Vcr Linear audio interference

[useraxe]

I'm trying to capture linear audio on a vcr, but sometimes the linear head captures some interferences at 25-30kHz range, and mirrors it down to 14kHz making it audible. I have tried to move the vcr and disabling the wifi, doesn't help. And it only artifacts when the tape has audio in it, no artifacting when audio is off/paused, or runs in HiFi.
The only way of removing mirroring at 14kHz is to hold a screwdriver near the head, but that's not a good fix, is it...

Tried a different vcr, almost the same thing.

Has anyone had success with reducing this kind of interference (in hardware)?

[pandy]

HW solution is low pass filter with attenuation at least 50..70dB before your ADC or alternatively increase sampling rate and later resample to desired sample rate.

[useraxe]

I can do LP in software (i don't have that as hardware) but then the audio loses a lot of treble information.. And I already changed 48khz to 96khz, but it makes no difference for the 14kHz artifacts. Will try 192kHz though.

[timtape]

Congratulations on wanting to capture excellent linear audio (few seem to bother) but VHS capture is good for 10 to 12kHz. Not much useful above that. If we can capture all of that with minimal extraneous noise we've done well. Unlike HiFi, linear cant go to 20 kHz and the signal to noise is not nearly as good. We have to make the best of what's there.

A lot of later model VHS decks add significant noise in linear playback, especially high frequencies. I've made some tests. Some of it seems magnetically induced into the playback head, and some can enter via the unshielded cable (!) from the head to the head preamp. Older model as well as (we hope) professional decks used proper shielded cable. Some of the noise can be induced from the video drum motor into the head.

Possibly you have an earth or ground loop problem in your setup. Have you tried disconnecting all video leads, leaving just the audio RCA cable and the (two pin?) mains lead to the VCR?

What make and model VCR(s) are causing the problem?

[useraxe]

I see 14.5kHz peaks in recordings (i know it's not hifi, just correct azimuth), so it's not as easy as using low pass to fix this.

And thanks for bringing up shielding. I'll try to wrap the cheap, thin, flat connector in aluminium foil tomorrow, hopefully the issue gets reduced... Previously, removing the metal cover of the deck did add a little more artifacting, so it might actually help by reflecting interference from outside which I'm certain is the source of the problem.

[timtape]

I see 14.5kHz peaks in recordings (i know it's not hifi, just correct azimuth), so it's not as easy as using low pass to fix this.

Are you sure the tone is on the tape's linear audio track. Did you peak the tone's gain by actually adjusting the A/C head's azimuth screw, monitoring the signal in real time?

And thanks for bringing up shielding. I'll try to wrap the cheap, thin, flat connector in aluminium foil tomorrow, hopefully the issue gets reduced... Previously, removing the metal cover of the deck did add a little more artifacting, so it might actually help by reflecting interference from outside which I'm certain is the source of the problem.

Unless it's grounded to the VCR's chassis with a wire of some sort the foil may make things worse. Without being grounded the foil may just act as an antenna.

[useraxe]

Are you sure the tone is on the tape's linear audio track. Did you peak the tone's gain by actually adjusting the A/C head's azimuth screw, monitoring the signal in real time?

I did adjust it myself, all 3 screws to sound the loudest. Interestingly adjusting the screw on the back side can cause worse interference.

Unless it's grounded to the VCR's chassis with a wire of some sort the foil may make things worse. Without being grounded the foil may just act as an antenna.

I'll try it and report back how it goes, thank you.

[timtape]

Are you sure the tone is on the tape's linear audio track. Did you peak the tone's gain by actually adjusting the A/C head's azimuth screw, monitoring the signal in real time?

I did adjust it myself, all 3 screws to sound the loudest. Interestingly adjusting the screw on the back side can cause worse interference.

Oh dear. Normally we adjust only one screw, the azimuth screw, leaving the others well alone. The others adjust height, and the forward and back tilt. Restoring perpendicularity can be difficult.

[useraxe]

Welp, I couldn't find information on the internet on how to adjust it. Maybe searched the wrong places....

[timtape]

If you type in "Azimuth" and "VCR" you'll probably only get references to "azimuth recording" which relates to the video heads, not the audio/control head. Instead go to an audio tape forum like Tapeheads (use Google search because it has a better search engine than the forum itself) and you'll find many references to head azimuth. Then apply that knowledge to the A/C head in a VCR.

Understanding what we're doing before jumping in is crucial. First learn what azimuth is re the audio control head. Understand the principle. Then we can identify which one is the azimuth screw - no matter which VCR or tape machine we're looking at.

[pandy]

I see 14.5kHz peaks in recordings (i know it's not hifi, just correct azimuth), so it's not as easy as using low pass to fix this.

And thanks for bringing up shielding. I'll try to wrap the cheap, thin, flat connector in aluminium foil tomorrow, hopefully the issue gets reduced... Previously, removing the metal cover of the deck did add a little more artifacting, so it might actually help by reflecting interference from outside which I'm certain is the source of the problem.

This is probably some signal not content related - i don't expect more than 10kHz..12kHz, there is strong peak from horizontal deflection so perhaps there overload in your capture card... normally i would not bother and try to apply low pass filter (your capture card normally should be insensitive for aliasing as it use probably sigma delta analog to digital converter - they oversample signal massively so aliasing is less problematic).

