testing to find out, but right now I'm going almost around the clock and just don't have time to sort out some test scenes.
Thanks again guys - lots of great help.
b
Comment
Comment