Shaky video and distortion line on bottom on DVD

[litlin]

I have used Hauppauge capture software and Colossus 1 card to capture 20 minute video from High 8 camcorder. IT is m2ts format and is 1 gig. The video looks really good after capture and audio seems fine (after some trial and error)
But when I bring into Premiere, export to mpeg-DVD -> adobe Encore and create the DVD with a nice basic menu system, the DVD when played in the VCR is:
little shaky
is grainy and less quality than the capture version which is both 720 so not sure why.
and has horizontal distortion border on bottom which is visible on PC and in the VCR I am playing it on.

I think I am using max quality and wide screen in the encore. What else can I do to make little better quality. less shaky and make the video fill the screen more and also
how to get rid of the horizontal distortion line on bottom. I heard of a masking technique that preserves resolution, please point me in right direction. I have been looking but no success.

Thanks again for your help in this forum

[johnmeyer]

The horizontal distortion at the bottom is probably head switching or vertical blanking interval. It is doubtless in the original capture and, for whatever reason, your NLE didn't show it. You can add a 2-3 pixel mask at the bottom to hide it. I would NOT re-size, because that introduces additional artifacting.

If you have any sort of shakiness, it makes me think you might have a field reversal problem. Remember that NTSC SD video is usually captured as lower field first, but that HD interlaced video is usually upper field first. If you have any HD settings in your workflow, that could introduce field reversal. This will be extremely obvious when the camera pans horizontally. It is awful stuff and must be avoided by making sure that field dominance is kept lower field first throughout the workflow.

If you provide a sample of the finished video, it will be easy to see if you have field reversal, and will also make it easy to see exactly what the other problems might be.

As for grainy and less quality, there should be nothing in the encoding to MPEG-2 process that should introduce any "graininess," but maybe you are using that word to describe some other artifact, like blockiness. If you are trying to fit more than 75 minutes of material onto a single-layer DVD, you must use a 2-pass encoder. If you try to fit more than about 100 minutes on a single-layer DVD, you will begin to see encoding artifacts. If you try to fit more than 120 minutes of video onto a single-layer DVD, you will be unable to do so without introducing quite a bit of blocky and "mosquito" artifacts.

[litlin]

Thank you for your help. so in premier, I can just overlay a rectangle on that distortion?

I included a short 20 second capture for you to see. IT is the bigger .vob from Encore. IS this ok? or how else should I do it.
Please let me know if you can see the .vob attachments.

THanks

[manono]

Crushed blacks, blown whites, stereo audio with sound only out of the left speaker, and 4:3 video encoded as 16:9 with added black bars (loses resolution).

I'd say the head switching noise is the least of your problems. To get rid of it you can cover it up if you wish.

[DB83]

Well if you really want to fill the screen without either the black side-bars AND the head-switching noise then you could always make a proper 16:9 video from the 4:3 source.

Of course you will have to crop many lines from top and bottom (forget the number for NTSC right now since I am in PAL-land) and you will compromise the framing which could be important for you.

Otherwise leave it at 4:3. Forget about 16:9 since there has already been done at least one re-encoding from the 4:3. And if the noise really bothers you then, yes, add a mask at the bottom. But then if you do want to leave it at 16:9 you get both side-borders AND a bottom border which makes it look worse than it is if you left it alone.

[poisondeathray]

1) manono mentioned levels (blacks,whites) - I would adjust those first because some manipulations like certain types of stabilization in some software work in RGB, and it makes levels manipulations in YUV less ideal (clipping)

2) To make it "less shaky" you would need to use a stabilizer . eg. warp stabilizer since you already have Adobe ; or you can use 3rd party ones like deshaker . A stabilized video will compress better (less noise) , but a side effect of stabilization is it will leave "black borders", so typically some combination of zooming in or border fill techniques are used. Zoom in will make picture softer. Eitherway there is always some sort of trade off . But do this next, because this might affect how much you "cover up" of the head switching noise.

3) Noise , compression and bitrate - John alluded to this above, but the bitrate is about 5Mb/s . Higher bitrate means higher quality, less noise. But that means fitting less running time on a disc (filesize = bitrate * running time) . A higher quality DVD would typically use higher bitrates , maybe 7-9Mb/s .

4) Yes, leave it 4:3 as DB83 suggested

[johnmeyer]

1. Don't ever capture 4:3 video as 16:9 because you are effectively throwing away resolution by including the black pillar bars in your limited allocation of 720 horizontal bits. Capturing as 16:9 isn't optional, is not a good idea, and is simply the wrong thing to do.

2. Yes, you could bring up the levels a bit. I found that increasing the gamma slightly (in Vegas) also effectively brought the black levels in spec, and only a very minor additional levels corrections was needed to "un-crush" the blacks.

3. I agree with everything poisondeathray suggests.

Here is an example of what you can do by changing the gamma and adjust the black levels; cropping it so it is encoded propertly as 4:3 video (although to avoid the huge resolution loss, this must be done when you capture, not afterwards). I then applied de-shaking using Deshaker. I did not do any de-noising, although I've posted my VHS de-noising AVISynth scripts many times and you can use one of those, if you want.

"Fixed" version

[litlin]

Thanks all for your help.
johnmeyer can you please analyze this original. THis is the raw capture video from the sony cam from my Hauppauge Colossus 1.How can I create nice result from this to DVD.

Thanks again

https://forum.videohelp.com/attachment.php?attachmentid=45571&stc=1&d=1525801383

[johnmeyer]

johnmeyer can you please analyze this original.

Why me?

[litlin]

Thanks johnmeyer, you took the time to make nice fixed version from my DVD result. I understand if you cant. I was thinking maybe starting with raw capture is best.

[LMotlow]

What I don't understand is why you're capturing noisy ugly analog tape to lossy codecs when what you intend to do is extensive repair and restoration. I don't know where you learned to work along those lines, but I hope you didn't learn it at videohelp. You're just shooting yourself in the foot from the start. And manono is correct, your illegal signal levels are all over the map, even for PC-only work. Maybe you should step back and do a little more research in the capture and restoration areas.

[DB83]

There are several threads on here that discuss what the Colossus 1 is capable of. But, generally, combine any Hauppauge card with Hauppauge software and you are limited to mpeg2/ts files.

I have used both a internal card and a USB (USB Live2) capture device. The internal device could capture huffyuv/lagarith using other capture software and I have read on here that the USB can do likewise. The program used was called Dscaler. Nothing to lose by trying. I suggest that the OP seeks this out.