VHS Tape Curly Edge Syndrome, Any Fix ?

[dellsam34]

I have a home video shot with a full size VHS cassette camcorder back in the 90's in PAL format SP, The family member who had the tape played it in different VCR's, When I acquired the tape to capture it the first 7 minutes or so exhibit unstable frames, Upon looking at the tape the only damaged area is the edge of it where the sync pulses are recorded, Obviously this has happened in a miss aligned VCR, The video and linear audio areas are fine (No HiFi), Except when in playback the audio has some sort of wow due to the VCR trying to adjust the timing with video tracks, But the audio track is not damaged.

The VCR's internal LTBC didn't fix the problem (JVC HR-S7600AM) and I know a full frame external TBC is not going to help because the VCR is struggling to lock the drum on the video tracks and it doesn't have the sync pulses needed for the job, So it wouldn't deliver clean video frames for the FFTBC to begin with, I already tried my AVT FFTBC.

I don't have a pro PAL VCR that accepts Genlock signal and buying one from Europe and shipping it to the states it would cost a fortune as those suckers are heavy.

I was thinking of a more invasive way, Playing back another PAL tape with another VCR in the same speed and solder 2 wires on the sync head of this dummy VCR and connect them to the capture VCR sync head after bypassing its own sync head (require cutting the wire to isolate its own head and installing a cutoff switch). While both VCR's playing I try to fiddle with the dummy's VCR pinch roller slowing it down until I got a perfect timing. But I though I'll post here first for better suggestions as I wasn't sure if this is going to work.

What is the less invasive way to recover those few minutes of video?

[JVRaines]

That is an heroic plan and one I doubt will succeed. How bad is the curling? You could try ironing the tape at low heat through a paper barrier, though if it is actually stretched, you are out of luck.

[dellsam34]

That is an heroic plan and one I doubt will succeed. How bad is the curling? You could try ironing the tape at low heat through a paper barrier, though if it is actually stretched, you are out of luck.

The curling is so bad that I don't think the ironing will do much about it.
Technically borrowing sync pulses from another source should work, But it has to be timed precisely and that could be a challenge.

[dellsam34]

No answers means unsolvable problem.

[JVRaines]

If you can't read control pulses from the tape, yeah, you are going to get bad tracking. The capstan and scanner motors will fall back to default speeds and there is little chance the heads will line up properly with the tracks.

[lingyi]

Post over at digitalfaq.com or PM lordsmurf. He'll likely be able to give you an answer, even if it's just yes or no.

BTW, if the tape is curling (i.e. rippled) it's probably stretched to the point where the control pulses are unreadable as JVRaines stated.

[dellsam34]

Post over at digitalfaq.com or PM lordsmurf. He'll likely be able to give you an answer, even if it's just yes or no.

BTW, if the tape is curling (i.e. rippled) it's probably stretched to the point where the control pulses are unreadable as JVRaines stated.

I know, That's a fact that pulses are not readable otherwise I won't be posting here.
Lordsmuf hangs out here too, I found it posting here gets more responses than there as almost every member at digitalfaq is a member here as well.

[lordsmurf]

more responses

Quality over quantity.

I don't envy your problem.
I'd cut my losses on that tape, unless it was the only footage of grandma, or proof of big foot.

[dellsam34]

It’s pretty close to “the only footage of grandma”

[ironwood321]

I'm trying to find the page but there was a probable solution on the web a while back using a similar method with a specific broadcast vcr.
They used two broadcast vcrs one of which in a certain edit mode you could feed a new control track timing signal the tape.
This would sometimes help the LOCT problems but it would depend on timings.

Hi I found the link it might help
http://utz2.com/videoexpert/artic1/237ctl.htm

[JVRaines]

It’s pretty close to “the only footage of grandma”

Another thought is to improve head-to-tape contact during playback. Take a Q-tip and press the lower tape edge against the audio/control head. Maybe you will get enough signal to lock tracking.

[pandy]

If you are desperate and sufficiently rich then you can go for signal processing in software... probably you need to hire some cheap but skilled East Europe coder plus use some HW within few hundreds of U$, if you are lucky and you can use some software like GNU Radio then you should be able to stay way under 10k of U$ .

