I'm trying to do a batch of low res test renders. I have a few pre-render scripts that are run to disconnect shaders, etc. to improve the performance.
Each file takes approximately 6 minutes to open. I'd like this to only happen once per CPU I dedicate to the job. That is, I'd like if I assigned 1 CPU to a job, it would open the file once, run my pre-render scripts, render the first frame in the range, then keep the file open and render the next frame and so on. As it stands, Qube wants to re-open the file for each frame, incurring a penalty of 6 extra minutes on top of a render time that's only about 30 seconds for that frame.
So far, the workaround I'm using is to simply avoid using the Qube --range flag, and instead hard code the ranges into the -s and -e maya flag in the maya Render command. The downside is I get no feedback on my jobs except by watching the StdOut window.
