Hi all, after 20 years, with just combination of MONEY, CPU power and disk storage (8TB NAS   ) i've just started converting my old video tape with analog video adapters.
Because my VHS and Video2000 VCRs use only composite video (no S-VHS, RGB or component), i've no other solution than using common yellow RCA cable for video transmission.
Now starts issues: video degradation is expected (30+ years old tape might have lost some magnetization), but while in TV all looks fine, on my grabber usb cards (i have many, from different years, all with same issue), i usually see discolored rows in upper part of screen (see attacked file, sky is blue in TV).

Do anyone have any clue to solve this issue?
Macrovision isn't expected as i'm not transfering copyright protected tapes.

I'm not an expert in this field but could this be the good old ground loop problem?

It could be the ground loop, but as well the tape stretched on its side.
The ground loop I would expect more on an older analog TV, where the stripe will slowly drift over the screen (as the phase shifts between vertical andmains), plus most likely there would be two stripes on the screen half vertical distance from each other and not close to each other fixed on top or bottom of the screen.
That cause the color carrier to shift in frequency at the top or bottom of the picture (depends which tape side is stretched), so causes incomplete synchronization so the phase shift. CRT TV's used to have rather fast settling analog PLL, designed to coop with the disturbance coming from the vertical sync pulse. The digital PLL in the computer capture software (part of the device driver, the HW is just a "stupid" ADC sampling the signal as is) uses more advanced gated PLL technique (would be very complicated in analog circuitry), allowing it to be slow in response to mainly noise or other similar disturbance.
The thing is, these advanced processing methods are designed with an assumption of the signal being generated accurately and so assume all deviations to come from additive disturbance. But as the consequence they can not respond to things like drifting frequencies over the picture.
If that is the case, try to look for variants od the device driver for the capture HW you have, I would expect you won't be the first one suffering from distortions coming from deformed tape.
Or the problem could be the ADC using just marginal sample rate while being synchronized on the color carrier, and because the frequency drift due to the deformed tape it does not sample in the correct phase relative to the color carrier (it has to be in 0, 90, 180 and 270deg points, to get the color differences as differences beiween 0vs180 and 90vs270.

By the way what color modulation are you using? I would expect these problems mainly in NTSC. With PAL these effects should be greatly suppresed by the phase alternation (the color shift distortion of one line will be to a big extend cancelled out by the opposite shift in the next line, as eyes (or even the SW) average them out due to their lower clor resolution.
SECAM (uses FM for color and not QAM as NTSC and PAL) could be directly confused by the frequency shifts, but I'm not sure if that could be ever recorded on VHS...

Hi all, thank for advices.
I'm using an debian machine with this https://www.amazon.it/gp/product/B0772GP4K8/ref=oh_aui_detailpage_o05_s00?ie=UTF8&psc=1 video grabber.
New (20 years old) VHS-C tapes are ok, but some older have this issue; i've not tryed 30 year old Video2000 tapes.
I'm using ffmpeg to convert analog signal (PAL standard, 720x576), uvcvideo drivers.

I'm not sure it's an ground loop, maybe it is tape issue.
I don't have any other VHS player.

I couldn’t say exactly what issues you’re having specifically but there is very little to no difference between composite and y/c, even SDI if there is quality equipment throughout the signal chain. Since it looks like you have vindicated the signal source and cabling I would look towards your capture and encoding setup.   I’ve always had good results with the blackmagic design h.264 pro recorder as it was designed for this purpose having all the processing built into the unit itself. Since I work strictly with NTSC systems I can’t share any personal experience with PAL, however I would believe it to work well in that situation as well. They are expensive, around $400-500, but if you can front the money and get a good deal on one you can turn around and re-sell it to get your money back when you’re done.

...Since I work strictly with NTSC systems I can’t share any personal experience with PAL,...

In theory the PAL should be way less sensitive so give better picture than NTSC due to the chopping (the phase related distortions are supposed to cancel out, at least in the eye or during the consequent JPEG compression after digitized), but that may be effective only to some extend...

