"Recording the Beatles" book + Vari-speed question

[Benmrx]

So, about a month ago I got the book Recording the Beatles.......best $100 I ever spent. The book is absolutely amazing, and I highly recommend it.

Here's my question for anyone that happens to know. In the book there's alot of talk on the use of vari-speed in all kinds of applications. In the book it mentions that the tape machines ran at a regular speed of 50 cycles. I'm curious as to how this translates to the more modern 15 IPS and/or 30 IPS settings found on most tape machines...like my Otari MTR90 mkII.

In the book they point out quite a few specific moments where they used vari-speed, and for example it might say (don't have the book right here): On the song "I'm Only Sleeping" that the rythem track was recorded at 52 cycles per second, so that when they played the machine at regular speed, the rythem track was slightly slowed down, and dropped in pitch, making it sound "dreamy" with a kind of slow motion feel.

If I was recording at 15 IPS (which I usually do) would I slide the vare-speed up to 16? Use your ears....etc., I know, and I do, but I'm just really curious if there's a way to find out what 50 cycles means in regards to 15 or 30 IPS.

Thanks in advance.

[max cooper]

50 cycles/Hz is the frequency of the alternating current. We have 60 Hz in the U.S.

When they synced two machines on "A Day in the Life" they striped one machine with the 50 Hz and drove the other machine with it (I think.)

I'm guessing a tape speed that correlates to 52 Hz if the norm is 50 Hz is 104% faster than normal.

15.6 IPS for 15 IPS

31.2 IPS for 30 IPS

but whaddo I know?

Don't you freaking people have Pro Tools?

[joelpatterson]

What's that phrase? "Bad workman blames his protools"?

[theblue1]

Originally Posted by max cooper

50 cycles/Hz is the frequency of the alternating current. We have 60 Hz in the U.S.

When they synced two machines on "A Day in the Life" they striped one machine with the 50 Hz and drove the other machine with it (I think.)

I'm guessing a tape speed that correlates to 52 Hz if the norm is 50 Hz is 104% faster than normal.

15.6 IPS for 15 IPS

31.2 IPS for 30 IPS

but whaddo I know?

Don't you freaking people have Pro Tools?

Max is more or less right. 50 Hz (cycles per second) is the standard for Brit/Euro AC. (And 220 volts still, yeah?)

They didn't stripe the deck -- 'cause they didn't have synch back then (by and large).

Since the tape decks ran off hysteris synchronous motors (which means rotational speed is tied to the frequency of the current driving them), they presumably used a variable speed oscillator in the power supply to apply a form of varispeed. It may well have been built into the machine. Higher frequency increases tape speed -- and of course pitch. So if you want to cut a vocal that will be "pitch shifted" downward in the final, you would record the overdub with the playback deck going the appropriate amount faster than normal so that when the track is played back at normal speed, the new part will be repro'd at a slower speed than it was recorded.

It's often called "formant shifting" in the world of DSP -- but in various forms, it was the source of all those amphetamine gazelle solos from Les Paul (who could already play ridiculously fast) as well as the chipmonk's voices in the famous David Seville Singing Chimpmonks recordings.

[max cooper]

Originally Posted by theblue1

They didn't stripe the deck -- 'cause they didn't have synch back then (by and large).

Pretty sure one four-track had a 50Hz tone on one track and that was used to drive the other.

Or maybe they used Pro Tools.

[thenoodle]

Ask Ken Scott since he hangs out at Hoffman's board a lot. He was there and worked those machines so he'd probably explain it pretty clearly.

My take on things from the Lewisohn book is that varispeeding was used all the time in ways that would just multiply. Like Penny Lane for example. They'd fill four tracks on one machine, varispeeding any one or several of the takes while overdubbing. Then maybe set the machine to play back the four tracks in some random speed (or not)...while piping the submix over to a track or two on the second four track (the reduction mixes as they called them)...while THAT machine might have also been taking the dub while in a non-standard speed. It must've been just crazy. Cool, but crazy stuff.