[dellsam34]

Another thought is to improve head-to-tape contact during playback. Take a Q-tip and press the lower tape edge against the audio/control head. Maybe you will get enough signal to lock tracking.

That's a good idea, I will give it a shot, Maybe a felt pad works better.

If you are desperate and sufficiently rich then you can go for signal processing in software... probably you need to hire some cheap but skilled East Europe coder plus use some HW within few hundreds of U$, if you are lucky and you can use some software like GNU Radio then you should be able to stay way under 10k of U$ .

To do a software solution you have to have correct frame data (2 fields a frame) and to get correct frame data the VCR needs the control pulses to determine what speed the tape is recorded at, PAL, NTSC, SP, LP, EP, SLP, also any speed variation due to shrinkage or stretch of the tape ...etc Once that determined the capstan servo locks on that speed and the drum servo locks on the start of a given field and the video heads start reading fields data.

The kind of problems that can be fixed by software or external TBC/Frame sync are the problems associated with the timing of those fields assuming that they were read entirely, If they are not read by the video heads correctly, In other words if the video heads were reading between the video tracks or reading the track and half way loosing it due to tape speed variation there is no way to fix those problems with an external TBC or software or even the VCR's internal TBC (Which corrects horizontal lines timing and requires the the heads to read video tracks correctly)

My initial idea was to install a control track pulse generator close to the control track head and use a switch selector to select between the head for good tapes or the generator for damaged tapes.

The generator will generate 60Hz for NTSC and 50Hz for PAL and a trim pot is used to compensate for speed variation, Basically watch the frames while turning it until a stable frame is achieved. an optional advance/delay pot can be installed but it is not required for VCR's with auto tracking.

Another way of achieving this is using the mobile app in the attached picture and from the headphone jack use a record head preamp to bring the level down from line level to a magnetic head level, The preamp is installed close to the control track pulse head and the line in wires are routed to the back of the VCR for an input jack. The same switch above can select between the pulse head or the input jack.

[dellsam34]

Since I'm no expert in designing electronic circuit boards I'm going to use off the shelf parts and components according to the diagram below:

[pandy]

To do a software solution you have to have correct frame data (2 fields a frame) and to get correct frame data the VCR needs the control pulses to determine what speed the tape is recorded at, PAL, NTSC, SP, LP, EP, SLP, also any speed variation due to shrinkage or stretch of the tape ...etc Once that determined the capstan servo locks on that speed and the drum servo locks on the start of a given field and the video heads start reading fields data.

The kind of problems that can be fixed by software or external TBC/Frame sync are the problems associated with the timing of those fields assuming that they were read entirely, If they are not read by the video heads correctly, In other words if the video heads were reading between the video tracks or reading the track and half way loosing it due to tape speed variation there is no way to fix those problems with an external TBC or software or even the VCR's internal TBC (Which corrects horizontal lines timing and requires the the heads to read video tracks correctly)

Nope, you need to acquire FM modulated, RAW video head read signal and process it later trough software, so you can address all problems related to servo but also other aspects like HQ software video signal processing (so software NTSC/PAL/SECAM demodualtion). Good area for some niche startup and perhaps scientific work as it is quite interesting for human heritage.

Since I'm no expert in designing electronic circuit boards I'm going to use off the shelf parts and components according to the diagram below:

Your idea may work and i see no bigger issue to try it. IMHO you should search for tracking generator with hold mode so generator with some form of frequency tracking and able to provide last good frequency sync pulses in case of error. Quite trivial from electronics perspective - probably doable in HW (with some 4046PLL) and perhaps better with Arduino board (then i would recommend you to create own code, measure frequency (or rather pulse width as this may be more accurate for such low frequency signal than measuring frequency) and generate own pulses of the last known good frequency (from range for example +-1Hz).
First shot: https://www.engineersgarage.com/contribution/how-measure-frequency-and-duty-cycle-using-arduino
http://www.auctoris.co.uk/2011/05/14/arduino-function-generator-part-1/

[dellsam34]

Nope, you need to acquire FM modulated, RAW video head read signal and process it later trough software, so you can address all problems related to servo but also other aspects like HQ software video signal processing (so software NTSC/PAL/SECAM demodualtion). Good area for some niche startup and perhaps scientific work as it is quite interesting for human heritage.

