[Table of Contents]


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ARSCLIST] National Recording Preservation Board (NRPB) Study



I honestly didn't know there were 80,000 unique 78 sides recorded! What is the material?

Steven, do you think you have more disks than Joe Bussard? Did I understand you correctly that you have 40 THOUSAND unique 78's (ie no repeats)? Or, how much of that is overlap?

-- Tom Fine

----- Original Message ----- From: "Mike Richter" <mrichter@xxxxxxx>
To: <ARSCLIST@xxxxxxxxxxxxxxxx>
Sent: Monday, May 15, 2006 4:36 PM
Subject: Re: [ARSCLIST] National Recording Preservation Board (NRPB) Study



steven c wrote:

I have about 80,000 78 sides to (eventually) archive in digital form.
As moderate-quality MP3 files of, say, 500KB each...I would need
about 40,000MB...or 40GB, which I could store on a single not-too-
expensive drive. Wav files would be (I think) about 20 times the
size, or .8TB (for I don't know how much...).

The numbers are easy enough and probably worth sharing here.


Redbook WAV (two channels, 16 bit depth, 44.1 ksps, PCM) is about 10 MB/minute. Minimum quality suggested is usually one channel, 16 bits, 22.05 ksps for 2.5 MB/minute. (In practice, 10-12 bit depth would suffice, but I know no software for that; 8 bits is pushing it.)

Lossless compression is more effective on quiet material than on noisy, so you can't count on more than 40% reduction.

The usual compression for MP3 is about 11:1. Good recognition is achieved from 78s at 32 Kbps (22.05 ksps, monaural) which means four minutes per MB. On average, a 78 side is likely to run 3-3.5 minutes - so be a pessimist and say 4 minutes. 80,000 sides would then be around 80 GB in 32 Kbps MP3 up to 3.2 TB in redbook WAV or equivalent (e.g., 32 bits depth but single channel). Go to archival extremes such as two channels, 96 ksps, 32 bits, PCM and you're looking at 16 TB - not impossible today nor even a great fortune, but probably more expensive than the benefits for access are worth. (Access is distinguished from archival purposes; the cost/benefit analyses differ for those two.)

Mike
--
mrichter@xxxxxxx
http://www.mrichter.com/


[Subject index] [Index for current month] [Table of Contents]