[RadioMoo]

Originally Posted by max cooper

Pretty sure one four-track had a 50Hz tone on one track and that was used to drive the other.

Or maybe they used Pro Tools.

Would have to have been Pro Tools because you couldn't stripe a tape deck in 1967. Well you could, but you'd be all dressed up with nowhere to go.

Actually. in Lewisohn's book, Sir George explains that it was done on the fly.

Pity the fly.

[Benmrx]

Thanks for the replies. And Max....yes I do have Pro Tools....LE that is. I just love the idea of vari-speed effects in the analog world though.

Tomorrow I'm gonna go down and record just a flat 440Hz A note, and then strap a tuner on the output of the channel, and varispeed the machine up and down and find out where exactly a semitone up or down is in relation to the speed. I'd love to be able to speed up or slow down the machine by a perfect fifth or something.

Any other info people have on vari-speed in the analog world would be much appreciated.

[max cooper]

Originally Posted by Benmrx

Thanks for the replies. And Max....yes I do have Pro Tools....LE that is.

Actually, kidding on the Pro Tools remark.

But I thought I remembered this was mentioned in Andy Babiuk's book Beatles Gear, so I looked it up.

Here's what it says:

(referring to "A Day in the Life")

"Some time during the morning before that session George Martin tackled me as I was walking into one of the studio restaurants," remembers Ken Townsend. "He said, 'Oh, Ken, I've got a little problem tonight. I desperately need eight tracks, but of course we only have four-track machines. Couldn't we lock two machines together?' I said I'd think about it."

No doubt finishing his cup of tea first, Townsend went off and tried to find a way of running two four-track tape recorders synchronised to one another. His solution was to put a 50 Hz tone on one track of the first machine - the one with the existing piece mixed into stereo on two tracks - and use the tone to drive the capstan motor of the second machine - the one that would record the three or four individual mono takes of the orchestra freaking out. "We marked up the starting points on each tape," explains Townsend, "so that both machines could be started at exactly the same point."

Townsend says the only hitch in the scheme came later, when mixing the song. "I remember George [Martin] getting slightly impatient about the fact that the tape machine in Studio 3, where he was mixing, would not lock up exactly in sync. Eventually we had to bring down the same two J-37s that had been used for the recording in Studio 1, because they had a slightly different lead-in time as they ran up to the right speed. That was the only problem," says Townsend. "On the session itself, the link-up worked fine. I think it's probably one of the reasons why 'Day in the Life' sounds like it does, because the orchestra tracks are not exactly in sync. They're probably only a millisecond or so out, but that's enough to make it sound stronger."

[RadioMoo]

Originally Posted by max cooper

"We marked up the starting points on each tape," explains Townsend, "so that both machines could be started at exactly the same point."

OK. So I guess what Sir George was talking about in the Lewisohn book was that there was no chase/lock at that time.

[dbbubba]

People have used Vari-Spped for ODs for years.
It was pretty common to speed up or slow down the multitrack for GTR, vocal or what-ever ODs for years. It's how you get that Dimebag Darryl tone or countless other double tracked GTR tones.
People have used it for BG vocal ODs for years, too.
They did it in radio ID and jingle vocal group sings back in the early '60s.
The first guy that asked me to do it was te guy that INVENTED radio/TV jingles (that'd be Bill Meek.) No, he's not realated to Joe Meek.... he pre-dates him.
The Beatles and Abbey Road poeple didn't invent it.

Syncing decks using SMPTE vs. the more crude 50hz / 60hz PILOT TONE sync isn't really all that different.
If you were/are using a non crystal controlled or AC capstan motor it derives it's speed from the mains freq. (50 hz over there/ 60 hz over here.)
All you needed was an oscilator that would vary in freq. around 50 or 60 hz and a power amp. That would feed the AC capstan motor. You can't vary it too much before it runs wild though.

If you recorded PILOT TONE on a tape track of one machine it could drive another deck like they did at Abbey Road, but IT WAS ONLY CONTROLLING THE SPEED WHICH MEANT THAT GETTING THE START TIME TO BE THE SAME ON BOTH DECKS AT PLAYBACK TIME WAS TOTALLY A MANUAL PROCESS. In other words... the decks were running the same speed during the recording, but since both decks started at seperate moments there was nothing that told the decks to START together. It would be diffficult.

