To be able to determine before render where there is nothing

V-Ray V-Ray for 3ds Max Requests
1

A solution to the problem of the long time to render where there is nothing (blank background)
a problem which is posed here : http://www.chaosgroup.com/forum/phpBB2/viewtopic.php?t=15643

good :idea: I think.

2

as posted in the other thread that probably wont work. If you request vray to take 4samples min, then vray will also have to do the visibility test with 4 samples. Otherwise, you face the problem of possibly non-matching alphas. An option would be to add a user-settable border value (like sample visibility with undersampling and add 5 pixels around it)…BUT the visibility check would also have to take effects like displacement into account. If displacement has to be sampled the visibility chack will be slow. And if you undersample the displacement, you have to REsample it when doing the final render…not sure if it would get any advantage…vlado ?

Thorsten

3

I can want a minimun of 4 sample on a simple object for a master key or an incrustation without wanting that Vray puts as many time for antialiasing where there is nothing. If it’s really not possible (and I think that nothing is it) then a simple “region selected” which work with animation will be enough. :roll:

4

That could be scripted i assume…if the region needs to move you´d have to use custom parameters linked to some dummies and tapes to visualise it in the viewport, animate the dummies as needed. Reverse Raycast them (or unproject?) them to get screenspace, and set the render region accordingly. then call the renderer. That´s quite some overhead i assume. But i´m pretty sure a similar effect could be done with finetuning QMC. I think lower min settings, but tweaking the sampler to be biased towards higher sampling. should keep quality while reducing empty background rendering. But i´m not a qmc wizzard so there´s prolly better ppl to give advice on that topic.

Thorsten

5

instinct, such thing has been attempted before. And it is possible, but there is a limitation in max. Such that its not possible to overcome with out hard coding. The limitation is that its not possible to clear the frame buffer from previous region to the next, resulting in image having multiple regions by the end of the animation.