[Info-vax] RMS internals?

[glen herrmannsfeldt]

In comp.os.vms Richard B. Gilbert <rgilbert88 at comcast.net> wrote:
 
> I think a better grade of tape was used for 1600 and 6250.  I no longer 
> recall if there were TWO better grades of tape; one 1600 and another for 
> 6250 bpi.  It has been many long years since I last used reel to reel tape!

It might have been better tape, but that didn't help for 800.

For 800, NRZI code, all the tracks clock together.  There either is
or isn't a transition on a track for a 1 or 0.  It has to be odd
parity or you can't have all data bits 0.  If the head azimuth
alignment is different than the tape was written on then the tracks
don't clock together (close enough) and data errors result.

For 1600 (phase encoded) each track clocks separately.  Logic in
the drive puts the bits back together again, but they can be read
even if the head azimuth is off.  (The flux transitions can be
up to two per bit, so 3200 FCPI.)

At 6250 they are GCR, with the actual bit density (flux transitions)
being somewhat higher.  Not twice as high (as for 1600) but 
maybe around 9000 FCPI with the extra going to error detection 
and correction. 

-- glen