Hello Tom Fine,

Let me comment on some of your good observations, which I have pasted-in below.

> The idea of "playing vintage tapes on vintage tape machines" is a mute point nowadays.

I wish this were so, but sadly it isn't. Witness that a fragile (and
obviously irreplaceable) Sinatra master was destroyed just last year,
simply by being played back on the wrong equipment. The machine in
question is regarded by some as being very modern, in relative terms.

Sadly, there are many such cases, although understandably not widely reported.

By the way, couldn't we argue that all analog tape machines could be
termed vintage today?

Wouldn't it be more useful to define some of the basic tape transport
architecture classes that distinguish these machines from one another,
and then classify the various machines accordingly? For example:
Recorder/reproducer or reproducer? Servo constant-tension or constant
torque? Servo-capstan or hysteresis synchronous capstan? Force-guided
or precision guided transport path? Pin lifter or roller lifter at

For practical examples of what these differing machine architectures
might mean in historic transfer work today, consider an acetate master
that has the oxide layer falling off. Or a mylar master that is
exhibiting soft binder syndrome.

Would mounting either of those master reels and re-winding on a
transport with fixed pin lifters be responsible practice?

What about edge-forced guidance through a headblock? Or needlessly
pulling the fragile tape over an erase and then a record head? First,
do no harm.

> By the time of all the later-generation professional tape machines,
> things like speed stability and scrape-flutter were well understood,
> so playback was more precise.

Yes, achieving satisfactory speed stability was in many cases
accomplished by adopting constant-tension and servo capstan designs.

But as I see it, the matter of scrape flutter is not so simple.

The project leader for the Ampex ATR-100 certainly understood scrape
flutter, but he knew that precision guidance through the transport
would add tremendously to the manufacturing cost. Further, he
correctly understood that you could not insist that customers use only
certain brands of tape. (Especially if these were tapes not
manufactured by Ampex!) So forced guidance through the headblock was
designed in. It was a huge scrape flutter generator, yet regarded as a
necessity. And it remains difficult to argue that his approach was

> It adds up and it's audible.

Correct. Scrape flutter is audible and the components of flutter add
vectorially. This makes the choice of the reproducer quite important,
exactly as you have observed. We have been saying (for about 20 years
now) that ideally, the reproducer's flutter contributions should be
lower than the recorded flutter on the tape, by an order of magnitude.

I hope that re-issue producers always first listen to a historic
master played back on an ultra-low flutter reproducer, before
committing to trying to fix something with subsequent processing in
the digital domain. People are often surprised at what can be
retrieved from an old tape when it's played back on a great

Fred Thal