Tree optimization

[joconnell]

Vlado, just curious about this method. It's very very convenient but with vray doing it's sampling optimisation based on how much something contributes to the final pixels, some of the passes can be a tiny bit noisy. The compers in here are using the render elements to come up with a look so they're pushing and pulling things beyond the levels that vray is sampling for so what's the handiest way to force vray to oversample and get really clean raw passes? At the minute I'm trying to get something that's reliant on AA sampling since there'll be a lot of 3d motion blur through the job!

[vlado]

Well, one way is to put more min. subdivs to AA, but of course this will put more samples on everything. For future builds we could add the "consider for AA" option to some of the default render elements perhaps with a spinner on how much the element should affect the AA.

Best regards,
Vlado

[joconnell]

Yeah it's a tough one - I'm kind of asking vray to totally overwork everything to allow fiddling afterwards which kind of negates importance sampling. Bloody compositors

[glorybound]

I am having a hard time cleaning up noise using the universal method. I have the clr threshold down to .0003, and there is still way to much noise.

[RockinAkin]

Is the minimum shading rate you refer to the same as the "min samples" under DMC Sampler?

Just to clarify: Min Samples is NOT the same as Min Shading Rate. The two behave very differently.

Min Samples should be thought of as 'Min Adaptive Samples' and is restricted by the DMC sampler.
Min Shading Rate is available in 3.0 and is MUCH more powerful since it's not restricted by the DMC sampler and will actually raise the maximum number of secondary samples.

Refer to the DMC Calculator to see for yourself.