VHS Family Wedding tape restoration

[2Bdecided]

It's odd enough that a "pro" photog would use autogain in variable-light situations and end up with the "pumping" seen in these videos

I'm not seeing that. There are a few shots where the auto exposure circuits of the time would have caused HUGE changes in brightness IF they'd been used, and it just doesn't happen in this video.

The white balance is dubious, but still pretty good for the time.

-- yet I'm having a difficult time trying to acccept that someone would pay out good coin of the realm for results that some members here have said looks "correct".

It's what you would have got on tape if you'd hired someone with some skill and experience to make a video of your wedding in 1985! He's even got rather decent video lights on the camera, which really help indoors (though they make a correct white balance even harder).


The only thing that seems to have happened during capture is that some noise has been removed. I don't use the DNR on my Panasonic machines, but it's not terrible.




The scanned photo is a great help - though the photo is the colour balance you get from 30 year old film, while the video is the colour balance you get from 30 year old video! The photo is MUCH closer to reality, but not perfect. It's a shot-by-shot colour correction job to make the video match, and even then different parts of the same frame need different correction, because different light is falling on it (direct sun / shade / wall lights / video lights).


Sometimes it's easier to accept something for what it is. I'm sure there are people today shooting videos, and introducing all the faults you're trying to remove, just to make them look like they were shot in 1985.


Cheers,
David.

[Skiller]

Sanlyn you should watch this Video of a tube camera used in very low light, and you might understand what "phosphor lag" looks like. There are no DNR artefacts whatsoever in these samples (Edit: scratch that), it's all phosphor lag and it was there the moment it was shot. It's typical and it happens only to bright moving objects.
Have a look at frames 130-200 of sample #3. There's still lots of chroma noise in the center left area, no DNR.

The colors aren't screwed up, they were recorded like that. The colors look absolutely typical for a tube camera (of course that doesn't mean one should not attempt to make it better). The cyan-greenish tint is very typical as well.

I think you are comparing these VHS samples to newer VHS recordings you had seen in the past which were shot with CCD sensor cameras – this really makes a huge difference.


Edit: Although I have to agree the samples do look slightly digitally processed in some way, not sure where it comes from.

Edit 2: OK, I have to apologize, there seems to be some DNR temporal smearing indeed, but it's really subtle compared to the phosphor lag. Where does it come from?

[sanlyn]

Selur, thanks for the sample. I understand phosphor lag from my radio/tv college days. Early RCA color TV cameras were plagued by it until broadcasters started using cameras from Europe and Japan. But how does that get translated into severe aliasing, DCT ringing, edge ghosting, and posterization, even on objects that don't move? They don't appear in the UTube sample, other than the usual UTube low bitrate noise. Of course, that UTube work is deinterlaced. and for some reason sample 3 is cleaner than the others. The two more troublesome samples look much cleaner when deinterlaced -- until something moves, when motion becomes juddery and clunky (and the double-imaging is still there). There is obviously some processing going on in all the samples: fine detail, subtle color gradations and tape grain are not in the images. Clay-face is rampant in all the shots, some worse than others, and in sample 1 the grass and shoes disappear into green soup during motion; those are typical of the archaic DNR in older VCR's.

Whatever the cause, filtering it is quite destructive. The filter paramaters get extremely critical, scene by scene.

[Tafflad]

This Panasonic was a very expensive top of the range VCR at the time (which was a couple of years after most of the tapes were created)

I could ask (very nicely) a 'local' forum member to take a capture of the same scenes for comparison.

[2Bdecided]

I've just noticed - unless I've missed something, I don't think you had an effective line TBC during capture. That VCR doesn't have one, does it? I think the AVT-8710 > ADVC-300 combination won't do that much to fix wobbly line-syncs either.

Yet there are no signs that a line TBC is really needed. If there is no line TBC in the capture chain, the lack of visible problems is another pointer to the fact that this tape is playing back as well as it did on the day it was recorded.




