VCR head output signal?
 

Greetings!

I'm hoping someone can point me towards information pertaining to output signals from common VCR video heads. The more technical the information, the better. Ultimately my goal is to find a close-to-perfect way of digitizing VHS tapes. By capturing the signals from the heads rather than from the VCR's output connectors, information loss is minimized. If the signals are sampled at an appropriate rate, the original tape will (essentially) be stored in digital format on a computer (according to the Nyquist–Shannon sampling theorem). From there, dynamic digital processing can occur (as opposed to static analog processing that generally cannot be fine-tuned). So if anyone has any information to help me out on my quest to modernize this whole process, I would greatly appreciate it! :congrats:

Lots of research and progress here: https://forum.videohelp.com/threads/...-RF-signals%29

All extremely technical stuff that goes completely over my head! :eek:

Note that it will be a long time, if ever, this becomes mainstream because of the knowledge and specialized equipment required.l

For reference, research on direct RF capture from LaserDiscs has been ongoing for years and seems to be functional, but I've never read about anyone outside the project using the technique: https://www.domesday86.com/?page_id=978

I started to loose hope too, As you said the Laser decode has been finalized for quite some time now and we still don't have a product that works outside linux. VHS decode may never see the light of day at all and it's more complicated than LD since you are dealing with 3 signals, luma, chroma and Hi-Fi stereo.

Speaking of weird ways of recovering analog media, I had some e-mail exchanges with this guy and he was thinking of doing his PHD thesis in non contact laser pickup of magnetic media, the way he explained it to me was that a universal player that can transport the tape with a laser scanner, The magnetic field on tape twists the laser beam slightly so one can detect the modulation (Kerr effect).

What exactly is your question? The thread title isn't coherent.

I believe he thinks he can just tap into the RF signal and digitize it like you would do with an audio signal, I don't think he knows the steps involved in turning that RF signal into useful video frames inside the VCR, Or he is a genius and wants to start a project like the VHS decode.

An interesting problem. The signal recorded on tape should be consistent among all VHS tapes (within tolerances established by the VHS specification). Different tape formulation can have somewhat different magnetic properties. The playback VCR then has the responsibility to read the signal recorded on the tape, decode it, and spit it out compliant with the applicable NTSC (or PAL or SECAM) standards. The signal read by the spinning heads includes the basic B&W video signal, the encoded color-under signal, and the two HiFi track signals if present for a total of at 4 modulated RF signals. (You can add to this the linear audio track(s) and control track not red by the spinning heads.)

The signal read off the tape can probably best be read from the output of the head playback amplifiers. and likely would be somewhat specific to the VCR's heads' designs. The VCR electronics would be matched to the heads, and then divine the video and audio signals from their understanding of the head characteristics, and use automatic signal correction (e.g., AGC, AFC) to ensure a standard output. The VCR playback system also interacts with tracking to ensure optimal tracking of the tape.

https://en.wikipedia.org/wiki/VHS#Video_recording provides a somewhat simplified explanation.
The references at the end of that article should provide additional information.

It is a daunting task with I suspect limited commercial market. (Unlike the professional/broadcast formats, VHS and Video8 were the original format for very little valuable intellectual property beyond home video and the market for most home video dies with the participants). A question becomes how much better an end product can DSP of the signal read off the tape be considering the limitations of the original video acquisition system. But it can be a fun project; I wish you well in your undertaking.

Actually the control track is not needed to be captured, it is just used by the VCR to determine the tape format (PAL/NTSC ....) as well as tape speed (SP, LP ...), It's not even used in most modern VCR's to auto-lock on tracking, they use the RF signal feedback itself to fine tune the capstan speed to get the best RF signal level possible (not enough infos to what VCR's use it or not). As to the linear track, it can just be captured with an additional audio ADC on board, it doesn't have to be mixed with the RF signal capture to simplify the process.

So as said above there is no need to start from scratches, such project already exists and if one has deep knowledge in coding he can pickup on the existing project.

VHS is a color-under format. All the video information is stored in one linear signal on tape, but separate.

See: https://www.av-iq.com/avcat/images/d...20-%204611.pdf

LDDecode would have to be modified to up-convert the 629khz chroma information instead of comb filtering a standard NTSC/PAL signal. After that its a matter of NTSC or PAL decoding, which is likely already done software-wise. Hi-Fi audio comes off a separate head on VHS and could be decoded on its own, or you could just record the audio directly as long as you have a way to keep the audio and video synced. NTSC Betamax records the Hi-Fi track with the video, but PAL uses a seperate head/track like VHS does because there was not enough bandwidth to fit it in the video signal.

That's the problem, the decoding and TBCorrectiing of the signal, if it was that easy we would have had a working app already.

Results are already out there with VHS sources: https://forum.videohelp.com/threads/...-RF-signals%29

Fork for VHS: https://github.com/oyvindln/ld-decode

LDDecode does have TBC functions already.

Still working on it, just gong a bit slowly for life reasons.

The way it is going it looks like it might take few years, by the time the interest to digitize tapes will probably fade off.

Yes I've heard of that as well! That would be freaking sweet. You could use it on any magnetic media. Decode the flux vs distance function read off of the tape by the laser to the RF signal by using the specifications of whatever standard was used to record on the magnetic media. I am guessing the signal change is TINY making good measurements difficult (but probably not for someone doing a PHD haha).

That would be especially useful for old open reel format like quadruplex since the machines that played them back are hideously large and complicated and require a lot of calibration.

I read about that Ph.D. thesis as well.
Using the laser to transfer the media from magnetic tape to a digital format.
But the machine would have to be calibrated every time a new tape is involved, wouldn't it?

No, the laser is not going to follow any magnetic tracks, instead it just scans the whole tape width in super high resolution and the software takes care of the rest.