----- Original Message -----
From: "Mike Richter" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Thursday, January 16, 2003 12:34 AM
Subject: Re: Record collection available
> At 08:08 PM 1/15/2003 -0600, Premise Checker wrote:
> >I hope someone takes this stuff and holds it until the time when it can
> >onto the net as MP3 files. I'm presently coordinating getting some 2000
> >C-90s that include most of the Victor and Columbia M sets to a friend,
> >Kevin Mostyn. Is anyone here now interested in helping out putting them
> >the net? Much, much easier than working with the 78s!
> It could get up onto the Internet quite easily. You're talking about less
> than 50 GB (figuring 32 Kbps). I'd guess that there are already electronic
> catalogues which can be used for documentation. Initial availability would
> then depend only on having bandwidth for whatever traffic develops.
> One could select and qualify a group to process the files after the
> posting, denoising and editing from the cassettes or (more reaosnably)
> WAV captures used for the initial posting. I would recommend similar
> deferral of database development. The initial system would provide access
> with the catalogue number or with full-text search either using HTML or a
> PDF version of the catalogue. In time, that could evolve to a sequel
> server. Similarly, upgrading the connection speed can be delayed until the
> initial link is saturated.
> [log in to unmask]
Actually, this relates to an idea I've long had (and filed in my mental
"someday" file!)...that of creating a sound-file archive of every 78rpm
recording...or possibly every known recording...still in existence!
The problem here, at least for us "civilian types," is the necssary internet
connection. Storage space isn't a problem...in fact, this is evolving into
an ever smaller problem, as hard drives become both larger and
cheaper...you can buy drive approaching 200GB for around $Cdn 400
(US$ 250), and likewise with computers. If we assume (arbitrarily) that
5 million 78 phonorecord sides exist, at an average length of 2.5 minutes
(150 seconds) each, you can do the math and figure out the storage
needed for our "ultimate archive."
The C8T files (admittedly awkward for some uses) use 160 bytes per
data record, or (in most cases) 320 bytes per phonorecord; that would
work out, given the above arbitrary figure, to just under 1600MB, or
>1.6GB for the database. Expand the data records and put them into,
for example, Access...where the higher space used for internal overhead
would be cancelled out many times over by the Access "packed storage"
format; if we go to 1KB data records per side (2KB/phonorecord) at
worst we multiply by 6 (probably much less, considering eliminated
training blanks!) for a 10GB database.
Deduction: It not only COULD be done, it SHOULD be done! In fact,
our biggest problem may be convincing RIAA lawyers we aren't planning
to compete with their member firms!
Steven C. Barr
(who has 30,000 phonorecords that could be transferred as a start...)
<this will be cross-posted to 78-L for further opinions>