Mellotronists

> From: ferrograph@...
   > Date: Thu, 6 Jun 2002 06:40:48 EDT
   > 
   > that seems reasonable- bias frequencies were generally lower in
   > those days, even on professional machines. 

I don't know that the bias frequency has changed all that much over
the years.

   > has anybody seen a really good explanation of how bias works?

"Bias" in electronics typically refers to adding a dc voltage or
current to move the operating point of an amplifier stage to where it
behaves best.  For instance, a single tube or transitor stage won't
clip in a symmetrical way, so the bias is adjusted to center things.
(That's a very simplified description.)

Tape bias is similiar.  In tape, we have nonlinearities.  Sure, if you
crank the record levels the tape will saturate, we know about that.
But there's also a little glitch in the magnetic curve near the
zero-point, a kind of "dead zone" which is unfortunately where an
audio signal typically spends most of it's time.

Tape bias is an ultrasonic sine wave that's summed with with the
regular audio record signal.  It keeps the audio signal from spending
much time in the dead zone, and the time it does spend in the dead
zone is averaged out and thus much smoother.  Tape bias effectively
removes this annoying form of distortion.

So every tape deck has an unltrasonic audio oscillator built in, and
it's level is adjusted for the size of the dead zone of the given tape
formulation.  And if you play back a tape much slower than you
recorded it you can hear the bias oscillator, and it sounds exactly
like that Moodies noise.  

  -- Don

-- 
Don Tillman
Palo Alto, California, USA
don@...
http://www.till.com