Hi, David,

The best duplicators were probably Ampex and Gauss, but you'll need 
to scrounge up lines that are being discontinued. Probably the only 
ones still working are for 0.150" tape for cassettes. Most of the 
0.250" ones have been parted out or are in landfills.

With that said, you need someone like Jay McKnight to set them up -- 
he did for Ampex and then wrote some papers about it. I don't know if the duplicator papers 
are there, but Jay is.

An issue with the duplicators is that they were tape-to-tape, or in 
the end, digital-to-tape. They were never tape-to-digital. While I'm 
not saying it cannot be done, it's expensive and custom work.

My posts have sort-of outlined what is possible with high-end pro 
decks like the APR-5000 and the A810/A820. Also probably doable on 
Ampex ATR-100s. But none of these are specifically duplicator machines.

If you're going to try and maintain 10 kHz response -- which is 
probably a better goal than 3.5 kHz which is approximately telephone 
quality -- then 2x is a reasonable compromise, also doing both 
directions in one pass for a 4x improvement in throughput. The other 
improvement in throughput is to have one operator simultaneously 
ingesting 2-4 tapes. With 4 tapes, both sides at once, at double 
speed, that is a 16x improvement over real-time. My personal opinion 
is that is approximately the limit for good-quality interview-quality 
ingest. I would not do that for music.

The next jump up which would double this throughput to 32x is to go 
to 4x transfer. The problem there, is we're now asking the system to 
go out to 40 kHz. The APR will probably get you to 28 kHz so that 
would limit the high end to 7 kHz. That's perhaps a reasonable tradeoff.

You've better get your metadata and file-name structure set up and 
well defined before going ahead as you're going to end up with lots 
of gigabytes real quickly in this scenario.

If the tapes are 1/4 track, then you can get another 2x out of doing 
all the tracks at once.

There is a small hit as you run the flipping routine on the 
backwards-running tracks, that reduces the throughput, but it should 
be less than 2-3 minutes per hour track flipped.

Going faster using real duplicator technology would require a custom 
systems design. Dale Manquen or Jay McKnight are the people to do it. 
It will be expensive.



At 06:49 AM 2/13/2006, [log in to unmask] wrote:
>In a message dated 2/12/2006 4:17:53 P.M. Pacific Standard Time,
>[log in to unmask] writes:
>Duplicators were designed to work at the higher  frequencies.
>the dialogue I raised has been very useful.  Your responses have been
>fabulous.  What are the best duplicating machines or the ones that 
>have  been most
>successful or the one's easiest to acquire?
>Continued thank you
>David Hoffman
>_www.thehoffmancollection.com_ (

Richard L. Hess                   email: [log in to unmask]
Vignettes Media                   web:
Aurora, Ontario, Canada       (905) 713 6733     1-877-TAPE-FIX
Detailed contact information: