3 3/4 IPS analogue tape question

[MessianicDreams]

Hi guys, not too sure which forum to post this in, but here goes anyway:



A client has asked us to digitize a bunch of 1/4" tape. The only issue is the tapes was recorded at 3 3/4 IPS, and our Otari MTR-12 only goes down to 7.5 IPS. What I've done as a test is recorded the audio in at 88.2kHz and played it back at 44.1kHz, there-by halving the playback speed.

Is there any issues in doing this? I'm having a hard time judging if there's much quality loss as I have no idea what the tapes *should* sound like - the mixes on them aren't great (pretty low-fi) and I'm worried how much my method is degrading the audio..

If anyone has any useful tips, pointers or comments, please feel free :)

thanks!


MD

[drtechno]

you night want to find someone with a good 1/4" you could rent/borrow.Specially if you have different head type. If the job pays well enough to justify buying a used one, go ahead. fast dubbing was used in cassette replication, but that always lead to reduced headroom and bandwidth. You might want to use puncher and/or stereo expander in wavelab to restore it if you go the fast OD, but you should be able to find a good process in premastering that will work on all those tapes, specially if they were recorded on the same machine.

[MessianicDreams]

Hi DrTechno, thanks for your reply.

Unfortunately, buying another 1/4" machine wouldn't be possible for this. For most applications, our current machine works fine.

So fast dubbing will result in bandwidth and headroom loss? How severe is this?

[Matt Syson]

Hi
Unless the tapes were recorded on a really well lined up machine in the first place and they are on good tape then the response will be pretty poor anyway.
No 'professional' would have recorded this slow unless it was only a 'rough' or whatever the term is for it.
Make sure your machine is as well lined up as possible and go for it as you are.
Are you sure the head gap and spacings are correct?
It may be worth cross posting to answer whether the equalisation for the two tape speeds will be 'compatible' while doing this transcription.
Matt S

[MessianicDreams]

Hi Matt,

Yours is a name I know well - for a period I worked at the Motor Museum in Liverpool (I started just after the 8232 was put in, and left just before the "second coming" of Mike C!). Thanks for your help on this matter.

Originally Posted by Matt Syson

Are you sure the head gap and spacings are correct?
It may be worth cross posting to answer whether the equalisation for the two tape speeds will be 'compatible' while doing this transcription.
Matt S

I'm not sure what you mean by "head gap and spacings" being correct - how would I know if they weren't?

The tapes actually have 2 mono tracks on them, rather than a stereo mix.

[cjmnash]

a better play may be to record it by playing back at 7.5ips, then play it back out of your DAW at half speed, not by sample rate conversion, but by a quick key. Pro Tools, for instance, does half-speed playback by hitting shift+spacebar.

-chris mara
www.welcometo1979.com

[MessianicDreams]

Originally Posted by cjmnash

a better play may be to record it by playing back at 7.5ips, then play it back out of your DAW at half speed, not by sample rate conversion, but by a quick key. Pro Tools, for instance, does half-speed playback by hitting shift+spacebar.

-chris mara
Welcome to 1979 on MySpace Music - Free Streaming MP3s, Pictures & Music Videos

I thought about that, but wouldn't that effectively mean that the sample rate would be 22.05kHz? Playing back 44.1kHz at half speed...I think playing back 88.2kHz at a 44.1kHz rate would be better no? Or is my reasoning off target?

[PlugHead]

First, do your results sound OK? Is there any way to verify the master copy sounds similar?

I've had to do the same job back when there wasn't 'hi-res' digital (44.1/48k only) - I played it dbl. speed, re-recorded it at 15 IPS and played back/Xferred @ 7 1/2 - fun times!

I think it should be fine using your method - shy of having the proper R to R for Xfer, once digitized properly, dropping the SR to 1/2 should be one of the least invasive methods.

If it's a high profile project with an appropriate budget, I'd find the right machine (and in good working order) but if not, your method should be fine...

My .02 c

Originally Posted by MessianicDreams

Hi guys, not too sure which forum to post this in, but here goes anyway:



A client has asked us to digitize a bunch of 1/4" tape. The only issue is the tapes was recorded at 3 3/4 IPS, and our Otari MTR-12 only goes down to 7.5 IPS. What I've done as a test is recorded the audio in at 88.2kHz and played it back at 44.1kHz, there-by halving the playback speed.

Is there any issues in doing this? I'm having a hard time judging if there's much quality loss as I have no idea what the tapes *should* sound like - the mixes on them aren't great (pretty low-fi) and I'm worried how much my method is degrading the audio..

If anyone has any useful tips, pointers or comments, please feel free :)

thanks!


MD

[MessianicDreams]

Thanks for your help Jay.

The project is neither high profile nor particularly well paying, so buying another R2R for this isn't really an option - particularly as (as far as I'm aware) 3 3/4 IPS is a fairly obscure speed.

The results sound......okay I guess. But not being able to compare them to any of these recordings, its pretty hard to tell whether the results sound like they *should*.