Did you capture any recent-ish broadcast source (e.g. a VHS tape you recorded from BBC One in the 1990s) using the same kit Tafflad? A sample of that will show whether there's anything fundamentally wrong with the capture set-up. It's not completely fair because IME with decent equipment an off-air UK VHS recording will always have perfect levels (while consumer camcorders almost never do), but it'll reveal how strong the DNR was, and it'll show the presence/absence of any other problems.


Cheers,
David.

P.S. Don't wear the tape out.

[Tafflad]

Correct tha the VCR didn't have one ... I purchased the AVT-8710 to provide this.
http://www.avtoolbox.co.uk/avt8710.shtml
Does this not provide required TBC ?

I just checked all DV files are off Cammcorder footage tapes, I dumped all my 'off air recordings'

[2Bdecided]

It's a frame TBC rather than a line TBC. If the line-sync pulses were all over the place, causing any vertical edges in the original video to become squiggly, it wouldn't help much.

I have a line TBC in two of my Panasonic machines, and with most of my own tapes it only makes a subtle difference when capturing with the ADVC110. I don't have a frame TBC. I've never needed one.


I didn't bring it up to criticise. I mentioned it to point out what good condition your tapes are in.


Cheers,
David.


P.S. In the following link, I don't agree with the opinion on DVD recorders' TBCs, and it doesn't mention the very good line TBCs found in some DV camcorders' analogue input stages, but otherwise this is quite helpful...
http://www.digitalfaq.com/forum/video-restore/1853-alternative-avt-8710-a.html#post9889

[DB83]

Maybe I am having a 'blonde' moment. But doesn't a ADVC 300 (the one the OP used), and the one that I have, have a rudimentary line-TBC ?

It certainly 'cleaned up' many a bad VHS I have handled over the years.

The question was asked, yesterday by me and again today by the OP, if anything is to be gained by recapping this tape without having ANY form of TBC or DNR in the capture chain ?

[sanlyn]

There's very little wobble that I see in the captures, at least none that seems to be a problem. I use old "RD" Toshiba's and a Panasonic ES15 for pass-thru -- far from perfect, but you can certainly tell they're working. I even have one old rebuilt Panasonic whose tracking defies description: I've never used a line tbc with it (but they certainly don't make 'em like that any more). And it is correct that a full-frame TBC like the AVT-8710 doesn't help line sync problems at all. That's not what it was designed for. It might be needed as frame sync to avoid lost frames with really chewed tape. But in this case it just gets in the way.

The VCR mentioned is not a low-end machine, but it does look as if DNR was turned on somewhere and was making trouble. The high end TBC-equipped JVC's created similar problems. tfflad can't be blamed for tape aging, of course, unless some sort of hermetically-sealed lead lined storage unit had been used for the last 30 years, and how many people would take those measures? The captures are just very soft and often smeared or blurred for some reason, and I see problems with capturing VHS to DV-AVI. But if that's all that's available, one simply has to work around it.

And please -- don't discard the tapes, even if you get a better capture. Store them carefully, away from heat, moisture, or anything with a strong magnetic field (such as a big subwoofer. Yikes!).

[Tafflad]

It's a frame TBC rather than a line TBC.

I didn't bring it up to criticise. I mentioned it to point out what good condition your tapes are in.

Thanks for explaining

[Tafflad]

And it is correct that a full-frame TBC like the AVT-8710 doesn't help line sync problems at all. That's not what it was designed for. It might be needed as frame sync to avoid lost frames with really chewed tape. ...

The VCR mentioned is not a low-end machine, but it does look as if DNR was turned on somewhere .

If somebody wants to advise on a reasonable capture card (or what to look for) ... I don't mind looking on fleabay, getting one and recapture without TBC. I did post earlier today that I have a Hauppage WinTV card - but would expect that to be worse than the ADVC300 I used.

Annoyingly, I had captured half of my tapes .. then after some advice on the forum, I went and bought the TBC and redid all captures with that in circuit.

There is nothing listed for DNR on the manual, the nearset is the Panasonic AI function ... but this was off.

However I have no way of knowing what settings or even what machine was used when the Videographer created the tape.

