WTB: Time Base Corrector for S-VHS

[brabus]

Hello everyone,

I am restoring some old VHS tapes, and I need to buy a TBC (Time Base Corrector), like the classic AVT-8710 or TBC-100 - TBC-1000.
Naturally I am interested in evaluating other models, if you have something different.

If anyone has an used item of this kind, I am interested to buy it. Yes, it is a quite exotic piece of gear. 

Thanks a lot!

[Wuerstchenhund]

I am restoring some old VHS tapes, and I need to buy a TBC (Time Base Corrector), like the classic AVT-8710 or TBC-100 - TBC-1000.
Naturally I am interested in evaluating other models, if you have something different.

If you can't find them then have a look at the Grundig GV-280S, Panasonic NV-V8000 or Blaupunkt RTV-1000 (rebadge of the Panasonic). These are SVHS VCRs that all have a built-in TBC, the Grunding for playback and the Panasonic and Blaupunkt for playback and recording.

Just an idea....

[brabus]

Hello Wuerstchenhund

Thanks a lot for your answer, your advice is right: I already have a Metz 9877 (rebadged version of the Panasonic NV-HS1000).
The internal TBC works great, but I keep on having issues with the signal integrity.

I explain the situation, if someone is interested:

Quality Vs. connection:
VCR --> TV: Perfect
VCR --> Capture Device: Bad chroma issues, sometimes OK, sometimes terrible. Horizontal coloured stripes all over the place.
SVHS Camera OUT --> TV: Perfect
SVHS Camera OUT --> Capture Device: Perfect
VCR --> SVHS Camera IN: Not bad, but colour stripes flashing from time to time, just like VCR --> Capture Device.

These trials seem to demonstrate that the four devices work great if analyzed individually.
My suspect is that the Sync and colour burst produced by the VCR are a bit weak; the TV* can acquire them without issues, but the acquisition device needs sharper signals.

I found the schematics of my TV, and tried to use the internals to regenerate my sync signals (just by whitdrawing the signal after the input matrix/preamplifier, removing a couple of 0402 resistors), with no success.

Today I am going to look at the Y and C signals with the oscilloscope, so we'll see the difference between the perfect signal produced by the camera and the problematic signal from the VCR.

I wanted to use a TBC like the AVT-8710 because it is a full-frame (instead of the simpler line TBC of the HS1000), and it provides AGC and sync regeneration for Y/C signals.

That's my last chance, I spent too much time on this stuff already, and I still can't figure out what is wrong. It hurts my nerves to see such a perfect picture on the TV, and the total colour disaster on my PC.

Thanks a lot if you read the whole thing.

*TV: Sharp LC32D65

[grumpydoc]

Quality Vs. connection:
VCR --> TV: Perfect

TVs are often quite forgiving of poor input signals.

SVHS Camera OUT --> TV: Perfect
SVHS Camera OUT --> Capture Device: Perfect

Is that the camera playing the tapes, or just live output?

If the former why not use the camera to play the tape into the capture board?

You did say buy - what budget were you looking at?

I have no experience with these things but would this sort of beast http://www.ebay.co.uk/itm/Evolution-G2-Multistandard-Time-Base-Corrector-SDI-Rack-Mount-/201421019427 be of interest?

[brabus]

Hello grumpydoc, thanks for your interest.

You did say buy - what budget were you looking at?

Let's say that the quality of the result has priority, so an investment of 1000€ is not an issue.

The issues are the following:
- I do not want to buy-and-try hundreds of € worth of gear, and discover that the problem is still there.
- I hate to try solving problems by shooting at them. I want to discover the root cause.

I wanted to try with the TBC, but I am really not sure about it.
I just observed the signals with the oscilloscope, in fact sync and (especially) chroma burst are a bit weak from the VCR, and almost ideal from the live camera. Maybe the color burst is really the cause, and the TBC can help reconstructing it.

TVs are often quite forgiving of poor input signals.

This is true, therefore I am totally lost. Why is it so? Why does a TV work better than a capture device?
This 5 years old Sharp LCD TV is amazing with analog signals (also CVBS!), but at the same time a brand new Samsung (in the other room) exhibits major issues. It literally sucks, I see grainy artifacts on the screen when selecting analog inputs. Maybe old TV's were designed to better accomodate analog signals; today almost no one has a VCR anymore, therefore producers tend to save on the old inputs, is this an educated guess?


