Questions - Neat Video with Vegas Pro

general questions about Neat Video
Post Reply
dalemccl
Posts: 6
Joined: Fri Jun 17, 2011 7:39 pm

Questions - Neat Video with Vegas Pro

Post by dalemccl »

Just got NV Pro 64-bit for use with Sony Vegas Pro 10d 64-bit with Windows 7 64-bit Professional. My clips are 1920x1080i AVCHD NTSC from a Sony camcorder. I have some questions.

EDIT: I deleted Questions #1 and #2. Further study of the manual answered them.
1. deleted
2. deleted


3. What is the meaning of the y-axis in the graph? I understand the x-axis to be increasing levels of brightness (right?), but what does the Y axis represent. Is the goal to get all the squares into the negative area of the y-axis?

4. It may be out-dated information by now, but in a Vegas discussion forum I read that in Vegas, using Temporal noise reduction results in audio and video going out of sync by the number of frames specified in the temporal filter "radius" setting. I understand this to be an issue regarding how Vegas presents data to plug-ins and is not a Neat Video bug. I want to avoid sync problems. But would it be best to leave the radius setting at the default of 1, or to set it to zero?

Or does checking the "No Lag Mode" allow using up to 5 frames with no sync problems? Does the No Lag Mode even apply to Vegas?

Thanks, and sorry for initially posting questions I before more thorough reading of the manual. I don't normally read manuals (usually don't need to) but Neat Video is deeper than most software and requires reading the manual and doing experimentation.
NVTeam
Posts: 2745
Joined: Thu Sep 01, 2005 4:12 pm
Contact:

Re: Questions - Neat Video with Vegas Pro

Post by NVTeam »

dalemccl wrote:3. What is the meaning of the y-axis in the graph? I understand the x-axis to be increasing levels of brightness (right?), but what does the Y axis represent. Is the goal to get all the squares into the negative area of the y-axis?
The y-axis is the level of noise. The goal is to get accurate measurements of the actual noise levels in the image, to cover all parts of the graph. Measurements can be negative or positive in relation to the primary analysis's noise level (0%).
dalemccl wrote:4. It may be out-dated information by now, but in a Vegas discussion forum I read that in Vegas, using Temporal noise reduction results in audio and video going out of sync by the number of frames specified in the temporal filter "radius" setting. I understand this to be an issue regarding how Vegas presents data to plug-ins and is not a Neat Video bug. I want to avoid sync problems. But would it be best to leave the radius setting at the default of 1, or to set it to zero?

Or does checking the "No Lag Mode" allow using up to 5 frames with no sync problems?
Yes, please enable the "No Lag Mode" option and all those issues should not arise to begin with. That option has been specifically added in the Vegas version of Neat Video to resolve the problems caused by the limitations of the host application.

Using the radius of 1, 2 or higher therefore depends solely on the noise reduction considerations: if you need stronger temporal filtration then you can use 5, if you want to disable the temporal filter, set it to 0. Any setting will work fine when the "No Lag Mode" option is enabled and will not cause sync problems.

Thank you,
Vlad
Joran
Posts: 47
Joined: Sun Jan 08, 2006 2:48 am
Location: Oslo, Norway

Post by Joran »

The new OFX video plug-in architecture in Vegas Pro 11 ande Vegas Movie Studio HD 11 does not have these limitations. So I'm looking forward to an OFX version of NeatVideo for Vegas...

Jøran Toresen
NVTeam
Posts: 2745
Joined: Thu Sep 01, 2005 4:12 pm
Contact:

Post by NVTeam »

They are still fixing bugs in the OFX interface in every update of Vegas...

Vlad
dalemccl
Posts: 6
Joined: Fri Jun 17, 2011 7:39 pm

Post by dalemccl »

Vlad, I tried the No Lag option and here are my observations.

I put three clips in the Vegas Pro 10d 64-bit timeline, each in its own pair of tracks (video and audio), all starting at the beginning of the time line so they are stacked vertically. I took care to make sure all three start at exactly 00:00:00:00.

The clips are stacked in this order:
1. Raw clip with no processing
2. Same clip rendered with no processing applied.
3. Same clip rendered with NV, 1 frame radius, No Lag on.

