[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [ARSCLIST] Prefered format for digital archiving?
At 03:37 PM 3/24/2004 -0600, Brandon Burke wrote:
I know this is a loaded question but...
Is there an accepted standard format for audio files that are to be digitaly
archived on a server of some sort? By this I mean a format that does not
condensing and/or any other manipulations that would in any way affect my
ability to go back, open up, and have access to "the whole thing". I'm not
talking about simply saving, let's say, a Pro Tools session en masse but
a file format that retains as much of the "sound" information as possible. Not
the session itself. I'm asuming that these are completed sessions. Also, I'm
not concerned with things like CD-Rs, etc; though I obviously would like to
preserve the ability to make them again later.
My guess is that we're talking about either a WAV or AIFF files, no?
In my non-professional opinion, there is no standard format.
There are three parameters of interest for a digital file with audio
information: sample rate, bit depth, number of channels. A CD uses 44.1
ksps, 16 bits, two channels. The highest quality at moderate cost today is
96 ksps, 24 bits, two channels. Purists will argue that information is
still lost with those parameters - and that is certainly true in theory. In
the other direction, spoken word is generally considered fully intelligible
at 16 ksps, 8 bits, one channel - my approximation to the target for
telephone quality. Somewhere in that 36:1 range, you are likely to find
your preferred operating point.
AIFF has lost popularity in modern times; it offers no advantage over WAV
and has been pretty much superseded even on the Apple computers which were
its 'home' in decades past. Depending on the operating parameters you
choose and on your considerations of storage space and convenience, you
should look into lossless compression. Both Shorten and Monkeys Audio
provide lossless compression. If the parameters you have chosen for WAV are
substantially used by your signal, compression of the order of 2:1 is
routine. If you have much wider bandwidth than is used for the bulk of the
audio (including noise), substantially higher compression is routine. Since
the compression is lossless, the decompressed file is identical with the