What drives me crazy is this incredible difference between the results.
Even the crappy USB grabber gives me clear, crisp, *perfect* results with the SVHS camera connected to it (in live output). My problem is that the tapes are full-size VHS, that I cannot play anymore with the camera.

I found information about people having the same issue with the best Canopus devices, so I am thinking that my issue is not on the capture side, but on the source instead.

Let me add this picture to show you the problem:


Look at that green bars and that shocking orange tablecloth. Just look at them.

The same picture on the TV is flawless, the colours are normal and detail is crisp.

Again, I explored all the settings of the device, with no success.
Still, SVHS camera is shown flawlessly. Just VCR signal gives me this horrendous result.

[aargee]

I do think if you nab the problem at the source, the VCR, you might get better results. Try getting a commercial Panasonic, like the AG-7350. I have two of them here, but I'm a bit far away!

For me they have produced rock solid outputs, even from the crappiest tapes, which leads me to also warn about the physical state of a lot of tapes now days which generally love depositing themselves all over the heads of the VCR.

The 7350's and their kind were used a lot in medical equipment, like Ultrasound. So a hunt through scrap/old gear such as this may turn up some bargains.

[desy2820]

I'm wondering if copy-protection is at play here.  Macrovision?  I think, it's been a while.  I had a similar problem with a Sony VCR when trying to copy tapes, if it was anywhere in the signal chain, my recording was garbage.

I have no idea how or if it can be disabled, as I just borrowed a different VCR, which worked for me.

Might be worth exploring.....

[grumpydoc]

TVs are often quite forgiving of poor input signals.

This is true, therefore I am totally lost. Why is it so? Why does a TV work better than a capture device?

I'd like to say that it is because they are designed to be robust in the face of severe signal degradation but I'm not actually sure that is true 

MacroVision has been mentioned which might be a possibility if these are pre-recorded tapes. Otherwise poor/marginal chroma and sync which is the avenue you are considering with TBCs

What drives me crazy is this incredible difference between the results.
Even the crappy USB grabber gives me clear, crisp, *perfect* results with the SVHS camera connected to it (in live output). My problem is that the tapes are full-size VHS, that I cannot play anymore with the camera.

Ah, OK so we're comparing live signal from the camera with playback on the VCR. Not quite apples-for-apples but perhaps useful when it somes to looking at the signal for possible causes.

I hate to try solving problems by shooting at them. I want to discover the root cause.

Nod, but sometimes you have to test a hypothesis by doing the experiment.

[brabus]

Guys, first of all THANK YOU for all your hints, they are all relevant and interesting. I am amazed by the fact that someone is spending so much time and energy to help me.

Let's go through your hints:

I do think if you nab the problem at the source, the VCR, you might get better results. Try getting a commercial Panasonic, like the AG-7350. I have two of them here, but I'm a bit far away!

Very interesting piece of gear, I am very tempted to try one of them!
I already tried a different VCR, a Panasonic NV-F65 with CVBS output. Same thing: crystal picture on the TV, but color flicker on the capture device and on the camera input*. Also the analysis with the oscilloscope showed a reasonable signal, without spikes or differences between even/odd fields. By the way, it was the first time I tried the "PAL TV Trigger" functionality on my 54621A, it works great!

I'm wondering if copy-protection is at play here.  Macrovision?  I think, it's been a while.  I had a similar problem with a Sony VCR when trying to copy tapes, if it was anywhere in the signal chain, my recording was garbage.

This is also an interesting hint; I alreay met this issue with commercial tapes. But my tapes have been recorded in the 90's with a normal VCR**, on brand new blank tapes. I am impressed by the quality of those tapes when played on TV: no drop-out whatsoever, even after 20+ years.
--> I assume that the tapes are OK.

Ah, OK so we're comparing live signal from the camera with playback on the VCR. Not quite apples-for-apples but perhaps useful when it somes to looking at the signal for possible causes.

Exactly. I wanted to perform a test with a "reference" signal, in order to exclude any issue on the acquisition device.

Nod, but sometimes you have to test a hypothesis by doing the experiment.

Unfortunately, that's right. I have no other means, at the moment. I can only suspect chroma/sync issues, hoping that the TBC can solve them.

I am considering designing my own TBC, based on a TVP5160: http://www.ti.com/product/tvp5160/description
This chip really seems full-optional, and paired with the ADV7311 they form an impressive signal processing commando.
The TVP5160EVM schematics are a piece of cake, a very good basis to develop the board.

