Got in a project for a client. It was a transfer done from a tape recorded in the 1970s by a college radio station. We did not do the transfer and it was sent to us in a digital format. A couple of problems with the transfer and I was hoping that someone on this list could provide some answers as to what may have happened to this transfer. Problem #1 was the the left and right channels were misaligned by 4 ms. Problem # 2 there were spikes at the following frequencies 7.85 KHz, 10.47 KHz, 18.45 KHz and a particularly strong peak at 7.7Khz. We were able to correct all the problems using Wavelab, Izotope RX and Fabfilter Q4. The college radio station is some place I have worked for and with and I know their technical chops are very good. I am wondering what could have happened in the transfer to cause these problems??? Thanks in advance for any and all replies.

Tape head alignment with tone was not carried out before the transfer (my guess).

Or tape is aged and strange chemical processes are happening...

If the original tape is NLA for retransfer I'm not sure what difference it would make trying to guess from the digital file what may have caused this. We can only do our best with the file we've been given.

Because I like to know more and more about what I am doing and how to solve problems like this in the future...

From what you've said over the years you must have transferred more analog audio tapes over the decades than I've had hot dinners.

Didnt you already know that a small play head azimuth misalignment to the tape recording causes a time displacement between left and right channels?

And that failure to align azimuth to the recording is very common especially amongst those not familiar with the older analog tape formats?

On another forum someone suggested a badly messed up A to D converter.

Also 4 ms is a lot for azimuth misalignment. I would also add that the high frequencies are good out to 10 K which would, I think, rule out severe azimuth problems.

There was a unit that I owned that was designed for analog playback of a Sony F1 PCM unit that had a built in delay between the channels.

Who knows what the person doing the transfers used???

To continually do the best job for my clients I am constantly learning about a wide range of subjects connected to my profession. So I ask a lot of question of other professionals in my field like YOU! If you find my questions not to your liking you can block me.

Does it seem a natively stereo or mono recording? Have you compared what it sounds like and what the spectrogram looks like, in both stereo and mono digital playback?

When we adjust play head azimuth to a recording we adjust while listening in stereo and then in summed mono. Sometimes we may need to make a compromise between optimal stereo azimuth and optimal mono (summed) azimuth.

Often it's best to adjust azimuth solely in stereo and leave correcting a time error between left and right until post where in digital it's so easy to do.

But the time error between channels is also the only error we can correct in digital post. Any losses within say the left or right tracks due to azimuth misalignment cant be restored. This digital transfer recording has program up to 10 kHz but how do we know if there wasnt program on the tape even higher up than 10 kHz which was never captured due to say an azimuth error, dirty head, worn head etc?

Perhaps they adjusted azimuth only listening (and viewing) in stereo but somehow it's been summed to mono which especially if it's a full track (or two track dual mono) analog recording can really cause some cancellations.

Yes exactly, and also how they went about it? In the end it may just be speculation on our part.

Thanks for your reply. Much appreciated. This is the type of information that benefits us all.

I am doing some tape and cassette transfers for another client. He has given me the tapes and cassettes he wants transferred so I know what the outcome will be. In the case mentioned above I was only given the digital file and did not do the transfers.

I am always learning even though I have been doing this for over 40 years. Thanks again for taking the time to reply.

The azimuth being out of alignment does not really answer the whole question. What is causing the spikes at the following frequencies 7.85 KHz, 10.47 KHz, 18.45 KHz and a particularly strong peak at 7.7Khz. Most of these spikes are over 6 dB and that could not be caused by head alignment. So the question is still what happened in the transfer??? WE will probably never know but it is still worth discussing.

I'm not sure what you mean by "spikes" and a "peak". It's probably easier to demonstrate them in pictures. Also as I mentioned is it a natively digital stereo or mono file? Can you upload a screenshot of say the RX spectrograph? If it's natively a digital stereo file, upload its stereo spectrogram, plus the same spectrogram after having summed the two channels together.

If it's possible, even better just upload a sample of the audio file?

I am getting some help on another forum. I just want to increase my knowledge so the next time this happens I will know quicker how to fix it.

I will not post pictures because no matter what they are you will find fault with them...

Speaking of azimuth, just chanced upon this closeup shot of a what looks like a very nice custom made and fitted repro head azimuth adjuster for probably a Studer A80. Here's the link to the full YT video.

https://youtu.be/XLguOahEuxE?si=3v9lgT_FCPjlIhhg

Thanks for sharing. I do my two track my transfers off an OTARI MTR-10/2 so the azimuth screws are very accessible. My other transfer deck is an Otari MX5050 BQ II 4 track reel to reel recorder and it would be impossible to add anything like this.

Thanks again!

Such an addition isnt essential but it makes azimuth adjusting faster, safer and easier, especially when transferring a lot of tapes.

PCM-F1 format does have an inter-channel delay, but it's only 11.35 µSec. This is resolved at the analog output (but not at the digital output of the PCM-601). However, many of these Sony units (501, 601, 701, etc.) have proprietary ICs which are aging out and dying. In the process, they are starting to emit random audible spikes. IDK if this is what's happening in your case, but something to consider.

Thanks for the info...That could be.

FWIW

I really wish my client had sent me the original tapes and cassettes but a lot of his material is in someone's "archives" and the various archivist did not want to send me the tapes so they had someone "local" do the transfers. In a lot of the cases they did a very poor job and in two cases the transfers are complete useless due to distortions and spikes in the playback. Grrrrrrrrrrrr!!!!

These fine azimuth adjusters were common on Studer A80 preview decks for lacquer cutting.

The azimuth was off... and this affects both the delay between channels and the high frequency response. This is what happens when you put a cassette into a machine and just let it run without manually riding the azimuth because cassettes are godawful crap.

The whistles could be anything, and if it's a cassette transfer I'll bet it's in the playback machine (but you can tell from the lead-in... if you hear the whistles before the machine is turned on you know it's on playback). I am surprised one of them isn't video sweep frequency since that creeps into everything. Could be a nearby computer monitor, could be a switching supply (especially a badly maintained one).

This is assuming they are consistent. If there are whistles and the frequency varies a lot, first replace the cassette slip sheets and try again.

Send back for proper transfers, especially if they are cassettes.
--scott

Various formats including open reel are vulnerable to azimuth error. The combination of slow tape speed and wider track makes it more likely.