How to fix audio flutter, warble?

[latreche34]

Quote:

Originally Posted by BarryTheCrab

Used multiple capture cards. And the internal speaker of the JVC camcorder clearly put forth the garbage audio.
It’s a lost cause. But it’s great to know so many smart people are willing to help.

If that's the case then it's baked in the tape, proceed with alternative solutions if the contents are important.

[radiokom]

There is tape slipping at pinch roller. So this is not constant wow what can be corrected by software.
And Celemony capstan never worked for me anyway, it correct well constant tone, from 3150 vinyl test record for example).
In this case it could happen because of bad VHS-C shell and or bad adapter.
If it is recorded you have a problem. The first try to replace particular shell with new good shell.

[BarryTheCrab]

Replacing the shell may be the only option remaining, but I think it’s just the format, the recording at SLP/EP, and age.

[fafeco]

Just throwing an idea, I don't know if this could even work and it requires a lot of time and effort both in hardware and software: the control track is recorded by a head placed on the same assembly as the linear audio head. The control track is a known reference waveform recorded by the original VCR, which in your tape should now appear with the same wow and flutter as the audio track. The playback VCR is already using this track to adjust its speed, but may be not using it to its fullest capabilities, which could perhaps be taken advantage of in software to counteract this flutter. I don't know the signal levels and frequency characteristics of the control track, but maybe you could tap into this signal and feed it with appropiate levels to one of the stereo channels of your sound card, and with the other channel record the mono audio output. You would end up with a two channel audio file, where the first track is the control track waveform, and the second track is the audio, both with the same flutter. Then, a software should be developed to analyse this control track against an ideal control track without flutter, extract the flutter characteristics to correct the original control track, and then apply the same correction to the audio to suppress the flutter. I am not aware of any software that exists that can do this, and the other concern I have with this is that the control track is a pulse at 50 Hz (PAL) or 60 Hz (NTSC), which may not be affected by the possibly higher frequency flutter of your recording and thus it would not help in suppressing it.

[timtape]

Quote:

Originally Posted by fafeco

Just throwing an idea, I don't know if this could even work and it requires a lot of time and effort both in hardware and software: the control track is recorded by a head placed on the same assembly as the linear audio head. The control track is a known reference waveform recorded by the original VCR, which in your tape should now appear with the same wow and flutter as the audio track. The playback VCR is already using this track to adjust its speed, but may be not using it to its fullest capabilities, which could perhaps be taken advantage of in software to counteract this flutter. I don't know the signal levels and frequency characteristics of the control track, but maybe you could tap into this signal and feed it with appropiate levels to one of the stereo channels of your sound card, and with the other channel record the mono audio output. You would end up with a two channel audio file, where the first track is the control track waveform, and the second track is the audio, both with the same flutter. Then, a software should be developed to analyse this control track against an ideal control track without flutter, extract the flutter characteristics to correct the original control track, and then apply the same correction to the audio to suppress the flutter. I am not aware of any software that exists that can do this, and the other concern I have with this is that the control track is a pulse at 50 Hz (PAL) or 60 Hz (NTSC), which may not be affected by the possibly higher frequency flutter of your recording and thus it would not help in suppressing it.

Yes I was thinking along the same lines, assuming the flutter occurred at the record stage. As you say the recorded control track should have the same flutter as the linear audio track. This sort of thing is the principle on which Jamie Howarth and John Chester have used in their Plangent Process for many years, although I believe much of their work has used potentially the way more detailed time base information contained in the recorded analog bias frequency which can extend up to 100 kHz or more. But I also remember reading that they have been able to use the much lower frequency audible tones in audio recordings in cases where the bias tone cannot be retrieved, which is quite often, for example in consumer analog audio and video cassette recordings with way lower tape speeds than professional analog audio and video recordings.

A rough measurement I made of the flutter in Barry's sample put it at about 8 Hz, so below the 50 or 60 Hz control track frequency. So maybe the control track would carry enough information to be useful in correcting that sort of flutter frequency, whether partly or fully.


But I know this is non standard practice and would likely run into more expense than the average person would wish to pay for reducing their home movie audio's flutter. I suspect Plangent has been used mainly on high profile audio recordings where profits from a popular artist's remastered album would justify the expense.

The attached spectrum below is of an open reel recording made in the early 50's where the tape slowed in the last minutes of the recording. A loud 50 Hz mains hum plus harmonics was also accidentally recorded which I guess could be used to correct speed to some degree. Interestingly here there is both a short term speed error shown as a cyclical wow or flutter similar to Barry's example, but here also a much slower, long term drift upwards, representing the gradual slowing of the tape when originally recorded.

Thanks for your comments.

[BarryTheCrab]

Ah yes, the old Plangent Process trick…how could I be so dumb…?