[Sigma]

plug the deck into a VCO and that to the wall outlet

or wrap tape around the capstans until it plays at the correct speed

[Matt Syson]

Hi
In describing the heads there are several configurations. You say you have 2 tracks of mono. Are there 2 different songs on the 2 tracks? Do you know what machine it was recorded on?
There were 2 track stereo heads where you turned the tape over and if you try to replay these you will get a bit of extra noise since the head will be 'listening' to half of the intended track and half that would be 'the other side' if you get my meaning.
Quarter inch 2 track has recording widths of just under one eighth of an inch (there is a guard band).
2 track stereo where you turn the tape over (domestic types) have rather less than one sixteenth track width (3 guard bands) .
Whatever you do in the digital domain will be rather better than from analogue at this point so if you can get any 'reference' sounds (known instruments) off the tape, EQ it to them.
It may be worth scouting around to see if you can borrow a machine (or blag some 'not quite studio time') with a machine at the right speed however many studios will not be maintaining their old machines so may be worse than yours.
Good luck
Matt

[MarvinDog]

What you're doing is fine. Any sonic limitations are already in the 3 3/4 ips source material. There's no concern with the 88.2 to 44.1 conversion as there's no dithering involved.

[MessianicDreams]

Originally Posted by Sigma

plug the deck into a VCO and that to the wall outlet

or wrap tape around the capstans until it plays at the correct speed

Thanks for your thoughts! However, I'd rather not start doing that...I wouldn't want to damage the machine through some sort of mis-use!

Originally Posted by Matt Syson

Are there 2 different songs on the 2 tracks? Do you know what machine it was recorded on?
There were 2 track stereo heads where you turned the tape over and if you try to replay these you will get a bit of extra noise since the head will be 'listening' to half of the intended track and half that would be 'the other side' if you get my meaning.
Quarter inch 2 track has recording widths of just under one eighth of an inch (there is a guard band).
2 track stereo where you turn the tape over (domestic types) have rather less than one sixteenth track width (3 guard bands) .

Sorry, I should've been clearer - I said its 2 tracks of mono as each track has a different song on it - either that, or this is some seriously avant-guard stuff! There does seem to be a lot of hiss on this, and the levels are quite quiet, but I'm starting to wonder whether the tape was recorded under less than ideal circumstance. Also, I have no idea what machine the tapes were recorded on. They've come out of someones archives and are dated in the early '70s.

Originally Posted by MarvinDog

What you're doing is fine. Any sonic limitations are already in the 3 3/4 ips source material. There's no concern with the 88.2 to 44.1 conversion as there's no dithering involved.

Thanks for your thoughts marvin, it's good to know that any sonic limitations are more than likely to be dictated by the tape rather than my methods..

[Matt Syson]

Hi
Don't do either of the 'machine mod' suggestions!
Any professional machine will be stabilised against mains frequency and voltage so varying either would not affect the speed, certainly not doubling (or halving) it.
Wrapping tape around the capstan while messy and it would kind of work if you were transcribing speech would introduce a massive amount of wow and flutter making music unuseable (unless you want that effect!).
The BBC used to re record their archives, I can't remember how many years between revamping but they were played and 'tarted up' before being put onto new tape.
7 1/2 ips is below a professional standard. 15 is OK and good work at 30 so expect it to be poor.
Don't forger to clean the heads VERY fequently, possibly after each song as tape this old will shed or be sticky in a BIG way!
Matt S

[TheFirstThird]

I just did a whole bunch of these for someone recently; same exact scenario. I transferred them at 7 1/2 and simply used an audiosuite time/pitch plug-in to drop them down to 50% after the transfer was complete. worked perfectly every time. I get the feeling that a lot of the 'poo-pooing' of this method is 75% 'know-it-all engineer' public masturbation.

[TheFirstThird]

Oh and if you're baking, make sure it's not acetate!

[rocksure]

It's funny that I should come across this post today because I have a similar but different situation. I have an old 1/4 track tape from the 70's -someone's archive of their kids. The dilemma is that part of the tape is music and has been recorded at 7 1/2 ips and the rest is talking/laughing at a much slower speed. My machines are 1/2 track machines that don't play slower than 7 1/2 ips and are no good for this because you can hear tracks in one direction in the right ear and other direction in the left ear. So I got hold of another old consumer 1/4 track machine and I thought that would do it, but unfortunately it's slowest speed is 3 3/4 ips and the recording is slower than that still....must be 1 1/2 ips. It's not a well paying job...just one of those fill-in jobs when you aren't that busy, but I hate to say "sorry I can't do it."

[TheFirstThird]

Some 1/4" machines, like the Otari MX5050, have 4 track AND 2 track repro heads, allowing you to transfer one 'side' at a time, while keeping you in the stereo domain. If your machine does not have this option, you'll have to transfer it in stereo, split it into two seperate mono tracks and then treat the backwards/forwards/speed issues seperately (in mono)...

[Matt Syson]

Hi
Many years ago I worked at a place that made 'talking books for the blind', an early cassette which had 3 inch spools. It used an 8 track head and ran at 15/16ths inches per second. I think 13 hours on a 3 inch spool of tape.
Matt S

[theblue1]

Originally Posted by Sigma

plug the deck into a VCO and that to the wall outlet

or wrap tape around the capstans until it plays at the correct speed

No...

The VCO would supply the wrong voltage to the tape machine's electronics potentially resulting in overheating and likely all kinds of other problems. (Yes, running at too low a voltage can lead to overheating. Imagine all those little capacitors charging up but never reaching discharge level. That's a bit crude, but hopefully it gets the idea across.)

And -- the horror of putting tape on a capstan aside -- putting tape on it would speed up the playback. Not to mention the problms of getting the seams smooth and speed calibration.

[Sigma]

Originally Posted by theblue1

No. No. No.


And a whole lot more...

really we ran our mci 16 track thru one for years as a vari speed device


and my father and many others used the tape around the capstain to vary speed also

[theblue1]

Well... you know, if you're stuck on a desert island and you have to play that tape to get a message that could save you, that's one thing... they'd applaud your savvy ingenuity when you got home.


I would strongly recommend against either approach in real life, assuming one cared about one's tape machine.




[EDIT: as I noted in an edit to my post above, like a dumb ass, I was thinking about voltage not frequency, but I would nonetheless not be hugely surprised if changing the AC from 60 to 30 Hz did not have other ramifications.]

[drtechno]

you should see a great deal of hiss because of the noise reduction systems that were used in the consumer world.

do your fast overdubb and bring it into the digital world with the hiss.
restore transients, also there is noise reduction plugingsa availle, but if you really want to get agressive with the noise:

output it to tape cassette w/ no noise reduction applied
then bring it back into digital using dolby c tape playback.

adjust tape cassette motor for any tempo variations.

Ive done digitizing from reel to reel; but the process I used was reel to reel then to a casette machine (pioneer I remember) and used the casette's noise reduction system then output that into a DBX 1BX ( all black 1 unit high I baught in 1985) expander then into digital recorder.

[theblue1]

Whoa.

Pretty unlikely this tape was recorded on a reel deck with NR on it. Possible, yes. But pretty unlikely. Very few consumer reel decks had either Dolby or dbx NR. Additionally, the classic TASCAM 1/4" 4 tracks (2340, 3340, etc) did not include NR, either (or 3-3/4ips for that matter). (Some narrow format [8 tracks on 1/4", etc] amateur recording decks in the mid-80s and later did, however.)

[John Eppstein]

Originally Posted by Sigma

plug the deck into a VCO and that to the wall outlet

or wrap tape around the capstans until it plays at the correct speed

Wrapping tape around the capstan would make it play faster, not slower.

[Sigma]

Originally Posted by John Eppstein

Wrapping tape around the capstan would make it play faster, not slower.

yeah i realized that after i posted..

i sorry!!!

[theblue1]

I beat you to it, John.

But I certainly wasn't on my game when I addressed the VCO issue... I was thinking voltage not frequency of AC (don't ask me why, I certainly know what a VCO is)... that said, I'd be sort of surprised if there weren't other issues arising from reducing the frequency of the AC from 60 Hz to 30 Hz in order to slow the hysteresis synchronous motors to half speed. You can get away with a little, but I suspect at some point you are going to have ramifications. That said, I'll leave it to others to figure out what they might be.

[Rick Sutton]

Originally Posted by theblue1

Whoa.

Pretty unlikely this tape was recorded on a reel deck with NR on it. Possible, yes. But pretty unlikely. Very few consumer reel decks had either Dolby or dbx NR. Additionally, the classic TASCAM 1/4" 4 tracks (2340, 3340, etc) did not include NR, either (or 3-3/4ips for that matter). (Some narrow format [8 tracks on 1/4", etc] amateur recording decks in the mid-80s and later did, however.)

Agreed with the small exception that the 2340 did have 3 3/4 speed.

[theblue1]

Originally Posted by Rick Sutton

Agreed with the small exception that the 2340 did have 3 3/4 speed.

I had qualms about making that statement -- since I'd run into variants before. I should have said that I had a 2340 that had 7.5 and 15ips.

I knew I had a picture of a 2340 around here and I thought it would be fun to see what the markings on the speed knob on this particular unit said:

In case it's hard to read, the speeds are: H and L

... so I'm guessing that they were, at the very least, keeping multiple speed set options open.


Oh, and I should also say that I've got VCO (voltage controlled oscillator) and VSO (variable speed operation) bouncing back and forth in my head and for a moment I almost started to doubt my notion that a VCO of some kind lies at the heart of of any circuit or module designed to implement VSO for hysteresis synchronous motors.

[Rick Sutton]

Originally Posted by theblue1

I had qualms about making that statement -- since I'd run into variants before. I should have said that I had a 2340 that had 7.5 and 15ips.

Yeah, even though I knew the 2340 as a 3-3/4 and 7-1/2 machine it wouldn't surprise me that there might have been other versions......hey, that was a long time ago.
Cheers,
Rick



And to the OP, 3-3/4 was so crappy that I doubt that you are doing any significant damage with your methods.