Correct, But that doesn't apply to this case though, I can't acquire FM modulated video and audio signals, The video head will never align with the video tracks and HiFi audio tracks underneath. It's lacking the control track. So without modifying the hardware (VCR) software doesn't do anything.
Here is a good VCR basics.

Arduino and coding is not for me and I don't think I will have the time to go through the learning curve.

[pandy]

Correct, But that doesn't apply to this case though, I can't acquire FM modulated video and audio signals, The video head will never align with the video tracks and HiFi audio tracks underneath. It's lacking the control track. So without modifying the hardware (VCR) software doesn't do anything.
Here is a good VCR basics.

Arduino and coding is not for me and I don't think I will have the time to go through the learning curve.

Of course you can acquire all signals and perform separation in frequency domain and later perform regular DSP (FM demodulation, color demodulation, etc), also you have access to servo pulses only they are distorted beyond servo capability - head drum rotate, heads reads signal, servo head reading distorted signal - only regular electronics in VCR is not designed for such case. Reading distorted signal in time domain means you need to transpose (shift) it in frequency domain.

[dellsam34]

Of course you can acquire all signals and perform separation in frequency domain and later perform regular DSP (FM demodulation, color demodulation, etc), also you have access to servo pulses only they are distorted beyond servo capability - head drum rotate, heads reads signal, servo head reading distorted signal - only regular electronics in VCR is not designed for such case. Reading distorted signal in time domain means you need to transpose (shift) it in frequency domain.

Yes with a working tape not a defective one. Again without a hardware modification there is nothing you can about it.
No the control track is not distorted it's gone, It's completely damaged beyond repair, That's why I kept telling you software or TBC is not going to work.

[JVRaines]

You can't "acquire all signals" when the heads aren't lined up with their tracks. You will be missing data.

[pandy]

Yes with a working tape not a defective one. Again without a hardware modification there is nothing you can about it.
No the control track is not distorted it's gone, It's completely damaged beyond repair, That's why I kept telling you software or TBC is not going to work.

TBC works in different domain. Servo is designed for particular scenario - can't deal with distorted signal but electronics can.

You can't "acquire all signals" when the heads aren't lined up with their tracks. You will be missing data.

Only with assumption that you perfectly misalign tracks.

[JVRaines]

You can't "perfectly" anything without control track. The helical tracks are not be perfectly spaced. At best, you will get a few whole tracks and lots of pieces of tracks.

[dellsam34]

Pandy, I think you have very superficial knowledge on how the VCR works, Any mod you will do it has to be invasive such as modifying the electronics of the VCR or adding to it which what's I'm aiming for. If the VCR is not reading the fields correctly there is no way in hell that you can fix such noisy signal. What basically you are saying is take an audio cassette player change the audio head position until it looses the audio track completely and feed that noisy hiss to a software and make a perfect audio signal out of it, Or drop a record player stylus between the grooves and for some how you can turn that into sound.

[pandy]

Pandy, I think you have very superficial knowledge on how the VCR works, Any mod you will do it has to be invasive such as modifying the electronics of the VCR or adding to it which what's I'm aiming for. If the VCR is not reading the fields correctly there is no way in hell that you can fix such noisy signal. What basically you are saying is take an audio cassette player change the audio head position until it looses the audio track completely and feed that noisy hiss to a software and make a perfect audio signal out of it, Or drop a record player stylus between the grooves and for some how you can turn that into sound.

Well, not judging your knowledge about how VCR's works and how they are designed. Based on your feedback i can reply only that you completely ignoring fact that video signal is stored on video tape as FM modulated and on audio tape as AM modulated (for hq some HF bias current is used to linearize system characteristic). Don't get me wrong but using external reference signal especially as manual generator you facing exactly same issue as you pointed (compare frequency error you feeding to system with width of tracks).

You can't "perfectly" anything without control track. The helical tracks are not be perfectly spaced. At best, you will get a few whole tracks and lots of pieces of tracks.

