Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Barry Hufker]

I think I heard of someone possibly doing this but it's been so long I don't remember.

What I'm discussing is: using the bias frequency on an analog tape recorder as the "sync" tone for reducing the ATR's wow and flutter.  The idea of course would be that as bias is laid down during recording, that signal is used to tell the capstan and reel motors what the speed should be.  On playback, the bias signal is read, performing the same function.

Wouldn't this reduce wow and flutter to something negligible?  180k or 250k should be a hell of a clock for the servo motors.

[bruno putzeys]

I'm unaware of anyone using the bias oscillator for syncing the motors. ISTR crystal locked motors though. Anyhow this only corrects some wow because most errors are caused by asymmetries in the capstan, pinch roller, stickiness etc. What does work is what these folks do: http://www.plangentprocesses.com/. While digitizing the tape they record the bias tone along with the audio. The bias tone is then used to correct the audio in the digital domain.

[John Roberts {JR}]

That's pretty clever... I would assume the 100k+ bias would be down in the dirt but I guess enough signal for parsing time errors in that several uSec range.

JR

[drknob]

I 'think' Plangent was at AES about 3-4 years ago. They played a demo on headphones of a Procol Harum mono recording (Salty Dog?) before and after processing. I remember being amazed at the ambience audible once the w&f was reduced. In mono!

[Barry Hufker]

Plangent has some pretty amazing sound files on their site and it is an amazing process.  I was just hoping for a method everyone could use to make better recordings every time without a special digital process.

[Jon Halverson]

Could that be done by "simply" locking MCLK to the bias with the appropriate filters in the loop?                                                  

[bruno putzeys]

That's the idea - in principle that should get you quite close, but in practice you really need digital techniques to recover the bias modulation with sufficient quality. AFAIK quadrature detectors are not easy to do in the analogue domain. Also, for fast modulation, sampling theory comes apart and you need to develop a new "nonuniform" sampling theory, which is at the core of the Plangent process.