Your computer is not good enough for you are trying to do. You need a combination of both high end CPU/GPU in order to render or working with 32 bit project settings. I can edit and render 32 bit project, no problem here. In 8 bit settings 4K footage on the TL I am getting 51 frames. In 32 bit project settings, I am getting 22+ frames in rendering. Working with 32 bit project settings. This settings is not for average Joe's computer. Its required a very powerful computer to that.

@ALO I would stay in 8 bit, why are you trying to use 32 bit?

You have to push on GP Protune a lot to decode it to an attractive image. 32 bit math helps reduce the banding/quantizing in (especially) blue sky gradients.

Bruce, I haven't noticed this problem before using less powerful computers. Sure, preview frame rate drops, but it was still useable. Suspicious also that I'm able to use 32 full mode with this computer. If it couldn't handle 32 video, it wouldn't be able to handle 32 full either, no?

@ALO what model GoPro are you using? Are you shooting in 4K, 2K, or FHD?

@ALO when I preview Best/Full GoPro 2k in 8 bit and 32 bit floating point (video levels) I get full speed (29.970 fps). If I switch to 32 bit floating point (full range) i get about 6 fps preview at Best/Full.

In 32 bit floating point (full range) have you tried changing the 'view transform' to 'Rec. 2020 (ACES), does that make an improvement?

Hmm...if you're able to to work in 32/video mode, that suggests this isn't ubiquitous.

I've been using the Hero 7 Black's 2k/Linear mode for non-POV shots. Really like the combination of wide angle with low (ish) distortion in such a compact package.

I only have access to a Hero 3, which if I remember correctly does Protone in 'CIN' mode. But I'm sure there have been many improvements through the models.

BTW: I'm not getting any banding, but that doesn't do you any good being on a different camera.

@ALO BTW, your Dell does have a NVIDIA GeForce GTX 1050, correct? How much vRam and are you on the latest driver?

32 bit has no effect at all on banding if your source video is not 10 bit or higher, same for your target output format.

OK thanks for the help -- looks like this is either a corrupted project file or a bug. I can't replicate it in a new project, so for now I'm going with corrupted file. In any case, I recreated the same GP stream with the same titles/generated media tracks combination in a new project. No issues going from 8 to 32 bit video modes. My system effortlessly handles either. Go figure. :)

32 bit has no effect at all on banding if your source video is not 10 bit or higher, same for your target output format.

That is simply incorrect*, and I strongly advise you to stop repeating it. Someone might actually believe you.

*if you are grading and/or otherwise adding post-processing to 8-bit source footage

This has been debated and debunked in this forum for many years. There's threads and threads covering it, many of which have benchmark tests disproving any benefit from using 32 bit mode on 8 bit source video. I suggest you read some of them.

@ALO

Those are strong words. I think you should back up your position up with some examples. I would love to see someone (anyone?) demonstrate the advantage of processing 8 bit camera footage destined for 8 bit output, using 32 bit mode. But I need to see it, not read about it.

I have asked numerous times, but nobody ever presents the evidence.

https://www.vegascreativesoftware.info/us/forum/plea-to-new-users-please-do-not-set-32-bit-pixel-format--114750/

This has been debated and debunked in this forum for many years. There's threads and threads covering it, many of which have benchmark tests disproving any benefit from using 32 bit mode on 8 bit source video. I suggest you read some of them.

There is a wealth of information available on the internet that will explain, in exhaustive detail, the differences between 8 and 16-bit math, and how that is relevant to post-processing in photography and videography.

You can also simply run tests yourself on 8-bit footage. Put a color curves on a clip and crank up the contrast. Look for areas with subtle gradients, like skies. You'll see a clear difference between 8 and 32-bit. You don't even have to render out. Just switching between modes will reveal the difference.

It is unfortunate that some people feel the need to argue otherwise.

Funny how the response is always to refer to "other" sources for explanation or tell us to run our own "obvious" tests.

If it's obvious, then come on ... post your samples. I really don't want to argue, I just want you to produce the evidence to back up your claim. Can you do that?

;?)

I'm using 32-bit pixel format (video levels) and do not know any issues at this moment with VP15 related to it (used to be an issue with SonyAVC encoder in earlier VP15 builds).

Here is a comparison of gopro protune footage 8 bit source to 8 bit jpeg. Applied a ridiculous curve to induce banding; original sky was nice gradient. One processed in 32 bit mode, but they look equally sh***y to me. :)

ALO, please post the information specified in part B1, C1 & C2 of this post: https://www.vegascreativesoftware.info/us/forum/important-information-required-to-help-you--110457/ This is very important in diagnosing your problems.

Here is a comparison of gopro protune footage 8 bit source to 8 bit jpeg. Applied a ridiculous curve to induce banding; original sky was nice gradient. One processed in 32 bit mode, but they look equally sh***y to me. :)

Kinvermark I ran a few tests to try to come up with a quick, definitive result, and what I find is it depends a lot on the scene, and even then, visually it's a subtle improvement--generally a smoother, creamier sky is what I notice. Or no difference at all. It all depends on the source and the grade.

However...you can easily verify that the math is giving two different outcomes, always, whether or not you can see the difference, and you can do this right in Vegas. Put some 8-bit footage on the timeline. Switch your scopes to histogram view (not the parade, the one-channel histogram). You should see a relatively smooth curve from light to dark. Switch to 32-bit mode. You should see absolutely no difference. Switch back to 8-bit mode.

Now: put a levels Fx on your clip, and start raising the contrast. The histogram curve will start to look jaggedy as the data is quantized by 8-bit's inadequate headroom. This is not subtle--you'll see it right away. Finally, now switch to 32-bit processing. The histogram snaps back to smooth.

Reasonable people can certainly debate whether or not any of this matters. The math yields different results because it *has* to. Whether or not that translates into a visible difference is highly dependent on the scene, the quality of the capture, and the editing choices. And even then...it's probably going to be subtle. But it is real.

ALO, please post the information specified in part B1, C1 & C2 of this post: https://www.vegascreativesoftware.info/us/forum/important-information-required-to-help-you--110457/ This is very important in diagnosing your problems.

Nick I can't replicate this issue reliably, so I'm not going to escalate it. If I can figure out a way to make it happen on command, I'll follow up. Thanks!

Kinvermark I ran a few tests to try to come up with a quick, definitive result

https://www.vegascreativesoftware.info/us/forum/this-is-not-about-grading--114804/

https://www.vegascreativesoftware.info/us/forum/10-bit-vs-8-bit-grading-the-musical--111748/