Adjusting a slow tape

[backwhereubelong]

I need some advice, helpful hints if you will, on how to adjust the running speed of some vintage cassettes I have transferred.  Some of them has obvious pitch issues.  Where's the best point of attack on fixing this issue.  I use a high-end tape deck, transfer at 24/96 resolution and I use a separate deck to "unpack" the tape by fast forwarding/rewinding it completely before I start the transfers.  The two main programs I use are Audition3 and iZotope RX, so advice using one or both of those would be appreciated.  Also,
is it best to adjust each track separately or the whole show at once?  Thanks in advance.

[yousef]

I saw an amazing piece of software reviewed a couple of years ago that was aimed at correcting even variable speed variations. I can't remember the name but I think it cost something like UK£4000 

[morst]

This is not a software solution, but the big-time professional answer to this is Plangent Processes - they clock the speed of the analog tape by digitizing at such a high sample rate that they can analyze the tape bias frequency, and clock-lock that in order to compensate for the original recorder(s) speed variations!

Simply amazing concept. I know they charge money for it. 
http://www.plangentprocesses.com/index_save2.htm

Also,
is it best to adjust each track separately or the whole show at once? 

I'd always look for a Whole Show At Once solution.

[bombdiggity]

This is a simple function in Audition. 

The not at all simple part is to determine the proper pitch.  If you're not musically trained (with the ear to correctly identify pitch) you're just pissing in the wind...  A fair percentage of people can tell something is not in tune but comparatively few can actually determine the offset that brings it completely on key...  If you can't hear a few seconds of a song and know what key it was played in you shouldn't attempt to apply an offset to a source. 

First you'd need to investigate to determine what key a few representative songs on the source were played in.  The best approach to then get the source playing on point is to play a keyboard at the proper key along with the source, adjusting it until you have the song exactly in the correct key as indicated by the keyboard (since keyboards are by default always in tune in whatever key you're playing).  Hopefully the tape stays stable and consistent (many don't). 

Sometimes if all the tapes are masters made on the same recorder and played back through the same chain there is a fixed offset and you can fairly safely apply that to all the like masters transferred if you start to find that.  It is rarely so simple though. 

Essentially no two tape recorders run at the same speed without being calibrated so every copy or every master played back in a different player than it was recorded in will be off speed.  Hopefully the difference is small and subtle...  The best ear on adjusting tape speed hands down has said to me more than once that if the difference is less than 10 cents it's not worth doing anything with.  More than that starts to be noticeable.  Once it is into the 30 cent range or greater many people may perceive something is wrong...  100 cents and it is in tune but in the wrong key. 

In Audition the best process is in Effects > Time and Pitch > Pitch Bender (apply a consistent offset at both ends of the line there).  Theoretically there a different process in there that can identify the pitch and offset but it will give you multiple different reads on different points in the same source so the low end automated processes are not at all useful (I don't know if any of the high end ones are). 

The automated tool for fixing warbling (varying speed) tapes is called Capstan (there's a thread here).  It's expensive but does work wonders.  The output isn't perfect but it can find something tolerable from what is completely unlistenable.   

[morst]

If you can't hear a few seconds of a song and know what key it was played in you shouldn't attempt to apply an offset to a source. 

I don't have perfect pitch, but I can adjust a recording so it's in the same key, and in tune with another recording, for example the studio version of the song!?

I must listen to both and switch back and forth to achieve this, but it's not overly difficult if you've ever tuned a guitar or a wind instrument. But yeah, without ANY musical training, I'm not sure it would be so simple. 

But yeah, fixing warble is not something I've ever dealt with, though I guess it might make some of my old cassettes into potential matrix material, if the speed could just hold still!!?? Capstan, eh? I'll look that up.

http://www.celemony.com/en/capstan
This is exciting! Capstan is pretty cool! $$$ 

[bombdiggity]

If you can't hear a few seconds of a song and know what key it was played in you shouldn't attempt to apply an offset to a source. 

I don't have perfect pitch, but I can adjust a recording so it's in the same key, and in tune with another recording, for example the studio version of the song!?

I must listen to both and switch back and forth to achieve this, but it's not overly difficult if you've ever tuned a guitar or a wind instrument. But yeah, without ANY musical training, I'm not sure it would be so simple. 

^ This is a logical approach and will work if your assessment switching back and forth is accurate.  Simple for some but a lot of people are not trained enough to do this effectively... 

[morst]

Simple for some but a lot of people are not trained enough to do this effectively...

And according to this NIH website, a small proportion of the population is actually tone-deaf by nature!

Take their test and find out if it's you!?

https://www.nidcd.nih.gov/tunestest/test-your-sense-pitch

Want to test your own sense of pitch? We've developed an online version of the Distorted Tunes Test, a standardized survey in use for over 50 years. In it, you'll listen to a series of snippets from well-known tunes—some of which have been distorted by changing various notes' pitch. Your task is to pick out the incorrectly played tunes.

[bombdiggity]

^ Cool.  I sort of think that actually tests a somewhat different application of the facility than adjusting tape speed involves though it is ultimately getting at the same skill.   

[morst]

^ Cool.  I sort of think that actually tests a somewhat different application of the facility than adjusting tape speed involves though it is ultimately getting at the same skill.   

yeah, it's a little different, but if you can recognize the wrong notes, you probably have the biological ears needed to fix tape speed, with practice!