Forcing Vegas to only use NVidia gpu

Comments

RealityStudio wrote on 12/4/2021, 6:49 PM

Ah that's a shame. I mean truth be told this hasn't been fixed in ages so realistically I think it will remain this way at least until they are forced to port Vegas to an alternate architecture like Arm.

I also just did the test with Resolve on my little M1 based Mac Mini, it was quicker than all the others and resultant output file size was around the same but the video quality was substantially worse. No idea what's going on there but I'd have to look into it more. It does make me wonder if all these Youtubers that love to compare render speed between M1 Macs and PC's ever actually look at the final product to see if it's comparable or if they just post numbers.

RogerS wrote on 12/4/2021, 8:24 PM

Thanks for the 3 way test, very illuminating. I hope Vegas can catch up with Resolve for GPU support.

Former user wrote on 12/4/2021, 11:10 PM

The Vegas team has a better relationship with Intel, so while Nvdia performance is better, it's not necessarily the best for Vegas (like AVC decoding).  

Who recalls this benchmark from Intel's site when trying to sell the very poorly received 11900K?

 

No other details are given, but we can guess 100% that a HEVC 422 10BIT file was used as the source file, and that's probably enough to make that somewhat unbelievable result true (8core cpu beating 10core). I mean it is true, but it's not representative of editing and encoding in general. It is a misleading benchmark.

Intel will be releasing their Intel GPU's next year, who's willing to bet they will use Vegas again, compare NVENC encoding via QuickSync encoding. Intel will win due to Nvidia hardware encoding having the problem it does as seen on the encoding graph above

Intel would be mighty unhappy if Vegas fixed their NVENC encoder before then..... so would they dare?

Howard-Vigorita wrote on 12/5/2021, 7:53 AM
I also just did the test with Resolve on my little M1 based Mac Mini, it was quicker than all the others and resultant output file size was around the same but the video quality was substantially worse.

@RealityStudio That's odd... same software and output file size similar but worse quality. MediaInfo might explain.

If you zip the clip and lut for download, be happy to compare runtimes.

RealityStudio wrote on 12/5/2021, 1:16 PM

I film adult industry content so definitely not a shareable video, but anyways having said that my Resolve issues are beyond the scope of this forum. I need to find a Resolve forum and ask there because I really can't figure out why the Mac M1 version of the same render, basically a 7 minute 4k h265 file around 600MB looks bad whereas the same one encoded with the same settings on the same app but with NVidia encoder on PC looks really good. I don't know Mac enough to sort this out so I'll look around.

Former user wrote on 12/5/2021, 7:24 PM

When reading reviews about M1 I recall M1 Resolve has a box in the render section that said something like 'Optimize for speed' that would sound like something you would want turn off, but otherwise you should find the resolve help forum. And read about what a buggy mess the last 2 releases of Resolve have been, so could even be a bug

RealityStudio wrote on 12/5/2021, 9:08 PM

When reading reviews about M1 I recall M1 Resolve has a box in the render section that said something like 'Optimize for speed' that would sound like something you would want turn off, but otherwise you should find the resolve help forum. And read about what a buggy mess the last 2 releases of Resolve have been, so could even be a bug

I tried unchecking the "Optimize for speed" box then re-rendered. It then actually look longer to render than Resolve on my PC laptop and the quality was still bad. I wonder if I have to install a separate encoder or something like that, no idea.