----- Original Message ----- 
From: "[log in to unmask]" <[log in to unmask]>
> Mass digitization-
> I agree that is a valuable and, possibly, the only currently viable way to
> preserve LARGE volumes of audio and video material into the future- once the
> material has already been digitized.  Other groups have done it and, while
> there have been some spectacular digital-related losses, it has seemed to
> "mostly" work.
> I do, however, have a few serious reservations.  While there is no reason
> you can't automate digital-to-digital migration, no one has been able to
> clearly explain to me how you effectively automate analog-to-digital
> migration.  To the best of my knowledge, any proposed (or existing)
> automated system uses existing analog equipment for the initial playback.
> This equipment was designed to be manually calibrated and
> maintained/serviced.  How do you automate or robotically ADJUST the existing
> equipment to get the optimal signal off the old analog tape?  While there
> are a lot of things that can be done to manipulate a signal between the
> playback and record, if the playback machine is not properly set up for the
> tape you are playing, you may be getting a very inferior or degraded signal
> to work with.
> If one is going to leave the world of analog and make a new digital master
> (a new master that is going to be the "ultimate" version to be
> migrated/cloned down thru the generations), doesn't it make sense to
> actually make a good digital master? 
> That's where the expertise of those on this list is so invaluable.  Of
> course, if someone has made a little robot version of Steven, Tom, Richard,
> etc., please let me know.  I'll set a few of them up in my lab and retire to
> the Bahamas.
Seeing my name mentioned...

1) My expertise is strictly based on 78rpm recordings.

2) The closest we have come to automating the playing of those was/were
the record changers...first available in the mid-twenties, but improved
and standardized in the later thirties.

3) Those were NOT 100% reliable...

4) and it would be difficult to create a hardware/software-based method
which would digitize each 78, cease at its end, and await the next disc's

5) I own about 50,000 78's (or about 1.7% of all 78's ever made...).
Assuming a record changer could successfully play one side of eight
records, in a real-time duration of c.24 would take
300,000 minutes...or 5,000 hours...or 1,250 eight-hour days...or
about 3-1/2 play this entire collection! 

6) Given my estimate of a total of 3,500,000 (+/-) 78's having
been issued between 1889 and 1960...the digitization of the complete
archive of 78rpm records (insofar as that is possible)...all those
duration figures need to be multiplied by 70 to apply to some
future digital archive of "Every 78 ever recorded!). Or...about
245 years give or take!

By which time global warming and/or Dubya's new wars and/or
I dunno wot may well have rendered the project rather useless
(unless our radioactive-cockroach successors are interested...)

Steven C. Barr