Reducing Ampex 440 wow/flutter on playback

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Icegoldnixon

Well-known member
Joined
Feb 3, 2015
Messages
79
Hi folks. I have an idea that I'm having a tough time validating.
I have a 4-track Ampex 440 and a 4-track Otari MTR-12. I'd like to mix down to the Ampex and play the tape back on the Otari for conversion to digital.

My question is, if I record a 9.6kHz pilot tone onto the Ampex tape at mixdown, could I use that tone as a reference for the capstan when playing back on the Otari? (see attached diagram)

And if I can do this, would the effect be to reduce wow/flutter/speed variation apparent on the Otari output? Or should I just play back from the Ampex and not bother? Thanks!
 

Attachments

  • otari capstan control.png
    otari capstan control.png
    69.5 KB · Views: 14
Unless you have a critical timing requirement, just play back on the Ampex. If you have a critical timing requirement, use SMPTE timecode and a synchroniser.
 
An Otari MTR12 should have lower wow and flutter than an Ampex 440. It should beat the 440 at every spec except headroom and subjective sound quality which the 440 will probably win.
 
Re-reading your post, why dont you record and playback on one of them? Your post is confused by the way.
An Otari MTR12 should have lower wow and flutter than an Ampex 440. It should beat the 440 at every spec except headroom and subjective sound quality which the 440 will probably win.

My goal for this set-up is to gain the benefits of Ampex subjective sound quality, while using the Otari to cleanly capture that recorded Ampex sound. But because the Ampex has significant speed variation across the length of the reel, the Otari would play it back "wrong" wherever the Ampex drifted from exactly 15ips.

I am inspired here by Plangent Plangent Processes - the finest tape playback on the planet. They use the bias frequency on old tapes to correct for wow/flutter digitally.

My idea is that (maybe) the Otari would be able to cancel out the wow/flutter as printed by the Ampex, because the pilot tone (printed using the Ampex) would encode all of these variations. By chasing the "fluttered" pilot tone the Otari transport would match the original Ampex behavior, thus making the Otari playback truer to the original source material (with some Ampex vibes left over).

So I guess my question is, does this idea make any sense? Would just playing back on the Ampex be more accurate than making the Otari mimic its idiosyncrasies? Thanks!
 
Playing back what was recorded on the Ampex on the Ampex would mean the same speed inconsistencies would be cancelled to a degree. But if you are really worried, the best way is to use a synchroniser and SMPTE code, as I said before. Then you can lock either machine to SMPTE. Of course you need the right hardware, like a Zeta 3.
 
Playing back what was recorded on the Ampex on the Ampex would mean the same speed inconsistencies would be cancelled to a degree. But if you are really worried, the best way is to use a synchroniser and SMPTE code, as I said before. Then you can lock either machine to SMPTE. Of course you need the right hardware, like a Zeta 3.

I have a Timeline Microlynx. But who would be locking to who in this scenario? We're working with one tape across two machines, so the timecode (or pilot tone) striped on the tape by the Ampex is what we want the Otari to chase (or resolve to) during playback. Sorry if I am not understanding your point.

I don't think the Ampex can be locked to SMPTE, it's a 440B with an AC capstan.
 
Last edited:
Well the Ampex can but its not easy or cheap. You can lock the machines to each other, Ampex being master, but thats not what you want. So you stripe the Ampex with TC as you record on it, and then you slave the Otari to a timecode generator. It can be done, its messy, and I really think you are over thinking it.
 
I think replay makes more use of the electronics than recording
so that said I would rather record with Otari and played up by Ampex.


But it seems like extra hassle to have two tape machines doing the one machine job.

Why don't get a tech to merge the Otari transport with the Ampex electronics and headstack?
 
Well the Ampex can but its not easy or cheap. You can lock the machines to each other, Ampex being master, but thats not what you want. So you stripe the Ampex with TC as you record on it, and then you slave the Otari to a timecode generator. It can be done, its messy, and I really think you are over thinking it.
It is definitely sounding like I'm not gaining much benefit for my trouble. Thank you for your advice!
Why don't get a tech to merge the Otari transport with the Ampex electronics and headstack?
That would be a cool idea, but I don't want to butcher both machines. Also, we wanted to maintain a separate transfer room and leave the Ampex configured for mixing/tracking. Hence using two machines for one job.
 
Having the second machine lock to a tone played back by the first would certainly make the long-term playback times equal - but considering the response times (lag) of the capstan motor servo loop, it would likely make wow and flutter worse as the servo overshoots on small changes.

I was thinking something like this would be the reality. Maybe some combination of lookahead+tone frequency could minimize overshoot? But from what folks are saying I should probably just stick to a single machine.
 
Playing back what was recorded on the Ampex on the Ampex would mean the same speed inconsistencies would be cancelled to a degree.
When inconsistencies are wow & flutter, they won't cancel, being erratic. Only drift related to tape tension variations due to change in spool diameter may be compensated.
But if you are really worried, the best way is to use a synchroniser and SMPTE code, as I said before. Then you can lock either machine to SMPTE. Of course you need the right hardware, like a Zeta 3.
No synchronizer can compensate the rapid variations of flutter. Wow may benefit a little.
 
I did hear of a Studer C37 mod which got rid of the mechanical tape counter which gave improved W&F performance . Of course the Ampex probably has a digital counter , so not applicable .
Making sure the mechanics/solenoids handle the tape gently and that none of the rollers/guides are worn and allowing vertical oscillatory motion of the tape might be the best course of action .
Brian Roth runs a fleet of old Ampex machines , he may have a few tips specific to the 440 .
 
I have never found the speed variation of the 440 to be objectionable. If you do a lot of music editing I could see having problems but for just mixing to I wouldn’t think so. Playing back on the Otari wont get rid of that speeed variation. Only something like Plagnet would.
 
It also rarely produces the expected results. People forget how much the transport contributes to the overall sound of the deck.
The transport contributes essentially in two ways:
Speed variations, particularly flutter is very important. I think a large part of the 'dryness" of digital recording is the quasi absence of flutter.
Head wrap is also a major source of variations regarding the signal stability and the LF response.
One may find, by grafting Ampex electronics to a modern transport, that some of the magic is gone.
 

Latest posts

Back
Top