Question about future option of VDub pluggin

general questions about Neat Video
Post Reply
jpsdr
Posts: 221
Joined: Mon Aug 11, 2008 7:33 am

Question about future option of VDub pluggin

Post by jpsdr »

Hello.

I can't stop wondering about several things...

If i understoud correctly, adding YCbCr support to virtualdub plugin is on your to do list.
I'll explain to you what i think, but i may be mistaken. So, if it's the case, correct me, and, if possible, i gladly take a little explaination where i'm wrong.

All my working footage came from DVDs (and in near future from blu-ray). So, my sources are all in 4:2:0 format, so YV12. I think that the best way to do things is to stay in YV12 during all the chain of process. Unfortunately, it's not possible for now, as plugin works only on RGB32. So, i must upscale 4:2:0 to 4:4:4. But, doing this, it's change the original noise profile, and on your site, you said that you can eventualy change constrat and things like that wich don't alter the noise, but resampling is not a good thing to do, as it alter the noise. And, unfortunately, going to 4:2:0 to 4:4:4 upscale chroma.
So, i was wondering, if waiting for the VDub plugin wich will work on YV12 worth, for having a more accurate denoising ?
Or maybe because the eyes are very more sensitive to luma than chroma, it will not have a great impact ?

When the future version with YV12 support will be out, i was assuming that noise profile builded with actual version will not be accurated, as they were builded from upscaled 4:2:0 to 4:4:4 video, so the noise would have been changerd. Am i correct ?

Or maybe, for technical reason, you can't support YV12 ?
One problem i think of is the size of the rectangle for noise profile building, sizes will be different in chroma an luma planar.
NVTeam
Posts: 2745
Joined: Thu Sep 01, 2005 4:12 pm
Contact:

Post by NVTeam »

There is not going to be any significant improvement if you reduce noise in Neat Video in YV12 or other similar format as compared with regular RGB. The noise component that we are normally dealing with is in all cases much larger than any small difference caused by color space conversion. I wouldn't worry about it really.

Regarding the comments on the website about resampling, it is not bad thing in itself, you just have to build a new profile after resampling, to be sure you analyze and actually filter the same type of data (resampled data).

Hope this helps,
Vlad
jpsdr
Posts: 221
Joined: Mon Aug 11, 2008 7:33 am

Post by jpsdr »

Ok, thanks.
Post Reply