Theoretically, a TBC works in the same way, right?
Signal IN --> ADC --> frame buffer --> processing --> DAC and sync/chroma reconstruction --> Signal OUT

I still have to find a good TBC teardown, to better understand how it works!

*Camera: Canon MV400i with CVBS/SHVS in/out capability.
**Sanyo VHR-290, long time gone.

[Psi]

I'm wondering if copy-protection is at play here.  Macrovision?  I think, it's been a while.  I had a similar problem with a Sony VCR when trying to copy tapes, if it was anywhere in the signal chain, my recording was garbage.

This is also an interesting hint; I alreay met this issue with commercial tapes. But my tapes have been recorded in the 90's with a normal VCR**, on brand new blank tapes. I am impressed by the quality of those tapes when played on TV: no drop-out whatsoever, even after 20+ years.
--> I assume that the tapes are OK.

My old Sony VCR had a switch labeled "Rental" on/off.
I found when copying tapes using it as the player the copy looked terrible unless i set it to off.
I also found if i used that VCR as the recorder they came out fine regardless of the setting. Except, when i tried to copy them again. They looked terrible if i had copied them with the setting on.

I think the switch turned the copy protection on/off for playing and recording 

[dexters_lab]

would looking at the signal on a scope help identify where the issue might be?

probably a good idea to identify the exact problem before spending $$

[SeanB]

I have some Panasonic professional recorders, which have a feature in that you can turn off the video AGC, which means they then do a superb job of making copies of tapes with Macrovision encoding, which was very useful duplicating tapes.  A lot of the semi pro Panasonic equipment also has the ability to turn off the AGC, which works well for video with poor sync, but otherwise good quality. Just means there is going to be a slight luminance variation if the source has one, but they also have manual level controls for both audio and video level.

BTW, a video recorder will work very well even with heads worn down to the point where they no longer will play back video clearly, but as the recording side only cares about the edge of the head at the point where it leaves, not the actual gap, it will run with severely worn heads. I still have some video recorders, which actually do not have the ability to play back the video off the tape other than by plugging in on the board inside a cable with the video and audio out. No front panel controls either other than power and eject ( and the associated cassette stackers) with only a video and audio input, along with the input to loop the remote signal through or to connect an external control panel. I automated this for duplicating under licence, though sales dropped off a cliff after DVD came out, so it sat in a corner.

[brabus]

Thanks a lot again for your answers, every hint helps to build up the whole picture (pun intended ).

I am bidding on an AVDC-55, theoretically a better video acquisition device than my USB dongle. I am not 100% satisfied with the overall noise of the USB dongle, so the AVDC will improve the situation in any case. If the problem persists, TBC is on my list.

I will keep the 9877 VCR, since TV playback really is flawless. I looked for copy protection switches, but I found nothing at all; plus, the problem persists also with the Panasonic NV-F65, so I would move the spot from the source to the acquisition device.

Let's see!

[brabus]

...and I won the bid on eBay! 

Let's see how this '55 performs.

[SL4P]

Another thought to toss into the mix...

A (consumer/prosumer) TBC is sometimes a short-term 'resynchroniser' to regenerate consistent output timing - based on the median input timing, but if the source timing (i.e. the reference) has drifted (line/frame/sc freq, drum, capstan speed etc), the output may be consistent - but still out of spec for transcribing etc.

The loose input tolerance of a tv/monitor can hide these visually, but when re-recording, the errors accumulate in microseconds, and can't be ignored.

What you may want to look for is a device that is slightly higher up the food chain - a combined TBC/Frame synchroniser that re-clocks and re-phases the output to a known-good reference input.   and yes, they can freeze/repeat - or drop lines, fields, frames to recover the timing integrity...

Just a thought.

[Richard Crowley]

Note that the Canopus (now Grass Valley) ADVC-300 model features built-in time-base correction.
The lower models (55, 100, 110) did not have TBC capabilities.

[brabus]

SL4p, Richard, thanks a lot for your suggestions. In fact the ADVC-300 was on my wish list thanks to the built-in TBC.

But after reading many controversial comments* on the additional filters in the '300, I decided to keep the chain simple and controlled --> I chose the ADVC-55.
If the result will still be disappointing, a full-frame TBC will find its place between the VCR and the ADVC - probably the AVT-8710.


*For example: http://forum.videohelp.com/threads/356262-Canopus-ADVC-100-vs-ADVC-110-vs-ADVC-300

You might find you need the 300's proc-amp, and wish you could defeat the 300's (supposedly undefeatable) primitive noise reduction.

