Hello
Here are 2 animations I did to test VRay Next update 2.0
the first one is regular Archviz scene like what I deal with daily .
All the technical details are in the video description, but in short I did use the default settings and Next GPU for this one .
I’ve noticed a flickering in the reflection so I re renderd it is with light tree options and that -or lower noise threshold - seems to fixed it
the second one was fun project inspired by Beeple .
for this one I used Next (CPU ) as I counter a bug that crashed the rendering (already reported)when rendering with Next GPU
I usually post these in my sketchbook, but I though creating new thread isn’t bad idea to maybe start a discussion : ))
That’s what I had for a recent animation, but to get something clean in Corona, it would have taken days (10 minutes per frame). Yours took, what, 20 hours? That still seems incredibly long. I have been unable to convert my Corona scene back to V-Ray to render on my Titan RTX so I am stuck. I might have to use a render farm because I have 4 rooms to animate.
Yes it is almost 20 hours ~ 3 night of rendering..I didn’t care that much about the time asI like to do those tests to keep my pc warm at night lool
My machine is old now, so I think with your hardware you can get the time to half at very least.
Rendering animation is always power demanding. In the studio that I used to work for, I had almost 400 cores to use for animation and it is always a race with time to meet the deadline.
We, in few occasions, had to use render farms out side, but we couldn’t use any of the online services (slow internet, unsopprted features.. Etc) so we used another studio computers for that.
I always ask for more time than what I excpect the render will take as any mistake can cost you lot of time and /or money.
I am not sure how you are doing it, but my kitchen, using GPU is a noisy mess. WIth Corona, 5 minutes at 1920 gives me a clean image. V-Ray NEXT GPU at 1080 is noisy and I can’t use it at 5 minutes.
Yes, that’s right .I’ve used the Standalone here , For 2 reasons :
1-It can load the sequence and read few frames before and after the current frame and then do the de-noising which will give better results from frame to another
2- I can batch all the cameras then try different values for each one if needed (The DOF shot for example might need more aggressive de-noising )- I might play few frames in aftereffect to see if I like the denoised results or not then go back and change things without the need for re-rendeirng .
You can read more about de-noising here : V-Ray Next: Denoising in Production | Chaos.
Regarding render time .. in my case it was direct light mainly with bright and simple materials ..but also you might have some issues in your scene that slow things down ..if you suspect some, I could take a look for you if you like or you can send it to support to check .