Also, remember that SMPTE wasn't really designed to sync decks. It was a timing code used in military missle tracking that was adopted to control video editing systems that used two video decks.
Since those decks were Ampex 2" quad-head video decks and were pretty much the same technology as Ampex's analog transports it was easy to use it to controll analog decks. Ampex transports in both analog and video decks are EXTREMELY similar in design.

You have to remeber that even with a pair of analog decks sync'ed the "modern" way with SMPTE still required a signal to operate the transport. SMPTE and Pilot Tone both only control relative tape speed. There are "tally" signals that send the STOP, PLAY, FAST FORWARD, RWEIND info and there is a tape position sensor that know where in the reel of tape you are. It just coounts revolutions of a roller/guide or reel revolutions, so it is all pretty crude.

If you have ever used SMPTE to sync two analog decks it is NOT an exact science and the transports have to be able to have the two reels of tape in pretty much the same starting place on their own. SMPTE (like PILOT TONE) does not physically position the tape. SMPTE will only chase up to a point. The decks have to be within a few second of each other. The futher apart you park them, the longer it will take the chase deck to catch up or lock. Too far and it either won't lock before the song starts or it won't lock at all.

Come to thin of it... ADATS and DA-88s operate pretty much the same way. They only use the digital ref. claock to accurately sync the tapes, but the postioning (stop, play, FF, rewind, locate) is pretty much like the syustem used in a SMPTE sync set up.

So, the hard part for the Abbey Road guys was to have the two tapes sitting in a parked positon that was already closely synced up. They had to do the hard sync and have them in sunc. They had to have two decks that would ramp up to speed from stop pretty much the same. They had to manually do what the tally and tape, position sensors do on a SMPTE system.
They were syncing the playback and the 50 hz PILOT TONE was just keeping the speeds constant for the duration of the recording, so it wouldn't wander out of sync.

Once DC servo decks came along with a crystal ref. freq. you didn't really need the pilot tone to pull this off. You can hand sync two modern analog decks and if you don't take any tape out for edits and the moon is exactly right they will stay in fairly good sync for up to three minutes. I could hand sync my JH24 and my Ampex 102 and they would run sync that was close enough for music for almost three minutes. I used to drop in stuff all the time this way. You've heard a lot of record over the years where the BG vocals were droped in this way, too.

THE STARTING WAS THE HARD PART!

By the way.... Abbey Road and Ken Townsend didn't invent PILOT TONE.
It had already been used to sync film audio to picture for years.
They were the first guys to use it in the studio on a REALLY famous ROCK GROUP.
It all becomes mythical don't it?

[Lave]

I saw an interview with George Martin a lot of years ago where he explained that the very first varispeedsessions with the Beatles was executed in a mechanical way.
They simply thread condomes over the capstan.
Then we have to talk about rubber thickness rather than electrical percentage.
I haven´t read the mentioned book.

[kriSh]

great photographs in that book! i really enjoyed looking through it.. sorry for hijacking the thread, but i'm also into vintage photography

[BrianK]

<<Abbey Road and Ken Townsend didn't invent PILOT TONE. It had already been used to sync film audio to picture for years.>>

Yes, MOST of what the Beatles are credited with "doing first" is mythology. In some cases, the people thought they had done it first because they had never known anyone else to do it, in their world. Still, it's a bold thing to come up with something without outside input.

Our Recording the Beatles book certainly clarifies these credits, rather than promoting the old legends. (BTW - it answers almost all the practical questions posed in posts above). It's better to get the bigger picture and realize there were people like Les Paul and Joe Meek (and others even at Abbey Road) doing many of the things the Beatles later made known.

(I love vintage photography too!)

[PhatStax Drums]

Just a brief digression...the "Pro Tools in `67" or whatever joke reminds me of what a co-worker said back in 1999 when we were anticipating the huge Y2K meltdown:

"Well, what did they do in 1900?"