I think it more likely that your capture card detect some cellphone or similar interfering signal source...

[useraxe]

Interesting that some think linear is always bad. Maybe it's because these are PAL tapes they sound better?

I know it looks noisy, but properly denoised, the information is basically restored.

[timtape]

Interesting that some think linear is always bad. Maybe it's because these are PAL tapes they sound better?

Actually since the tape speed is 20% slower in PAL re NTSC, linear audio in PAL is roughly 20% worse.

I know it looks noisy, but properly denoised, the information is basically restored.

It depends on the case. What you probably have there is program which is already intelligible, just a little noisy. The problem comes when the wanted program is more seriously masked by random broadband noise. Cleaning that noise away to reveal the speech or music underneath is, in a word, impossible. Try a sample where the speech is so masked by such noise that it is unintelligible. It's a very different ballgame.

The reason I expend some effort get the audio to play back optimally at the VCR is because I think I have some understanding of the limitations of even the best audio "restoration" software.

The tone on the spectrum display (horizontal orange line) appears to be 15,625 Hz which is a horizontal video signal frequency. Very unlikely to be from the tape.

[useraxe]

Oh, yes of course, I wasn't trying to imply all of the original sound information can be recovered, but theoretically linear can have a lot of audio info which may not be well audible when not restored.

[timtape]

Oh, yes of course, I wasn't trying to imply all of the original sound information can be recovered, but theoretically linear can have a lot of audio info which may not be well audible when not restored.

Well we'd have to compare listening to some actual before-and- after audio examples, not just looking at spectral displays.

To get the ball rolling, here's a couple to illustrate:

LP PAL linear is about as lo-fi as PAL VHS can get. This example audio is just from playing back the tape close to optimally from the VCR. No processing, no restoration. Source: radio station VHS logging tape.

Mixed sample has a voice almost completely masked by noise. Anyone's welcome to try cleaning/restoring it and upload their best result.

[useraxe]

I agree that your LP sample is as good as it can get but, when the treble has more rolloff at higher frequencies (occurs more on SP speeds it seems like), those Can be improved with some EQing and denoising to sound sharper. That's what I was trying to say. Only when the signal to noise ratio is high enough it can be achieved.
It's the same how it's possible to remove grain on a video, but not a hardcoded overlay like subtitles.

[pandy]

Actually since the tape speed is 20% slower in PAL re NTSC, linear audio in PAL is roughly 20% worse.

Not necessarily - usually video tape is better quality (finer magnetic particles, higher coercivity than most common for audio Fe2O3, overall better quality of tape and better transport).

The tone on the spectrum display (horizontal orange line) appears to be 15,625 Hz which is a horizontal video signal frequency. Very unlikely to be from the tape.

Yep, probably from nearby TV (CRT) or from VCR electronics.

[timtape]

Actually since the tape speed is 20% slower in PAL re NTSC, linear audio in PAL is roughly 20% worse.

Not necessarily - usually video tape is better quality (finer magnetic particles, higher coercivity than most common for audio Fe2O3, overall better quality of tape and better transport).

Yes video formats soon adopted the higher performing tapes. That is a given, and with good reason. It allowed smaller form factor cassettes and therefore camcorders, to capture the much larger data stream of both picture and sound, than just sound.

But I was comparing the difference in linear audio quality wrt different linear tape speeds on VHS, controlling obviously for all other factors. But regardless of the tape's performance, usually slower speeds have lower fidelity, as I'm sure you know.

[pandy]

But I was comparing the difference in linear audio quality wrt different linear tape speeds on VHS, controlling obviously for all other factors. But regardless of the tape's performance, usually slower speeds have lower fidelity, as I'm sure you know.

Of course it is lower but i would argue if this is purely outcome of lower linear tape speed, this is probably least important factor - simply for VHS HQ audio was not as much important as possibility to deliver consumer video recording/playing possibility - so audio was just good enough for RF cable usually used for most of consumers to record (aerially/cable broadcasted) source.

[timtape]

But I was comparing the difference in linear audio quality wrt different linear tape speeds on VHS, controlling obviously for all other factors. But regardless of the tape's performance, usually slower speeds have lower fidelity, as I'm sure you know.

Of course it is lower but i would argue if this is purely outcome of lower linear tape speed, this is probably least important factor - simply for VHS HQ audio was not as much important as possibility to deliver consumer video recording/playing possibility - so audio was just good enough for RF cable usually used for most of consumers to record (aerially/cable broadcasted) source.

Whatever the reasons we're stuck with original linear audio format limitations. I'm certainly not blaming anybody for those design limitations. Rather as with the OP, I'm interested in extracting the best out of those audio tracks when digitizing. Unfortunately, when compared to the vision element, when it comes to playback for digitizing, audio tends to get the short straw, which is a shame.

Some of it is just ignorance. This ignorance seems less common in the audio only world, perhaps because when the audio is all we have, we have to make the most of it.

There is a tendency to look down upon slow speed "consumer" audio formats as not worth taking too much trouble over because the sonic final result will probably never win any hifi awards. But that's not the point. As with the picture, the serious digitisation task is to do what we can with what little is there, regardless of whether it's consumer or pro. Generally linear audio as on VHS and Beta can be the most difficult to work with and requires the most skill and best equipment because we're dealing with modest audio quality at best but terrible quality when handled badly.