[2Bdecided]

Maybe I am having a 'blonde' moment. But doesn't a ADVC 300 (the one the OP used), and the one that I have, have a rudimentary line-TBC ?

Yes, and people have different opinions on that. I haven't used it.

The important thing is that (in my experience!) a line TBC won't work after a frame TBC. On the frame TBCs I've used (professional ones, not these) there's no way a line TBC could do anything with the output because the frame TBC replaces all the sync pulses with clean ones. It doesn't fix the video content itself, and afterwards there's no way a line-TBC can fix the video signal itself because the badly-timed line-sync pulses which matched the badly-timed video (and hence, could have been used as a reference to fix the badly timed video - that's exactly what line TBCs do) have all gone.


Anyway, we're way off on a tangent. I think we're all agreed that there aren't really any significant time base errors in Tafflad's videos!

[2Bdecided]

The VCR mentioned is not a low-end machine, but it does look as if DNR was turned on somewhere and was making trouble. The high end TBC-equipped JVC's created similar problems. tfflad can't be blamed for tape aging, of course, unless some sort of hermetically-sealed lead lined storage unit had been used for the last 30 years, and how many people would take those measures? The captures are just very soft and often smeared or blurred for some reason, and I see problems with capturing VHS to DV-AVI. But if that's all that's available, one simply has to work around it.

OK, I'm going to resist arguing about the adequacies of capturing VHS to DV and the lack of tape ageing for the third time.

DNR. If there is DNR in these captures, it's pretty mild. Some of the areas that look wiped out are illegal RGB combinations. They're getting clipped for display, but if you greyscale() them you can see there's still detail there. Some areas appear to be clipping the in camera. There might be some other areas where DNR (or the external TBC) is smoothing too much, or something (maybe that external TBC? I really don't know) posterised them a little (before more noise was added back in, strangely). It would be nice to avoid that if possible.


The biggest problems by far are the aliasing (which is on the tape), and the weird 1985 camcorder colours (which are on the tape). I reckon it could do with a lower gamma too, e.g. levels(0,0.9,255,0,255,coring=false) but that's kind of subjective.


I'm just trying to be realistic about how much difference re-capturing will make. I don't know how good or bad the AVT-8710 is though.


Cheers,
David.

[sanlyn]

I don't disagree with any of that, and thanks for more detail on the issue. VHS is a nightmare anyway, even when it's "perfect" (sic). I think the AVT-8710 does have something to do with the posterized effect, and on some tapes it pumps gamma like crazy -- it does when I use it on really ugly tape, and I have to be rather desperate for anti-macrovision measures. Often it helps to lower saturation with its proc amp control, then restore it later. Fortunately the pass-thru units I use defeated most forms of copy-protection on old tapes, so the AVT-8710 sees little use.

[Tafflad]

Maybe I am having a 'blonde' moment. But doesn't a ADVC 300 (the one the OP used), and the one that I have, have a rudimentary line-TBC ?

It certainly 'cleaned up' many a bad VHS I have handled over the years.

The question was asked, yesterday by me and again today by the OP, if anything is to be gained by recapping this tape without having ANY form of TBC or DNR in the capture chain ?

I bought the TBC specifically to use with ADVC300 after reading the 'sticky' on this forum:

https://forum.videohelp.com/threads/313473-NOTE-A-Canopus-DV-box-does-not-replace-a-TBC!


I assumed this was the right approach

[sanlyn]

The third paragraph in post #1 of the thread you quoted contains this statement:

When you know you need a standalone full-frame TBC....

Unless you have serious playback issues that require a full frame TBC, such as audio sync problem or serious dropped frame issues, that TBC won't do much for non-copy protected home made tapes. And the concesus is that under most circumstances, the less hardware/wire/processing you have connected in your capture circuit, the better.

Full frame TBC's are used for the reasons described above when those problems become issues, and to defeat many forms of copy protection. they work on a frame-by-frame basis (that's why they're called full-frame) by correcting any frame timing issues that arise during playback. As a gross oversimplification of what a full frame TBC does, they retain a certain number of frames in memory and use circuitry that corrects the output timing signals imbedded in the tape's output. They create a new timing signal that usually erases copy protection interference. Then they output the frames using that corrected signal.