I said they had to re-light all the gas lamps and things like that (well, they did have electricity, but of course nothing with a CPU...)

[walsh350]

L

Originally Posted by Lave

I saw an interview with George Martin a lot of years ago where he explained that the very first varispeedsessions with the Beatles was executed in a mechanical way.
They simply thread condomes over the capstan.
Then we have to talk about rubber thickness rather than electrical percentage.
I haven´t read the mentioned book.

I can easily believe it would have been acchieved like this
I mean they did stuff like making tape loops and holding the tape away from the erase head with a pencil so they could record doubles on top add the natural wow and flutter you get by varying how far you move the pencil ontop of the doubling effect which would also be approaching saturation with each new recording pass and presto. I doubt they even thought of it in these terms. I've physically done this myself with an old B77 for shits and giggleslol.


Really don't think they used much specialised kit for their fx, humbling when you consider what they did with the technology they had at the time.

Wonder what mammoth number of plugin I'll get Upto on my next project, ha!

[Sigma]

Originally Posted by dbbubba

People have used Vari-Spped for ODs for years.
It was pretty common to speed up or slow down the multitrack for GTR, vocal or what-ever ODs for years. It's how you get that Dimebag Darryl tone or countless other double tracked GTR tones.
People have used it for BG vocal ODs for years, too.
They did it in radio ID and jingle vocal group sings back in the early '60s.
The first guy that asked me to do it was te guy that INVENTED radio/TV jingles (that'd be Bill Meek.) No, he's not realated to Joe Meek.... he pre-dates him.
The Beatles and Abbey Road poeple didn't invent it.

Syncing decks using SMPTE vs. the more crude 50hz / 60hz PILOT TONE sync isn't really all that different.
If you were/are using a non crystal controlled or AC capstan motor it derives it's speed from the mains freq. (50 hz over there/ 60 hz over here.)
All you needed was an oscilator that would vary in freq. around 50 or 60 hz and a power amp. That would feed the AC capstan motor. You can't vary it too much before it runs wild though.

If you recorded PILOT TONE on a tape track of one machine it could drive another deck like they did at Abbey Road, but IT WAS ONLY CONTROLLING THE SPEED WHICH MEANT THAT GETTING THE START TIME TO BE THE SAME ON BOTH DECKS AT PLAYBACK TIME WAS TOTALLY A MANUAL PROCESS. In other words... the decks were running the same speed during the recording, but since both decks started at seperate moments there was nothing that told the decks to START together. It would be diffficult.

Also, remember that SMPTE wasn't really designed to sync decks. It was a timing code used in military missle tracking that was adopted to control video editing systems that used two video decks.
Since those decks were Ampex 2" quad-head video decks and were pretty much the same technology as Ampex's analog transports it was easy to use it to controll analog decks. Ampex transports in both analog and video decks are EXTREMELY similar in design.

You have to remeber that even with a pair of analog decks sync'ed the "modern" way with SMPTE still required a signal to operate the transport. SMPTE and Pilot Tone both only control relative tape speed. There are "tally" signals that send the STOP, PLAY, FAST FORWARD, RWEIND info and there is a tape position sensor that know where in the reel of tape you are. It just coounts revolutions of a roller/guide or reel revolutions, so it is all pretty crude.

If you have ever used SMPTE to sync two analog decks it is NOT an exact science and the transports have to be able to have the two reels of tape in pretty much the same starting place on their own. SMPTE (like PILOT TONE) does not physically position the tape. SMPTE will only chase up to a point. The decks have to be within a few second of each other. The futher apart you park them, the longer it will take the chase deck to catch up or lock. Too far and it either won't lock before the song starts or it won't lock at all.

Come to thin of it... ADATS and DA-88s operate pretty much the same way. They only use the digital ref. claock to accurately sync the tapes, but the postioning (stop, play, FF, rewind, locate) is pretty much like the syustem used in a SMPTE sync set up.

