MD

[Robert E Englebretson]

Folks,

(I'm reading Linganth in digest form--which means someone might have
already addressed this...)  A couple of issues from various snippets of
yesterday's discussions:

> 	... but it is my understanding that MD technology is not useful
> for anyone who plans to do acoustic analysis of sound recordings later
> on, or to preserve that possibility for other researchers, due to the
> digital compression that alters the signal beyond the possibility of
> restoration to the original spectral characteristics.
>	I haven't tried the technology myself but have been so advised

This came up as a major issue in our department last year, as a number of
people were deciding whether to use MD for field work--some of whom do
acoustic phonetics.  We contacted several phoneticians regarding this
issue (at the UCLA phonetics lab, at a local speech recognition company,
and the tech people at Sharp), all of whom claimed there is no significant
disadvantage for doing acoustic analysis based on MD recordings.

Several people on this list have claimed that MD still uses a lossy
algorithm for compression.  Can any one substantiate this--i.e. real specs
from presentday MD models?  It is my understanding that this was true of
older models, but that the ones which have come out during the last 2-3
years do in fact use non-lossy compression.  In which case it's no more an
issue than using a program like Winzip to compress the files on your
computer...

> I tell them not to use MD-- at this point, analog is better.

"Better?"  What kind of analog?  Maybe the studio-quality reel to reel is
still a good contender (but certainly not portable!), but with analog
cassettes you still have to deal with:  rumble and hiss, wow and flutter,
print-through, a contact medium which degrades each time it is played, and
you don't have instant random access to any point on the tape.  If you're
looking for a completely faithful representation of the signal (if there
ever is such a thing), analog cassette has a much smaller spectrographic
and dynamic range than both DAT and MD (I don't have the numbers right at
hand, but can easily look them up...)

As far as the DAT versus MD question, I have used both, and each has
advantages over the other.  I used DAT extensively while working for the
Santa Barbara corpus of spoken American English, and also in my own
fieldwork in Indonesia.  The advantages are that you can record 4 hours on
one tape (at 32K), without having to mess with it or turn it over.  This
is especially nice when you want to not be co-present for the recording.
In Indonesia, I would set up the portable DAT and stereo microphones in a
camera bag, give it to one of my research assistants, and we would go our
separate ways--and I would get four hours of data.  (In the Sony D7 i was
using, the battery pack would last six hours--so running out of power
during a recording was never a problem.)  By comparison, the main
disadvantage of MD is that we're limited to 74 (or 80 with the newer
disks) minutes of recording on one disk--not as convenient for doing this
type of recording.  The advantages of MD over DAT are numerous though.
The main problems I've experienced with DAT are:  the heads get out of
alignment after a while; the tapes are still a contact medium which wears
out and degrades and tangles and such; some people find DAT recorders to
not be user friendly.  The advantages of MD (in addition to not having the
disadvantages I just mentioned for DAT) include:  portable MD recorders
are much smaller and more durable; instant random access to any point in
the data, and ability to easily perform editing functions in real time;
the recorders and disks are much cheaper than DAT.  As far as the
unavailability of disks (or DAT tapes) in the field:  bring enough with
you so that isn't a problem.  Same for high-quality batteries if they're
hard to come by where you work.  There are several good places which sell
these items in bulk.  And of course no matter which recording medium you
choose, you need high-quality mics (sort of goes without saying).  And
archive the data when you get back from the field (I used direct digital
onto high-quality CDR--not CDRW).

>> which you can easily  digitalize directly as WAV
>> through the computer's Line In, with little noise.
> Which means that you depend on your computer's analog to digital
> converter, therefore retransforming the signal.

If you want direct digital, you obviously would not want to do this!
Soundcards with optical or SPDIF ins are pretty cheap these days; if you
want direct digital then hook up the digital output of the MD to the
digital input of your soundcard...

It was also suggested that people would want to use MP3 or RealAudio--but
both are lossy, and I'm not sure why one would want to do this, given that
you can do direct digital transfers from MD onto your computer anyway.

Just my two cents from my own experience with this topic.  Okay, I'll sign
off and return to being a Linganth lurker rather than a poster...

--Robert Englebretson
UCSB Dept. of Linguistics