Agree to some point (stochastic sampling, statistical signal processing), however we talking about situation where regular VCR servo began to prevent proper signal processing due faulty (corrupted) servo signal.

[dellsam34]

It doesn't matter what the signal recorded on tape is, FM, AM, PWM, Caption, Teletext, Macrovision .... The VCR needs to lock on the video track and without the control track it cannot and therefore there is no signal fed to the $10k software, End of story.

I believe you are confusing the control track pulses (which are used only by the VCR motor servo's) with the vertical field synchronization pulses (which are part of each field video track and are used by TV, external TBC, capture card, the software you are talking about, another recording VCR...etc).

Each frame is made of two fields the odd and even, and each field is recorded in one full track on the video tape by one head, The opposite video head on the drum records/plays the other field track. Each field or video track contains the following :

- First 1-10 horizontal lines is the head switch signal, This is to switch from head A that is at the end of its track to head B that is just at the right angle to start scanning the next field.
- After the 10th line is the vertical synchronization pulse (This pulse also contain the first few horizontal sync lines modulated into it)
- Then the horizontal line synchronization pulses or bursts for the remaining field scan lines (each horizontal line starts with a pulse to move the scanner of a TV from one side to the other side of the screen to scan the next horizontal line, These pulses are also used by the VCR's internal Line TBC (If equipped) to fix line timing by using a digital buffer)
- After each horizontal line sync pulse comes the line video data (chroma, luma ....etc)
.
. 262.5 lines/field for NTSC, 312.5 lines/field for PAL
.
- At the last line video data comes the next head switch pulse to switch from head B back to A and then off course another vertical sync pulse to prepare for the next field and so on and so fourth.

Off course this is just a brief and symbolic explanation of how the mechanics of reading video data off the tape is done.

So when the VCR and the control track recorded at the bottom edge of the video tape are healthy the field data are scanned correctly so even with minor errors to the horizontal and vertical synchronization pulses due to tape drop out, dirt spots on the tape ..etc they can be corrected by a TV, a full frame TBC, Capture card TBC, DVD recorder, an editing VCR, A software ... This I agree 100% with you.

So to summarize:

- Control track pulses on the edge of the tape cannot be fixed if damaged or erased without a modification of the VCR
- Horizontal and vertical synchronization pulses recorded on the video portion of the tape if played by a crappy VCR and/or crappy capture card or physically corrupted due to tape damage can only be fixed to a certain extent by a forensic frame recovery software, good external TBC, a pro mixing console, Good DVD recorder, an editing VCR. Just to name few.

[pandy]

I don't confuse anything and i know how VCR's works internally. You can't ignore that VCR's are designed to deal with mechanical tolerances, temperature drifts etc, whole servo control loop is affected by many factors and without this you will be unable to read tape after while even on same VCR and not able to read tape on different VCR at all... Why? Because servo head is not tightly linked with video head...
Take oscilloscope, some VCR service manual and check how VCR works... is way more robustly designed than you think.
TBC doesn't need to be digital and can be implemented as analog device (CCD/BBD).

[dellsam34]

I'm glad that you understand my point now, Let's move on with the thread.

[pandy]

I'm glad that you understand my point now, Let's move on with the thread.

Well i'm not sure what do you mean by this but anyway from my side this is EOT, wishing you good luck.

[dellsam34]

Ok, Now to be able to test this surgical procedure I will have to get a cheap VCR from goodwill and recreate the damaged tape by erasing the control track at the edge of a donor tape, Probably using a small weak magnet and barely touching the edge while the VCR is playing the tape, When I see the capstan servo going nuts then I know I perfectly duplicated the problem. Then I will rewire the control track head for my new system.

I will report back when I get all the parts.

[JVRaines]

We look forward to your next journal entry, Herr Doktor Frankenstein.

You can probably make a control trackless-recording by putting a piece of tape over the audio/control head.

[dellsam34]

We look forward to your next journal entry, Herr Doktor Frankenstein.

You can probably make a control trackless-recording by putting a piece of tape over the audio/control head.

You know what I don't even need to erase the control track, I will be cutting one of the wires of the control track head anyway, That should do the trick, Unless it triggers the system control processor and gives out an error, In that case I have to go other routes.