Are all tapes behaving the same way?
Or if you inspect the tape itself, are there visible any scratches/worn out stripes close its edge?
The luminance signal is FM modulated before recording, but the color is recorded as is. So when the tape gets damaged and the noise increases, the FM may show its strengths in recovering the signal (the BW component), but the AM color information remains noisy.

Or you may inspect the tape head drum (now I'm talking about the VHS, the 2000 system I do not know in that much detail), if there is not any damage or so (I would expect some speck on the stator part, at or near the edge towards the rotary part, causing e.g. the tape being "lifted up" in that section, causing the signal problem).

We have the VHS machine hooked up in such a way that we can put the output through the DVDR machine.

Are all tapes behaving the same way?
Or if you inspect the tape itself, are there visible any scratches/worn out stripes close its edge?

Or you may inspect the tape head drum (now I'm talking about the VHS, the 2000 system I do not know in that much detail), if there is not any damage or so (I would expect some speck on the stator part, at or near the edge towards the rotary part, causing e.g. the tape being "lifted up" in that section, causing the signal problem).

Not all tapes act as this, in previous post i wrote about older tapes (25 and more years old).
Tape head drum maybe ok, otherwise i would have a worse issues than this.
For Video2000, drum is ok: because it's dynamic track follower (this kind of tapes don't need synchro track: every information about tracking is stored with video track), little issue with piezo video head actuator or head itself will result in loss of video and tracking issues (also piezo actuator pantographs), i don't dare about messing with such of rarity, because Video2000 tapes were recorded with different machine (i tried to use, but never worked), issues maybe caused by tape degradation (one newest tape have Pink Floyd live in Venice on it ) with original machine issues

589, thank for your advice, your suggestion is useful, but this piece of hardware cost too much for my use (professional transfer services from tape to DVD are quoted at 8€/tape, i don't event have 60 tape to get return of investment).

The head tracking is working, because the rest of the screen is OK. It is the same head picking up the complete half-frame, so with anything wrong with the head or its guidance, the whole screen would be affected.
I was thinking more about some peck of a dirt or something similar lifting up the tape edge away from the head, so the head then gets weaker signal.
Or the tapes being jus damaged on the edge. That will lead to a bit different distortion with each tape, but still may be very similar, as the tapes may have been damaged by the same defective player. Some line would be visible along the tape...

Thank you Medved, i'll try to inspect tapes to see damage.

In those days i'll get my (first) oscilloscope, i'll try to see signal levels (can weak color burst cause issue on color?)

(can weak color burst cause issue on color?)

Yes, the synchronization of the color carrier is then more affected by the noise, so then the demodulation ends up off phase, so with some color shift. Plus on VCR should be some amplitude AGC to match the color depth (AM, so varies with signal strength) with the frequency modulated luminance signal (does not depend on the pickup signal strength, because of how the FM works). So this will boost up the color signal if weak, but it will boost the noise too. You will see how the burst looks like, how stable it is (you will see it after the AGC)...

Hi, i've just tested oscilloscope on video signal, triggering on row sync pulse.
I've tested without 75Ω load.
Chroma burst seems to be unstable 

What the "unstable" means? Unstable in amplitude or in phase (or both)?


With some digitizers there could be some other variant of the device driver/software with faster PLL (or settings in the sftware you have), what may make it able to follow the phase "shaking", so be able to better recover the color signal.
This type of distortion is rather common with worn/damaged tapes, so I would expect there would be some SW variants designed to better coop with these.

How these records play using another machine?

And another idea:
I understood when displayed on an oder CRT TV, it plays well. What about tapping the decoded RGB signals and digitize from these? If possible, of course...

Unstable, in amplitude.

I don't have any other VHS in my home, not to mention Video2000 (discontinued in 1985).

I don't have any crt tv in my home, all lcd; i can't even record RGB signals, because both cards accept only S-Video or CVSB.

About software, both Windows an Linux version works the same (but Linux seems have better picture quality).