To check for any timing differences I soloed individual video tracks, and watched the preview window. This is what I saw: when switching between #1 and #2, the picture in the preview window remains static - no movement. When switching between #3 and either of the other two tracks, the preview jumps, as if it is on a different frame. In fact, it is exactly one frame different than the other two tracks that don't contain NV. Also when I look at the wave forms of the audio, #1 and #2 are lined up perfectly with each other. #3 does not quite line up with the other two. It is also different by 1 frame.

The NV video and audio tracks are both off by one frame compared the original, but the video and audio are off in the same direction, so as far as I can tell, the NV clip is still in sync within itself - it's just 1 frame different in timing than the original clips.

I tried the same thing again but replaced #3 (NV with 1 frame radius) with a rendering of the same clip, but using a 5 frame radius with No Lag on. Same result - 1 frame different timing than the original clip or the original clip rendered with no FX. So a radius of 1 frame or 5 frames had the same result.

Is what I am seeing consistent with what you would expect, given how NV and the No Lag option work as they try to overcome the plug-in limitations of Vegas ?

I'm not complaining about the 1 frame difference; I'm just trying to understand whether what I am observing is consistent with how NV works. I'm glad the No Lag option is available in NV to try to minimize the challenges that Vegas creates for plug-ins authors. I am very pleased with Neat Video and am finding it a great addition to my editing process.

Dale
NVTeam
Posts: 2745
Joined: Thu Sep 01, 2005 4:12 pm
Contact:

Post by NVTeam »

Dale,

On the machine which I am using at this moment, I have an older version of Vegas and it shows no such 1-frame delay whatever I do. The clip with NV and clip with no effects are always in sync.

I will ask my colleague who has Vegas Pro 10d 64-bit to directly repeat your test. In the meantime, please try to replace NV with another filter and see if using an arbitrary filter also introduces a delay in your test case.

Is it sufficient to just place a clip in a track, duplicate the track, apply the filter to the clip in the new track and then switch the Vegas preview back and forth between tracks? Or you actually need to separately render the clips/tracks, compress (using which codec? - it may introduce delays too) and save the output clip and then watch that new clip? Do you apply NV to the clip (event) or to the whole track?

Thank you,
Vlad
dalemccl
Posts: 6
Joined: Fri Jun 17, 2011 7:39 pm

Post by dalemccl »

>In the meantime, please try to replace NV with another filter and see if using an arbitrary filter also introduces a delay in your test case. <<

Wow.. you pointed out something I had never noticed. I placed a Sony Sharpen filter as the only filter on the raw clip and rendered it. It also shows a 1 frame difference compared to the raw, unrendered clip. I tried two different rendering codecs: 1) Sony AVC 2) MainConcept mpeg2. In both cases the output was set to 1920x1080i upper field first, which is the same as my camcorder's clips. I then tried a different filter, Sony Levels, and got the same results. In fact, it turns out that just rendering a clip/event with no FX in it will cause the 1 frame difference compared to the original.

So, this 1 frame difference is not caused by NV. I'm sorry for the false alarm and taking up your time trying to recreate this.

I'll post about this issue in a Vegas forum and see if others are experiencing this. I suspect they aren't or they would have raised the issue. I read the forums everyday and no one has mentioned this. So maybe I have a setting wrong somewhere.

In answer to your other questions:

1) The 1 frame difference does not show up with a pre-rendered clip in the timeline with NV (or other filters) added. It only shows up in rendered clips.

2) I apply the filter at the clip/event level, not to the entire track.

Thanks for your fast response. And again, I'm sorry to have taken up your time on what turns out to be something unrelated to Neat Video.

EDIT/UPDATE: just to close the circuit on this issue: I had done all the testing described in my other posts with one specific clip. It occurred to me that maybe the clip itself has a problem. So I tried a couple other clips. Result: NO 1-frame difference. Apparently there is something about the first clip that results in the 1 frame problem. Maybe there are other clips that I haven't yet discovered that have the same problem, but at this point it does seem clip-specific.
NVTeam
Posts: 2745
Joined: Thu Sep 01, 2005 4:12 pm
Contact:

Post by NVTeam »

Thank you for checking!

It is possible that the original clip has an incorrect field order tag, so the render/codec skip some 1-2 fields in the beginning. Just a speculation.
If that is the case then this should not reproduce with any progressive clip.

Anyway, it is great that you have localized the source of this issue and can inspect it more closely now. Perhaps the field order is indeed specified incorrectly in the clip.

Thank you,
Vlad
Post Reply