Reducing vertical jitter with standalone DVD recorder?
 

I have this huge collection of tapes with some horrible jitter problems (and/or wrong field order) all over the tapes, and it's really driving me nuts! :mad4:

But recently, I made a nice discovery when testing my "Pioneer DVR-530H" standalone DVD recorder. It seems like it actually reduces the amount of jitter quite a bit on most of my tapes, see the attached files for examples. Both files was captured simultaneously from the same playback from my VCR. "File 1" captured from the VCR's S-Video output (to the "datavideo TBC-1000" to "Philips DVDR 3360H" recorder), while "file 2" from the VCR's SCART output (set to S-video) and straight to the "Pioneer DVR-530H" DVD recorder. As you can see, if comparing the two files frame by frame, there is a HUGE difference with a lot less bouncing frames/fields in "file 2" that was made by the "Pioneer DVR-530H" recorder.

Strange enough, I haven't noticed any discussions on this subject elsewhere in this forum? Now, what I really want to know is if there's any other DVD recorders with this nifty feature out there? If so, it would be really nice to create a list of what models to look for, wouldn't it? :-) So, has any of you guys made similar discoveries with your DVD recorders? If so, I would love to know what brand and model it is! :congrats:

PS: According to the specs sheets of my recorder (Pioneer DVR-530H), it does have some TBC technology built into it. It also seems like it's possible to use it as a pass-thru device. The only problem is that it does drop out easally on tape errors so I'm really interested in a list of other recorders that can do the same thing! :wink2:

Well done, but I've used oboard DVR tbc's for 10 years. Most DVR's have tbc activity on some or all line inputs for recording. But placing the TBC-1000 in front of the Phillips caused the Phillips to see no impulse errors. You should have used the TBC-1000 on both tests, then ran both tests without it. So you've faulted the Phillips thru lack of control testing. Thus, invalid conclusion about the Phillips. There's nothing wrong with the field order; it's quick-n-dirty NTSC-to-PAL and involves originally telecined or interlaced video with added field or frame blending.

Didn't you post one yourself and get some answers? http://www.digitalfaq.com/forum/vide...rders-tbc.html

You'll find other posts if you scour the archives here: http://www.digitalfaq.com/forum/vide...-es10-dvd.html.

In my own use:

Panasonic DMR_ES10
Panasonic DMR_ES15
Toshiba RD-XS32
Toshiba RD-XS34
Toshiba RD-XS35
Toshiba RD-2, 3, 4
Panasonic DMR-ES20
Those at the top of the list are better than those at the bottom.

sorry, the Pioneer and several others were already tested. Like most DVR's, it failed for pass-thru. The models I listed above are used for pass-thru line tbc and some level of frame sync.

There were several posts with tbc pass-thru demos and partial restoration from losssless media. One post is here (Toshiba preliminary test samnple): http://www.digitalfaq.com/forum/vide...html#post43209

Or these test restorations from bad tape played on a non-TBC VCR thru an ES15 or ES10, captured to lossless media for repair and restoration: http://www.digitalfaq.com/forum/vide...html#post44631

But why repeat what's been discussed thoroughly elsewhere? For instance: http://forum.videohelp.com/threads/3...hat-do-you-use.

As you know, the usual recommendation for VHS capture is to cap to lossless media if you intend to do any fixup or repair. Considering the blown-out levels and other problems in your samples, jitter is the least of its worries. Was this a bad tape tape recorded to lossy media for archiving? I've had many a run-in with those myself using my now-deceased DMR-ES20 with its LSI chip and onboard tbc. I wonder how it would look if capped to lossless media? Have you tried playing this tape on other players with real tbc's built-in?

sanlyn: I think you possibly missed the "vertical" part? :unsure: Most discussions about the DVD recorders don't compare their abilities to prevent frame/field bounce.

hysteriah: One problem is that no one has a standard "control" capture device to compare against. Vertical jitter is caused by the ADC failing to determine the start of the frame. So every design out there will vary in its tolerance for malformed signals. If recorder A can grab frames 1-3 properly while failing with 5-9 and recorder B grabs 6-9 while failing with 1-3,5 which is better at handling jitter? :hmm:

- Horizontal jitter = timing error = video jargon definition of "jitter" = includes tearing/flagging/skew errors ;)
- Vertical jitter = not technical jitter, bouncing, layman description; I always forget the jargon

The error appear to be tracking based. Know that it can change from playback to playback, even under identical workflows.

