Hey guys, just wondering if this is normal behavior:
Since Nvidia Denoiser implementation in Corona, I hardly use it because to me Corona Denoiser is exactly the same speed on my dual Xeon 2696 v3 against my GTX 1070 Ti. My question is: is this expected? I seem to remember in V-ray the difference between Vray denoiser and Nvidia denoiser with the same GTX 1070Ti was more evident meaning Nvidia denoiser was faster than V-ray denoiser. Is Corona denoiser more fine-tuned speed-wise than Vray denoiser?
That sounds more like issue with how you use it, are you sure you have it on ? It’s confusingly currently on two separate system tabs. On all our cards in office (1080ti’s,1070, Titan Maxwell), it takes miliseconds to clear up to 3k image easily while in IR. It’s snap of a finger, except 10 times faster than I can snap fingers.
Definitely does sound like it is not the NVIDIA denoiser being used as Juraj says There is the checkbox in the Render Settings (Performance tab) which enables it during IR, then there is the dropdown list of Denoisers to use which is for final renders (which means it is entirely possible to have NVIDIA denoising during IR, but Corona denoising during final renders - which is why the two separate controls; just having NVIDIA checked in the Performance tab won’t use it for final renders). BTW just how long is the “same time”?
I should correct myself saying that both denoisers clear in less than a second also. But, again, they’re both equally fast with no noticeable difference in feedback speed on IR as far as I can tell. Just wanted to know if this is to be expected when using dual CPU workstations. I suppose that Nvidia denoiser is more useful along with a fast CPU with excellent single core performance?
Exactly, it doesn’t. You can have two independent denoisers set, one for IR and one for regular rendering. But for IR, only Optix is available. For IR, it doesn’t matter what the dropdown selection is set at.
I guess this is where the misunderstanding came from.