For anyone that's wondering, it's not like my footage is trash, overall it looks really nice, it's just that it had a lot of color changing backgrounds with stage lighting on a live performance (which of course means that I didn't have any control over lighting conditions and had to just deal and use high iso, as shooting wide open was not an option due to the need to capture a large area and have deep depth of field) so some things look totally fine, but some of the backdrop colors, especially if they're a little darker, get noticeably noisier than others.

Is this noticeable noise present in the footage or only after final rendering?

@3POINT

The original footage is in VLog on a S5iiX (full-frame) at sorta high ISO (couldn't shoot wide open due to necessity of deep DoF) so the original footage unedited is of course noisy compared to what things look like in final form...the rendered test videos don't really bug me that much, everything overall looks nice but the colored backdrops exhibit more or less noise sometimes (see above for full details). I considered a slight luminance reduction but that introduces issues in terms of losing elements of the video in the darker parts, and as fr0sty and RogerS are probably (painfully) aware, I have much to learn about the details of color work so I try to only do relatively light adjustments to things or else I am venturing into areas that I either get lucky or don't in a sense.

As I have stated somewhere above, I know I could just theoretically deliver 1080 instead of 4K and likely see a drop in perceived noise, but (BIG BUT) client expects 4K render and my source files (6K HEVC) certainly justify a 4K render.

This is neatvideo of your image, but I'd imagine NeatVideo like many others use data from frame before and after to build the noise reduced image. This is straight default because I don't know what that rear screen is supposed to look like, what actually is the noise? I wouldn't know.

Neatvideo surely has the free trial mode still, download it and see how it looks for yourself

@UltraVista That does look nice and smooth...Thank you for running that, and point taken. @RogerS also spoke about how a video would be a better test...FWIW I did state way above (likely buried cause I tend to write walls of text) that I did already DL and try Neat demo...problem being, lower resolutions automatically by nature show less noise, and the demo only allows for 720p....whereas I am constrained by the gig to do a 4K render (which the source files surely justify as well).

The raw files are VLog so that would be hard for you to duplicate my look...I suppose I could send people a rendered file but then the Neat would be operating on that instead of in the edit session.

I'd imagine NeatVideo like many others use data from frame before and after to build the noise reduced image.

Unlike many others, Neat doesn't just use data from the frame before and after, but up to 2 frames before and 2 after, for a 5 frame spread. This helps it preserve many details many others will filter out.

I'm going to try another experiment...render to ProRes LT (4:2:0 just like my source files). Why not? I have the space, and then I'll have a nice solid intermediate to play with. Then I aim to try transcoding via Shutter Encoder to one or more delivery formats, and giving the noise reduction built into Shutter Encoder (which I believe is from some form of ffmpeg) a trial spin. If it doesn't come out great, nothing is lost except disk space. Plus, my PC seems to chug right along rendering to ProRes, and it is also probably easier for it to do the NR while transcoding the rendered file to H264 or H265 than by trying to do all that at the render stage.

Am I crazy?

Since you start with H.265 10bit 420 (or is it 422?), you should try to maintain the color subsampling. So Prores 420 (or Prores 422 if you start from 422) can be nice.

Voukoder also has noise reduction algorithms built into it which you can do at the time of the render. You can't see what it does before you render so I'd rather use NeatVideo and dial it in precisely.

So I've just bought Neat. Waiting on the download link and watching tutorials.

Those of you who use it...generally speaking:

1. Do you typically use it as a track FX, an adjustment layer, a master video bus FX, or a per-clip effect?

2. In a chain, do you typically put it first, at the end, or somewhere in between?

2a. Before or after Color Grading (if any?)

3. Do you ever put it on two different tracks or clips with different Neat settings on each?

4. If you are working with intermediates (like for example rendering to ProRes then rendering to delivery formats after) do you use Neat at the intermediate stage, or the "delivery format" stage?

Just curious how people tend to use it, specifically in Vegas (lot of internet forums about using it in Premiere etc. but I want to hear about it from the people here).

1. Do you typically use it as a track FX, an adjustment layer, a master video bus FX, or a per-clip effect?

Per video event. In fact, I create a separate project solely to process video events that need Neat NR added as an FX and then render each video event individually (i.e. intermediates) for use in the main project. The main reason for this is that endering a video event with Neat NR applied drastically slows down the render time. Just testing a 6 seconds long 4K .mp4 video event without any FX applied rendered in 7 seconds, but adding Neat NR resulted in a render time of 2 mins 30 secs with an average frame rate of 1 fps (21 times longer than with no FX applied). Of course, different media/render template/computer specs will have different render time results. Nonetheless, a lengthy project with lots of Neat NR being used will result in a very long render time for the final project.

I've not used Neat NR as an FX at the track or project level. I imagine that that would be possible if each and every video event on the track or in the project had the exact same noise reduction processing requirements, but in my experience I've found that video events more often than not need slightly different Neat NR settings specific to the noise reduction needs of each video event. Neat NR is very flexible with its settings - its a rather surgical process IMO.

Keep in mind that if Neat NR is placed as an FX at the project level, it will affect everything in the project including titles, artwork, etc.

2. In a chain, do you typically put it first, at the end, or somewhere in between?

2a. Before or after Color Grading (if any?)

As my workflow has Neat NR as the only FX used before rendering, it is placed right after Pan/Crop. Color Grading is done later on the render/intermediate in the main project.

Neat NR has a good user guide (F1 or via the Help menu) which includes some Q &As.

I also use it sparingly at the event level or media level on clips that really need it. I would add it after color correction if the footage is so flat I can't see the noise (log formats). I don't do a lot of multiple renders so would likely just temporarily disable the Fx and enable before the final render.

1. Do you typically use it as a track FX, an adjustment layer, a master video bus FX, or a per-clip effect?

I only use Neat Video on the track or project level if every camera is in the same room and they are all the exact same kind of camera with the exact same ISO/fstop settings, where I know the noise profile will work universally across all cameras. If that isn't the case, then using it at the event or media level ends up being the best way to do it.