So, the hard part for the Abbey Road guys was to have the two tapes sitting in a parked positon that was already closely synced up. They had to do the hard sync and have them in sunc. They had to have two decks that would ramp up to speed from stop pretty much the same. They had to manually do what the tally and tape, position sensors do on a SMPTE system.
They were syncing the playback and the 50 hz PILOT TONE was just keeping the speeds constant for the duration of the recording, so it wouldn't wander out of sync.

Once DC servo decks came along with a crystal ref. freq. you didn't really need the pilot tone to pull this off. You can hand sync two modern analog decks and if you don't take any tape out for edits and the moon is exactly right they will stay in fairly good sync for up to three minutes. I could hand sync my JH24 and my Ampex 102 and they would run sync that was close enough for music for almost three minutes. I used to drop in stuff all the time this way. You've heard a lot of record over the years where the BG vocals were droped in this way, too.

THE STARTING WAS THE HARD PART!

By the way.... Abbey Road and Ken Townsend didn't invent PILOT TONE.
It had already been used to sync film audio to picture for years.
They were the first guys to use it in the studio on a REALLY famous ROCK GROUP.
It all becomes mythical don't it?

30 seconds to lock for a 3 second ride ..who misses that?

and 'time aligned"??? LOL when they PHASED it was all good lolz!!!

[Kiwi]

Originally Posted by theblue1

...It's often called "formant shifting" in the world of DSP -- but in various forms, it was the source of all those amphetamine gazelle solos from Les Paul (who could already play ridiculously fast) as well as the chipmonk's voices in the famous David Seville Singing Chimpmonks recordings.

I beg to differ slightly. Formants are the natural resonant frequencies of a musical instrument or a singer. No matter what pitch that instrument or singer produces, the formants stay the same and add their particular boost or cut, and this gives the unique character to that voice.

Varispeed raises or lowers every frequency by an equal percentage, and as a result also gives the illusion that the formants have changed.

Digital can do varispeed simply by playing back the data stream at a faster frequency. However, as the project sample rate is usually fixed, to give the effect of varispeed, a sample rate conversion must be performed. For example - if we record a 1kHz sine wave at 48kHz, we could raise the pitch by an octave by playing that exact stream of numbers back at 96kHz. The frequency would double, and the time of playback would be halved. But to get the same effect at 48kHz playback, we would need to shorten the data file by exactly one half, so it would play back in half the time. We would also have to represent all the cycles of the sinewave in this same period of time - so the SRC algorithm has a lot of number crunching to do, and there is a possibility of introducing artifacts.

With most digital 'pitch shift' algorithms, the developers have assumed that we don't want the Time of playback to be stretched, as it normally is with varispeed. They also assume that we don't want this illusion of the formants being shifted either.

This is where the concepts of Time Stretching and Formant shifting have been introduced, and confused together with Pitch Shifting.

To raise Pitch without shortening Time involves 'creating' some filler audio data to stretch it out. This is a much bigger opportunity for adding artifacts. Or if lowering the pitch, we need to throw away valuable audio data in order to avoid the Time being stretched out longer. Also a big source of artifacts and quality loss.

Formant shifting is different again. It involves identifying the formants of the audio (basically the fixed EQ shape) and then creating a different EQ shape that will counteract the effect of the varispeed. I think of Formant Shifting as a fairly aggressive EQ effect.

With modern pitchshifting alorithms, such as Melodyne, you can choice to use these different effects together, or individually.

I sometimes like to use Formant Shifting by itself (without Pitch Shift) as an effect that has some of the qualities of tape varispeed, without the Pitch and Time effects. It can be very cool for drums and stuff.

FWIW.

[BrianK]

I agree, as I understand it, formant shifting will not change the note/frequency of the track, but it will change the resonant overtones. You can shift the formant series (resonant peaks and dips) so the voice singing C# can sound like a BIG body voice or a SMALL body voice.

It's a neat trick on pianos, guitars and other acoustic instruments that have a noticeable "size"; the pitch can be kept the same, but the piano overtones can make it sound 20' wide, or 20" wide - neat stuff! NOT good for classical music tho!

[Kiwi]

Formant shifting is sometimes called 'gender' bending ...