I'll come along for that ride.
If my math is not faulty...
The difference between 10,000 hour-long cassettes (for example) captured in 48 kHz / 16-bit files and the same cassettes captured in 48 kHz / 24-bit files is 3,218.75 GB.
The difference in data between 10,000 hour-long cassettes captured in 44.1 kHz / 24-bit files and the same cassettes captured in 48 kHz / 24-bit files is 785.15625 GB. That's one hard drive's worth of difference in the amount of data.
So do we throw away 785 GB of potentially valuable data forever (because it is apparently only marginally valuable) or do we save the price of storage costs that continue to drop?
Have a nice weekend!
Bruce J. Gordon
Audio Preservation Services
Cambridge, Massachusetts 02138
tel. +1(617) 495-1241
fax +1(617) 496-4636
On Feb 22, 2013, at 1:33 PM, Tom Fine <[log in to unmask]<mailto:[log in to unmask]>> wrote:
Totally agree about 48k! For talk radio shows? Come on, even that's overkill.
24-bit is important, however. The reason, better DSP performance if you have to go in and do severe cleanup.
I'm thankful I've never even SEEN 10,000 cassettes, much less had to deal with them. As I said, good luck to ya! The upside -- it could be 10,000 Exabyte cartridges or 10,000 DATs.
-- Tom Fine
----- Original Message ----- From: "Richard L. Hess" <[log in to unmask]<mailto:[log in to unmask]>>
To: <[log in to unmask]<mailto:[log in to unmask]>>
Sent: Friday, February 22, 2013 12:35 PM
Subject: Re: [ARSCLIST] [GRAYMAIL] Re: [ARSCLIST] Digitizing 10,000+ audio cassettes
On 2013-02-22 10:59 AM, Joel Alperson wrote in part:
I have to confess, I'm caught off guard a bit by the recommendation for
24/96 files for voice recordings, although given the cost of storage,
it's probably not that big of a deal to go with that bit and sample
I have become more of a proponent of 48 ks/s for speech recordings from cassettes since I have a spectrogram shoved in my face on a more regular basis in iZotope (which I have moved to for cleaning). There's no audio above 20 kHz coming off these cassettes. It rarely happened back in the day and that was only within the Nakamichi line (pretty much). I've given you links of reading at my blog -- there is one article near the top about the 4 dB ambiguity at 16 kHz...and that was cooked into the non-standard back in the day. Post-recording HF loss makes that even a larger ambiguity.
Also, may cassette decks had 19 kHz multiplex filters in them so the Dolby wouldn't get confused (among other things), but some were not defeatable.
IASA TC-04 states: ( http://www.iasa-web.org/tc04/key-digital-principles )
2.2 *Sampling Rate*: The sampling rate fixes the maximum limit on frequency response.When producing digital copies of analogue material IASA recommends a minimum sampling rate of 48 kHz for any material. However, higher sampling rates are readily available and may be advantageous for many content types. Although the higher sampling rates encode audio outside of the human hearing range, the net effect of higher sampling rate and conversion technology improves the audio quality within the ideal range of human hearing. The unintended and undesirable artefacts in a recording are also part of the sound document, whether they were inherent in the manufacture of the recording or have been subsequently added to the original signal by wear, mishandling or poor storage. Both must be preserved with utmost accuracy. For certain signals and some types of noise, sampling rates in excess of 48 kHz may be advantageous. IASA recommends 96 kHz as a higher sampling rate, though this is intended only as a guide, not an upper limit; however, for most general audio materials the sampling rates described should be adequate. For audio digital-original items, the sampling rate of the storage technology should equal that of the original item.
I mentioned other sampling rates as they are, in my opinion, acceptable unless these cassettes are the very highest quality AND they are the inherent built-in sampling rates of reasonable affordable tools that will get the job done in an acceptable manner. Only the Otari high-speed digitizer is likely to handle sample rates not related to CD quality.
The 10 kHz upper limit imposed by the 22.05 ks/s of the relatively inexpensive 8 X British system is a function of the 8 X record option. It does produce 44.1 ks/s files at 4 X as I pointed out. The question is whether you want to spend the time considering that the likelihood of a substantial amount of program material being reliably recoverable much above 10 kHz from 10,000 cassettes is problematic.
I just looked at the spectrogram of the RE-10 mic demo with a male voice at http://www.coutant.org/evre10/index.html and the only significant energy above 5 kHz is in the "S" sibilant sounds "thiS iS..." and that goes out strong to the upper limit of the file around 15 kHz.
The British ingesting system is targeted towards churches that have a large sermon ministry on cassettes and want to make this back catalogue available digitally. I first learned about them through Technologies for Worship Magazine.
I agree with Bruce Gordon that for Harvard, I would ingest everything at 48 ks/s minimum and most items at 96 ks/s (all at 24 bits) and I'm trying to move many of my clients to 48 ks/s ingest rather than 44.1 ks/s, but many factors are involved, notably how the client will store the files. I assume that anyone moving forward with 10,000 cassettes will develop a way of managing multiple TB of data which is now easy to do considering the availability of multiple multi-slot NAS units. My current thinking is that my next NAS disk purchase will be WD Reds which are optimized for this use on a non-enterprise basis. Attempting to do this on optical media would be semi-suicidal in my opinion. Remember, three copies in three different locations if possible.
I just do not see the need if using equipment that performs well at 48 ks/s to ingest at 96 ks/s for spoken word cassettes. Any music deserves 96 ks/s as do grooved media if for no other reason as it helps separate clicks from program.
My big question is, what is the easiest way for me to learn to use a
software package like Samplitude (recommended by Richard Hess)? I've
seen an instruction manual or two for these types of programs and
they're massive and seem very complex. Given that for now I'm just
interested in recording (not editing) material, I'd hope there's an
easier way for me to get familiar with these products.
You can download a 30 day demo.
For just recording, it is very easy...sort of. Samplitude treats odd/even pairs as stereo pairs as a default and since the best available cassette machines are stereo, I strongly suggest ingesting in that mode and making a decision in post as to which channel you are going to preserve. You'll find that it may vary through a cassette.
You open a new virtual project (VIP) and select sample rate and number of tracks.
In the record option menu (red light surrounded by a gear) you select the formatting of the file name which will be the VIP and the track name (at least that is my suggestion) AND make sure you're on 24 bits.
In the VIP layout double click the track name and put the file ID in that.
Make certain that the input routing is correct and each track is record enabled.
Press record and then start the cassette machines.
At the end, save the VIP...the WAVs are already saved.
I probably left a few things out, but this month's tutorial on the Samplitude site goes into more details.
Richard L. Hess email: [log in to unmask]
Aurora, Ontario, Canada 647 479 2800
Quality tape transfers -- even from hard-to-play tapes.