The TBC-1000 is known to have a slightly positive effect on vertical jitter.

This can vary wildly from tape to tape. I often deal with this, and it's one reason I have a stack of varied VCRs, TBCs, and other filters. Eventually, some combo will make it behave. Usually.

A very good point, I neglected to mention it. Only one of the samples I linked to had noticeable vertical jitter and it was made without the benefit of any kind of tbc. The "fixed" version of that sample used both a line-tbc pass thru and an AVT-8710. The others were all made with pass-thru and external frame tbc's.

Very good point as well. Could be the basis for a whole battery of tests with problematic tapes and players.

Added later:
A sample of obvious and severe vertical jitter referred to earlier is in a post at http://www.digitalfaq.com/forum/vide...html#post35345. Attached to that post is a "bad" sample with vertical jitter and other severe problems. The partial test "fix" is also posted as an attachment and required a tbc pass-thru as well as an AVT-8710 frame-level tbc.

Hi guys and thanks a lot for all your replies. Gosh, you guys really are the GREATEST! :congrats:

Sanlyn: I think you misunderstand. This is all about "vertical jitter", I mean bouncing frames/fields and that means bouncing frames/fields only. See my sample clips to see what I mean ;-)

Aha... I think I understand... so in this case (on File 1), it's probably the "datavideo TBC-1000" that is "causing" the vertical jitter? I think all my capture devices acts pretty identical when connected after the "datavideo TBC". It has 4 video outputs and I'm sure I've recorded from them simultaneously onto several devices with quite identical resault, especially when it comes to vertical jitter, so the problem in this setup must be the datavideo TBC, right?

Thanks for clearing that up so perfectly. I'm always confusing them :huh1:

Yes, in this particular sample, it does have tracking issues. This tape was recorded with my old Fisher VCR that sure was misaligned and suffered from tracking issues. But, this was just the best (or maybe worst) example that I could find to upload for you right now in a hurry. The problem is pretty much on all my tapes, no matter what VCR that was used when recording them. See the new attachements for better examples ;-)

In this case, I actually think it's the oposite. I've never experienced as much vertical jitter as I've done after adding the "datavideo TBC-1000" to my setup. At least, the "datavideo TBC-1000" doesn't even come close to the "Pioneer DVR-530H" DVD recorder at defeating my vertical jitter :question:

I still don't think it's the same problem we are talking about. Please take a look at this new attached sample clips. If you open up the "File 1" in VirtualDub, you will notice that frame #13, #31, #59, #80, #289, #324, #613 (identical), #644, #658, #675 and #703-> does bounce. If you compare it with the "File 2" (recorded with the Pioneer DVR-530H DVD recorder), you'll see that it's only #613 and #703-> that bounces in both version.

So, what to do? I guess I should stop using the datavideo TBC-1000? But what to replace it with? :question:

Mmm, isn't that what's happening shortly after the start and near the end of the "bad" sample in my link? The fact that the VCR I was using tracked a bad tape well enough to avoid complete screen blackout doesn't mean the problem isn't up and down jitter. The final video using a better pass-thru is even cleaner than the "fix" version posted, but I no longer have that video which is in the owner's hands now.

There's little or no vertical jitter in the other samples I referenced because both were made with pass-thru and frame-level tbc's, as a demo of how a good capture setup and restoration techniques can retrieve watchable video from a bad tape. I did have another high end VCR with that tape but it refused to track well enough to suit the tape. Thus, many members here have more than one VCR. If you don't see vertical jitter in those other samples it's a sign that the setup worked. Without those tbc's, frames were all over the place. But the bad captures were useless, so I discarded them.

Yes. The first device in the chain that samples the video into frames will lock-in its bad-signal decoding ability/disability for every downstream device. Everything after the TBC-1000 is seeing a nicely-formed, valid signal like a DVD player would output. Devices that only sample whole fields or less (line TBCs without frame sync i.e. 98% of VCR TBCs) only lock-in their own horizontal stabilization abilities, not vertical. All of this is the same reason that some devices prevent frame drops downstream while others don't. A frame drop is the complete failure to decode a frame and a bounced frame is a partial failure.