A line-level TBC doesn't work that way. When a tape plays it outputs each scan line in each frame at a certain rate. Tape playback is not nearly as precise in sending out those scan lines as they should be for purposes of digital recording. If the VCR tracks poorly, or the tape is in poor physical condition, line timing errors get even worse. Some of the lines arrive at the recorder a little late, some a little early. We're talking about minute timing errors, but they are enough to make vertical and diagonal lines appear to wrinkle and warp, and they make some objects appear to wobble from frame to frame. You'll also see horizontal ripples, wavelets, moire, stairstepping or "rips". A line-level tbc collects incoming scan lines in memory, then releases them at exactly the correct rate. The correction effect can also reduce some playback noise and some forms of chromatic or motion smearing, or at least somewhat reduces them.

All full-frame TBC's at consumer price levels are imperfect. Some have dot crawl, softening, or sometimes odd effects on gamma. That's the price paid for some forms of processing. But when frame corrections are required, these shortcomings can usually be corrected later with software.The built-in TBC's or pass-thru units usually recommend for hobbyists are sometimes used by pro shops, but a properly equipped pro setup would use far more expensive equipment, which requires other gear and the training and expertise to use it.

[2Bdecided]

The third paragraph in post #1 of the thread you quoted contains this statement:

When you know you need a standalone full-frame TBC....

Unless you have serious playback issues that require a full frame TBC, such as audio sync problem or serious dropped frame issues, that TBC won't do much for non-copy protected home made tapes. The consesus is that under most circumstances, the less hardware/wire/processing you have connected in your capture circuit, the better.

To be fair to Tafflad, I think I've read posts from several "experts" here that imply you'll always need a full frame TBC when capturing VHS with certain devices (ADVCs included) because otherwise it's almost inevitable that capturing will stop when there's a problem on the tape. Some "experts" also suggest it's the default best capture path.




I have a line TBC in my VCR, no frame TBC anywhere, and my ADVC110 will happily capture through blank sections of tape and damaged sections of tape. It also does something else that the "experts" here say it does not. You have to take some experts with a pinch of salt.


Cheers,
David.

[sanlyn]

Agreed. I don't know where some of he experts get some of their information.

Just to note, I did a couple of restorals from VHS that did indeed require both a line tbc pass-thru and a full-frame TBC, because those two tapes were in really godawful shape. One rule of thumb: if you use a VCR that has no built-in line TBC but you work around that with a tbc pass-thru device that reads the VCR's output, the line-level TBC comes first in the circuit, then the frame TBC. If you use the frame TBC first in line, the line tbc will see no errors in the frame tbc's output and, thus, will do nothing.

[Tafflad]

David ... that was the way I 'read it' ............ and I'll fully admit this was from a position of ignorance.

From a few posts here (and on DigitalFAQ) I had understood that I needed to get an 8710 as the ADVC300 did not have full frame TBC, and to ensure that the signal presented to ADVC was of full correct strength (1V ?) and with frames in full synch I needed to get one.

It was mentioned 'I may not need one' but I honestly did not read anywhere that using one would degrade the capture.

I no longer have this ....... and am looking at getting a PCI capture card (anybody care to recommended one) ... I'll not get a TBC this time (pun intended)

[Tafflad]

Agreed. I don't know where some of the experts get their information.

If you use the frame TBC first in line, the line tbc will see no errors in the frame tbc's output and, thus, will do nothing except recycle every scan line and send them out the same way they came in, wrinkles and ripples intact.

Thanks for another piece in my education puzzle

[sanlyn]

...and that tip came to me by connecting it wrong way and wasting an hour of my time. Nothing unusual there.

[Brad]

There's no technical reason that a full-frame TBC can't do everything a line TBC does in addition to its own functionality. It has all of the data and memory necessary; the handful of devices sold to us consumers as full-frame TBCs just don't happen to implement that processing.