[Richard Crowley]

Proc Amp, Noise Reduction, And Time-Base Correction are three completely DIFFERENT and INDEPENDENT functions.
Proc Amps and Noise Reduction will do NOTHING for time-base errors. They will potentially make it worse.

I saw NOTHING in that forum that appeared to be written by an adult who had half a clue what time-base error is.
What a pathetic mess of misinformation.

[SL4P]

Proc Amp, Noise Reduction, And Time-Base Correction are three completely DIFFERENT and INDEPENDENT functions.
Proc Amps and Noise Reduction will do NOTHING for time-base errors. They will potentially make it worse.

HALLELUJAH brother!

[brabus]

Richard, do you believe that my chroma disaster can be related to a time base error?
I really can't figure out the origin of that mess (see picture in previous page), and most of all, I can't figure out why the same picture looks flawless on TV.

Yes, maybe that website is not the best example; in fact I spent many many hours here: http://www.digitalfaq.com/forum/video-restore/2251-tbc-time-base.html
Still, there is no clear definition of what an ideal signal chain should look like, every contributor has a different setup, and basically all of them found their way by trial and error.

I guess I am going to have the same destiny. If the ADVC-55 performs bad even with the TBC, ADVC-300 is next on the list.

It's a bit annoying having to try each and every single device on the market, but I see no alternative.
But most of all, it hurts my nerves having no clue about the root cause of my issue (see previous picture).

[Richard Crowley]

Yes, that looks very much like time-base error.  It was very common for old tapes to look fine on a monitor/TV, but not be anywhere near stable enough for copying, editing, or digital capture. This was why TBC was invented.  Displaying video on a screen is a far less critical function than accurately capturing the original Y and C signals into digital form.  And newer digital TV/monitor screens are quite likely less tolerant of time-base errors than older models. Just because it is no longer an issue here in the modern, digital era.

Your experiments with the camera, etc. are excellent troubleshooting procedure. The fact that the tapes look good on a screen mean that you probably have only simple time-base errors and not something more serious.

That does NOT look like Macrovision or any other kind of copy protection. They used several different schemes to "diddle" the frames outside the active area that would molest the ability of a VCR to decode the video signal, while leaving TV screens oblivious. But they did NOT use any kind of time-based diddling that I ever saw.  It was mostly throwing "garbage" in the vertical blanking interval to confuse the automatic gain functionality of consumer recorders.  Many commercial TBCs would very handily remove any Macrovision artifacts.

You have not revealed any background about these tapes. The more you know about them, the better chance you have of devising a scheme for replaying them successfully.  For example, do you know who, when, how, they were recorded?  Do you possibly have access to the machine they were recorded on?  The "calibration" between the recording machine and the playback machine is called "interchange". They tried to maintain some level of interchange to allow tapes to be played back on other machines properly.  And parts of the interchange problem would exhibit as time-base errors upon playback.

When I look at Ebay (at least here in USA), I see a great many commercial TBC units for sale at a tiny fraction of their original cost. I see many great candidates for < USD100.  With the availability of discarded surplus commercial gear like that, dunno why anybody would want to fool around with cheap plastic consumer gear?  Of course, those are likely all NTSC, so you would have to research your European marketplace to see if there are similar things available for PAL (assuming that is what you are using?)

Building your own TBC seems extreme just to capture some old tapes. Especially when discarded commercial/broadcast gear of exceptional quality and capability is being tossed in the rubbish.  Besides that TI TVP5160 says it can do only line-by-line TBC and has a limited range of +/- 80 pixels.

[SeanB]

A lot of the commercial units will be both PAL, NTSC and SECAM. Often with the ability to do both format changing to and from each standard with the TBC ability. Often used in news rooms as there they often got a lot of satellite footage in in assorted formats from a worldwide source, plus they got in tape from all sorts of format, from Umatic to Betacam SP and all the consumer formats.

[Wuerstchenhund]

Not sure if this is still an issue but maybe this could be another alternative to an external TBC:

http://www.sensoray.com/products/611.htm

The website blurb says this:

"The Model 611 uses special circuitry to achieve horizontal and subcarrier synchronization to incoming video. It guarantees generation of the required pixel/line from a video source in which line length can vary by several microseconds. This circuitry allows the 611 adapter to quickly and accurately lock on to video signals, regardless of source. Since the technique is completely digital, the 611 can recognize unstable signals caused by VCR head switching and adapt the locking mechanism to accommodate the source."

