Sound recording

[Neil Coffey]

On Mon, 6 Aug 2001, Joe Pater wrote:

> I recently spent a fair bit of time researching this question. For some
> reason portable Minidisc recorders are rarely, if ever, made with a digital
> output (presumably, either for cost-effectiveness, or to discourage digital
> copying).

As I understand things, there is a genuine technical reason for
blocking digital-to-digital copying of Minidiscs. Because of the way
the compression process works, if you repeatedly compress then
decompress data via a minidisc, this actually leads to a
degradation of the recording. (A bit like if you continually
compress and decompress a JPEG image this gives a slight
degradation of the image each time.) So the only real way to
do a faithful digital copy would be for the machines to output
the actual compressed data stored on the disc, and Sony presumably
don't want to do this for reasons of propriety. But of course,
the fact that it discourages copying is no doubt a handy bonus
for the music industry...

> You can get a "Pro" minidisc recorder for over $1000, but at that
> price, you may as well buy a DAT recorder, which doesn't use any compression.

There's a standard protocol for marking digital data as being
'non-copiable'. As I understand things, consumer goods with
digital inputs check for the presence of this marking and act
accordingly; "professional" equipment generally provides the
facility to ignore the non-copiable status of the signal and
let you copy it anyway. So presumably that's what your "Pro"
minidisc recorder does.

Some digital soundcards will also let you do intelligient
things with the encoding of digital data -- e.g. strip the
data of the 'non-copiable' marking. The card we have in our
lab for example comes with a program which simply takes
sound to the digital in, strips it of its non-copiable flag,
and pipes it to the digital out. So I assume you could hook
a minidisc machine up to each end and hey presto. If your
main reason for not going for minidisc is the non-copiability,
then this might be an alternative solution.

Incidentally, has anyone actually studied the suitability
of minidiscs for linguistic analysis in detail? The compression
algorithm potentially does some fairly hefty munging about
with the signal, and I'd have thought one would have to be
at best pretty cautious about any acoustic analysis done on a
minidisc recording.

Neil