I would also expect the 1060 to do better than 5 FPS at 100% utilization. So it has me wondering a little bit if there isn't a percentage cap written in? A guy with a i7-8700k shouldn't be pegging at 20% if a Threadripper 1950X is also pegging at 22%. And with a processor that mine would walk all over. Chances are I have a boatload more stuff running that isn't DR compared to OP. Just thought I'd offer my observations since I seen this post from a user who is right at 20-22% CPU like me. I haven't peeked at the current manual yet, maybe the answer is in it. I have already verified the GPU is set to CUDA mode. I used last March, I recall seeing my GPU nearly pegged during rendering. I'd be curious to see if GPU consumption holds steady around 54% no matter the GPU? In otherwords older GPU's at 54% are going to be a lot slower than the latest and greatest at 54%. I think my previous best was 224, so a definite improvement. (1920x1080 P 59.94 to 1280x720 P 59.94) When no color correction is required it shoots north of 320 with my current source material. When color correction is required, encoding to Cineform varies at 17-24.5. The software must have just gotten more efficient. I think my encoding and playback are all about the same as the last time I fired it up, if not maybe just a bit better. Maybe fusion uses more, I haven't played with that component much. I am sitting at 22% CPU and 54% GPU, you folks with 3090's must be crying to see how little is being used. It's comforting to know my frame rate would double if it did. Perhaps I have failed to understand exactly what it is you want to happen.I have noticed that DR isn't using all my resources either.
Of course not as good as if I had zoomed in with the camera, but pretty good enough, in that example, I may have been able to save the shot. Not ot long ago I tinkered with Super Scale, but zooming in on a 1080p clip from 1.0 to 13.0, then applied the Super Scale 4x, with sharpening and noise reduction set the medium. Whether the clip is acceptable will depend on how much Super Scale you have applied, how good the original clip was how much sharpening and noise reduction was used as well as other corrections you may have made, etc. Or, I suppose you could use the Super Scale to improve the quality of a low quality clip and keep the resolution the same for the render. Then the rendered file will reflect the appropriate attributes depending on how you change the resolution. What you will have is an uprezed clip, which you then can render out to a higher resolution and the clip may retain more quality as compared to upscaling. The clip attributes are going to stay the same, as is the project settings. It is not going to change the attributes of the clip. Perhaps I’m incorrect, but using the Super Scale is like applying an FX.
0 kommentar(er)