The thing is made for cameras but maybe that's a good alternative for your use case.

I have a new Sensoray 611 card (bulk) here but I've never used it and I don't have any analog video sources either, so it's essentially for sale.

[TerraHertz]

Pity I didn't see this thread before, and now there are no images showing up in the thread of your video error effects. I used to work in professional video, but with no pics, can't advise.

Also a pity you don't have a proper video waveform monitor and vectorscope. These are fairly cheaply available now analog video is out of fashion. The waveform monitor is *almost* like a scope, but has a lot of composite-video related features a scope lacks. A vectorscope is for analyzing chroma carrier problems, and you can forget about using a scope for these.
With the right tools your problem could be identified in seconds.
Not that saying this is any help to you.

[brabus]

Hello Terrahertz, welcome to the discussion! You can see an example of the issue I am facing on the previous page: Chroma Issue Example
If you don't see it, I can send you the frame via mp.
Wuerstchenhund, thanks a lot for your hint, let's talk about it later and see if it could help.

Guys, I finally obtained the ADVC-55, and the result is: no change.
Exactly the same video artifacts, chroma saturation over the roof, horizontal chromatic lines out of control... On still image, colors flicker like crazy.
The great difference between the cheap USB grabber and the ADVC-55 is the noticeably lower noise floor. So far, a small +1 for the ADVC-55!

This confirms what all of you said: the problem lies upper in the chain.

• Very important to notice: the S-VHS signal produced by the Metz 9877 and the CVBS signal produced by the Panasonic NV-F65 bring the exact same result.
  
• On the TV the video plays flawless.
  
• Therefore, I suppose that my VHS source is OK and I don't need a new VHS player.
  

Now, my preliminary defect analysis:

• I have no horizontal/vertical jitter at all, so this leads me to think that the luma sync signal (i.e. line and frame sync) is OK.
  
• In fact, if I try removing the color after capturing the video, I obtain a nice and clean B/W video, with no noticeable defects. Just the chroma is really tremendous.
  

During my search I found this piece of information, which gave me a lot to think about:

The two major problems are incorrect sync and burst levels that causes increases or decreases in luminance and chroma levels that get worse down the line. Specifically, video recorders. Images get brighter or darker, color levels get higher or lower. Usually the chroma winds up going through the roof because of the poor frequency response of consumer recorders, causing increases in color levels.

Source: VideoHelp Forum, post #13 by videobruce.
I can't figure out how the supposedly limited bandwidth of consumer recorders can affect chroma so much. Yes, the chroma burst from the VHS looked a bit weak if compared to the robust signal provided by the camera, but can this really make so much difference?
If this is the root cause, I guess that TBC is the way to go: rebuilding the chroma burst may solve the issue.

Before proceeding, I would like to collect your opinions. TerraHertz, if you think a video analyzer could help, let's follow that road! Do you have something to suggest?

I just add a quick overview on the european TBC market:

Old professional gear:
   PROs: Overall high quality, low price (200€)
   CONs: Difficult to obtain, mostly comes from USA: ~100€ import taxes + shipping on a 100€ value, aged components, possibly limited capabilities; in general not recommended in the forums (why?)
   Example: interesting JVC gear on eBay

AVT-8710 and rebranded
   PROs: Cheap (250€), picture controls available (saturation, hue, luma, etc.), seems to be the market leader
   CONs: Low quality assembly, reported overheating issues, reported poor performance even with good tapes
   Example: AVT-8710 on eBay

DataVideo TBC-100 / TBC-1000
   PROs: Higher quality, better reported results
   CONs: Difficult to obtain, more expensive (300€), no controls at all
   Example: no examples available at the moment

DataVideo TBC-x000
   PROs: Higher quality, professional equipment, controls available
   CONs: Video distribution amplifier issues, known also on the TBC-1000
   Example: TBC-4000 on Amazon

TvOne 1T-TBC
   PROs: Is there any? Seems identical to the other TBCs, controls and specs are the same.
   CONs: The most expensive on the market (400€ - up to 800€), lack of reviews.
   Example: 1T-TBC on eBay

I am very tempted by the TBC-4000 on Amazon (only 200$!), but... is there any good reason to prefer the inexplicably expensive TvOne 1T-TBC?
Should I get a video analyzer before, and then choose the right TBC?

Thanks a lot.