Page 2 of 2

Posted: Sun Oct 23, 2011 1:35 pm
by NVTeam
Thank you, we will check the project in the coming working days.

Vlad

Posted: Mon Oct 24, 2011 10:22 am
by NVTeam
Unfortunately we cannot directly reproduce the problem using this project. Please provide a step-by-step instruction on how to reproduce it.

Thank you,
Vlad

Posted: Mon Oct 24, 2011 11:14 am
by MoldCAD
Hi Vlad,

The required steps are obvious:

1. Load the GOPR034.MP4 clip into VP T/L
2. Apply NV filter (with or w/o No Lag option)
3. In my case, there is also a WB filter added, but that's not necessary (you may not have this filter, and this may be causing errors when loading my project)
4. Select Tools->Render to New Track, and render the denoised clip to any format you wish (in my case, it's 1080/25p, but feel free to use 1080/29,97p as in the original clip)
5. Check that the rendered clip length is greater than the original, and so they are NOT aligned on the tracks

This makes keeping synchronization with a master audio track impossible!

Posted: Mon Oct 24, 2011 12:34 pm
by NVTeam
Still see no difference when we do that.

Please upload two clips rendered with enabled NV and with disabled NV.
We will inspect these clips and check what makes them different.

Thank you,
Vlad

Posted: Mon Oct 24, 2011 12:43 pm
by MoldCAD
Vlad,

With such a short clip (30 secs), the difference is very small (less than a frame). In order to see it - having the 2 clips one above the other on Vegas TL - you need to turn Quantize to Frame Off, and zoom in considerably around the right end!

If you still don't see the difference, I'll upload the 2 clips for you...

Posted: Mon Oct 24, 2011 12:56 pm
by NVTeam
In Vegas 10.0d, after rendering to a new track:

- original clip: 36.570
- rendered with NV: 36.600
- rendered without NV: 36.600

as shown in track properties > media > attributes

Is that what you see?

Vlad

Posted: Mon Oct 24, 2011 1:02 pm
by MoldCAD
Exactly. The difference grows with time - and believe me, with a pianist hands on the keyboard every split second matters...

Posted: Mon Oct 24, 2011 1:04 pm
by NVTeam
The difference between original and rendered? Or the difference (which is currenty zero) between the version rendered with NV and version rendered without NV?

Vlad

Posted: Mon Oct 24, 2011 1:41 pm
by MoldCAD
Vlad, please give me some time to check again what I thought I was positive about...

Right now it seems it's not the NV filter that's causing this, but believe me I've run so many tests with different render formats and fps'es, and I could swear it was only present with NV:)

Will return to post my final findings; cheers

Piotr