The "pro" full-frame TBC/frame sync I have does correct line timing... but not as well as the DMR-ES15!! I also have a capture card that does a very good job correcting line timebase, but it likes to drop frames as soon as there is any disturbance to vertical sync -- and of course I can't place a full-frame TBC inline before it for the reasons outlined above.

[davideck]

There's no technical reason that a full-frame TBC can't do everything a line TBC does in addition to its own functionality.

Absolutely! My Philips HDD/DVD Recorder is another example of a TBC/Frame Synchronizer with excellent TBC performance. It is noticeably better than the "pro" FOR-A 300 TBC/Frame Synchronizer that I used for a while.

There has always been confusion on this Forum regarding TBC functionality. Long ago, I wrote a brief technical explanation for those interested in more detail;

https://forum.videohelp.com/threads/221219-TBC-Fundamentals

In short;

1) Appropriate Vertical Blanking can eliminate some forms of Copy Protection

2) Frame Synchronization can eliminate some causes of Audio/Video skew.

3) The actual Timebase Error Correction capability (what I refer to as "TBC performance") of a given "TBC" is a function of the input sampling as described in the link above. The sampling process required to digitize an analog video signal always does so with some inherent TBC performance.

More discussion about TBCs and their performance;

https://forum.videohelp.com/threads/331681-s-video-artifacts?p=2141140&viewfull=1#post2141140

[Tafflad]

The "pro" full-frame TBC/frame sync I have does correct line timing... but not as well as the DMR-ES15!! .

I have a DMR-E95H should I consider having that in the capture cct .... or only when a problem (or is DMR-E95H not suitable)

[sanlyn]

The Panasonic has a built-in line tbc and does some basic frame sync as well. You can use it as a line-tbc pass thru device. That means connecting s-video from your VCR to the Panasonic s-video input, then connecting the Panasonic's s-video output to your capture device. Turn the Panasonic's power on, but don't hit "record', and you wont't need a disk. Just allow the signal to "pass thru" the Panasonic to your capture setup, and the Panny's tbc will do its thing.

Use the Panasonic's Line 1 input only. TBC isn't active on the other inputs.

The ES15 has the superior tbc, but the '95' is better than none. Good luck finding an ES15 that still works.

[Tafflad]

Would you use this as standard for all captures or minimise record cct and only use if a problem ... or is it better to have a cleaned up signal hitting capture device (which was my original thought plan for AVT-8710 ... which seems was in error)

[sanlyn]

If you don't have a VCR with bullt-in TBC, use it for all tape captures. You might also find that using it as pass-thru will defeat many forms of Macrovision on VHS, eliminating the need for a frame TBC for that purpose. Unfortunately, that works only via pass-thru: try to record a protected tape into the machine, and Panasonic will very rudely refuse to do it.

[Tafflad]

I can find no mention that my NV-HS800 vcr has any tbc function ... unless someone knows different ? .... doies not mention it in the manual.
Luckily no commercial tapes - other than this wedding one and this has no Macrovision.

[2Bdecided]

Would you use this as standard for all captures or minimise record cct and only use if a problem

You need to try it both ways and look critically at the results.

(Which goes against my advice of not wearing your tapes out).


If you want to spend money on the job, unless the current machine is the one that recorded the tapes you want to transfer, then getting a decent S-VHS machine with a built-in line TBC is a good start. It's difficult to find ones that you can be sure are still working well though.


Many people report that decent VCRs (without TBCs) looped through an ES10 or ES15 give as good results. Other people are convinced the ES10 or ES15 will ruin the image. I haven't used mine yet.




I think you might be getting your hopes falsely raised about the increase in quality you can expect by re-capturing. Still, at least most of this equipment is easy to re-sell, so it might not cost you anything overall. Except time.


Cheers,
David.

[Tafflad]

OK ... taken the plunge just bought via fleabay a Canopus ADVC110

Once it arrives I'll try captures with & without DVR in the capture cct.

A comment was made previously that the SONY codec (default on my PC) used when I captured with WINDV was not a good choice ....
could someone advise what codec I should install ?