Vray RT darker than production

[timmatron]

This is driving me nuts! Can someone tell me why I'm getting a much different brightness or gamma with RT vs production render? If I switch to BF/BF in production mode it seems to be a much closer match to RT, but still a little brighter. I switched to linear multiply (from reinhard) and set both bright and dark at 1.0 just to make sure that isn't a problem. I'm clearly missing something fundamental here, hopefully easy. Please help! :/

Vray 3.50.04
Sub pixel mapping and clamp output is off

[Alex_M]

This might be due to low GI depth in RT Active Shade. Increase it from the default 3 to something like 10 and see if it's closer to the Production rendering. BEWARE that this might decrease the render speed and interactivity of RT because Vray will compute a lot more Brute Force secondary light bounces.

By the way, may I ask why you use RT CPU? If you want interactive rendering that is closer to the production result why not use IPR? RT CPU is generally slower than IPR. If you insist on using RT then you really want to be using CUDA which is many many times faster than RT CPU and even IPR if you have a fast graphics card/cards like 1080ti, 980ti 1070 etc.

[timmatron]

This might be due to low GI depth in RT Active Shade. Increase it from the default 3 to something like 10...

Good suggestion, thanks Alex. I'll try it first thing in the morning. As far as why I use CPU, uhm, I think it stemmed from a time when GPU RT had more limitations than CPU. Perhaps they are equals now, especially in 3.6 which I look forward to trying. I have a 1080 now so I should probably give it a shot again. The rest of my computers don't have good cards though so maybe CPU RT with DR is still faster for me, not sure. I only use it for low res preview renders but maybe it's time to expand my usage of it. But first things first, I need to get to the bottom of this huge disparity between the two render methods! Hoping you are right about GI depth.

[timmatron]

Hey Alex, I think you hit the nail on the head! Increasing GI depth to 10 gave me a much more similar result between active shade and production. GPU RT does quite well too in that test scene, except for one difference in a material that had vrayedges tex in the bump slot. But probably good enough for testing purposes. Thank you!