Nicely stated. Thus a line tbc precedes a frame tbc, if I'm not mistaken. Much of what I see in the submitted samples is skew error and jitter mix, if I interpret the problems correctly. If not, correction please.

I opened both in VDubMod and trimmed 2 frames from the start of File 2 to align them. I think your frame numbering differs from mine.

Then I got bored. ;)

I'm sorry for this late reply, guys. I've been busy doing some important comparisions between all my capture devices :)
I'm sorry, you are perfectly right, Sanlyn. I thought you were talking about the first sample clip, the one from "zamme", the first post in that thread. I now see the same kind of jitter in the "bad sample link" that you posted. I'm sorry, it was my misunderstanding :-)

Really nice explanation. Thank you ;-)

Yes. As you can see, there is really a lot less vertical jitter from the Pioneer DVD recorder than what's outputed from the datavideo TBC-1000. I've now done a little comparision of all my capture devices on how they react on this vertical jitter problem of mine. The only device I didn't bother to test is the "Panasonic DMR-ES10" because I never really liked the resault from it. The resault from my test was as follows:

- The WORST unit with LOTS of jitter was my LSI chipped DVD recorder, "Philips DVDR 3360". Without the datavideo TBC in front of it, the picture from it really bounces all the time, with large jumps (1 to 8 pixels up or down). The resault was a total disaster.
- Next worst is actually the "Datavideo TBC-1000". It's a lot better than the Philips DVD recorder, but it still has way to much jitter on most of my tapes. They are all similar to the sample clips that I posted earlier in this thread. All in all, the "datavideo TBC" is pretty useless on most of my tapes, UNFORTUNATELY :depressed:
- Next best on this vertical jitter problem is actually the "Canopus ADVC-300". It has noticably less bouncing, maybe half as many frames as the the datavideo TBC-1000, and the frames/fields only jumps 1 pixel up or down every time. The ADVC-300 is still not good enough for my taste though :mad4:
- The WINNER with large margin is without doubt the "Pioneer DVR-530H" DVD recorder. It only bounces once or twice for every 10th bounce from the datavideo TBC. I'm quite happy with this amount of jitter in my recordings and it's probably the best I can get from my tapes, I guess? :question: What kind of TBC is inside of this unit, actually? Is there anyone who knows?

Now, there's just one more thing that I don't understand:
Are you really sure about this? I wonder because... when I connect my VCR to my Pioneer DVR-530H's input... and connect it's outputs to f.ex. my canopus ADVC-300... it certainly outputs a signal as free from vertical jitter as when recording with the DVD recorder itself. Isn't that what defines a pass-thru?

The only problem I've seen when using the Pioneer DVR-530H as a pass-thru (except that the picture goes black on tape errors), is that it outputs some kind of "dotted line" on top of the screen to the left (see attached screenshot). Where does this white dotted line come from?
http://www.digitalfaq.com/forum/atta...1&d=1475800719

It's the WSS used on PAL line 23. The Japanese engineers screwed up when they specified PAL DV, causing non-picture Vertical Interval data to be encoded into the frame.

See the two links at the bottom of this post for further technical reading, if you'd like: http://www.digitalfaq.com/forum/vide...html#post45810

After opening a thread about annoying vertical jitter myself (https://forum.videohelp.com/threads/...ield-starts%29), I was pointed at this page.
I've contacted the TS (hysteriah) and was kindly strengthened in the idea of trying a Pioneer DVR-530H or DVR-630H.

After obtaining a DVR-530H-S (Silver?) and inserting it as a pass through between my VCRs (Philips VR1100 rebranded NV-FS200 and Blaupunkt's rebranded AG-1980P) and my full frame TBC (Kramer FC400), the frequent vertical shifting of fields disappeared entirely!
I am very enthusiastic (and surprised) about this remedy.
Apparently, these Pioneers indeed have something unique in their field reconstruction HW/SW.
Actually, it made my TBC quite useless, because also the horizontal jitter reduction of these Pioneer DVD recorders is pretty good for my old VHS camcorder tapes.
Highly recommended!