wjm (12/20/82)
To all you folks on the net ...
I'm glad to see all the discussion about the pro's and con's of digital recording.
I'd like to take this opportunity to flame some more about some of the comments
made this past weekend.
1. I for one CAN hear TV raster and those (expletive deleted) ultrasoninc
anti-shoplifting devices (they give me an Excederin headache and if possible
I try to patronize stores which don't use them). I can also hear (or perhaps
sense) ultrasonic vehicle detection devices when walking past intersections
with traffic signals that use them. I think the frequency used in both store
and signal detectors is about 25 Khz.
2. The comments about phase shift introduced into the digital process by
timing errors between channels are well taken, but I agree with an earlier
comment that more error can easily be introduced by head-azumuth misalignment
in an analog recorder. This problem is one of the main factors in obtaining
good high frequency (here I mean > 15 KHz) response from a cassette deck and
as far as I know, only one manufacturer, Nakamichi (who else ?) has solved
it by including automatic head azumuth adjustments in its decks.
3. One must keep in mind that the advantages of digital audio will only
be achieved with a well designed system (the same is true of analog recording)
and I suspect that the same record companies who currently produce noisy
analog disks will try to cut corners in the digital process (with audible effect).
4. In answer to the question of cassete decks with response > 20 KHz (to the
best of my knowledge) is that certain (higher-end) Nakamichi decks (the 600,
700, and 1000 series modesl
with automatic azumuth adjustment (see above) and separate record and playback
heads) can provide response to > 25 KHz. However, since the leas expensive
of these decks lists for about $2K (and Nakamichi decks aren't usually discounted
you don't find them in most audio systems. However, you can hear the difference
provided by that last treble octave.
5. The digital recording process will bring about a need for amplifiers with
dynamic headroom (to meet the peaks of the 90dB S/N ratio) and for transducers
(speakers, as they used to be called) with a response to 20 KHz. There are
several supertweeter designs that can convert 40 KHz electrical signals into
air pressure waves (the stores use them to nail shoplifters and give me head-
aches).
6. In theory, ANY periodic signal can be reproduced as a digital series of
numbers via Fourier Transform techniques, provided that the criterion of
Shannon's sampling theorem is met (the sampling rate is at least double
the highest significant frequency component in the input signal). So much
for theory. In the practical world of real audio, we've got to answer two
questions: a. Is a complex audio signal periodic ? As mentioned above, the
low and mid frequency components change slowly enough (with respect to the
sampling rate that this condition is true, but it may not be the case when
discussing the high treble). b. What is the highest significant frequency?
If we set it too low, we lose some information (I'm inclined to believe that
the low order harmonics (2nd & 3rd order) of the highest octave (10-20KHz)
contribute something to the timbre of the sound (even if they can't be heard
directly). On the other hand, if we set it too high, we waste bits and
reduce the capacity of our digital storage medium unnecessarily. My personal
opinion is that 16 bit encoding with a 50 KHz sampling rate (like the
Soundstream system used by Telarc, among others) is the minimum bit rate
that should be employed. (Of course, the number of bits governs the S/N
ratio rather than the highest frequency that can be reproduced.)
7. Mark Terribile's comment is well taken. Every concert hall has its own
unique characteristics. Frankly, I enjoy concerts in both Carnegie and Fisher
Halls, although I try to sit in different places to compensate for the delays.
In Fisher Hall, I tend to sit in the rear third of the orchestra, while in
Carnegie I prefer to sit further forward. (Then there's Symphony Hall in
Boston, where there are few 'bad' seats, and I really don't care, except that
I don't like those side seats in the 2nd balcony where you have to kink your
neck to see the orchestra.)
8. End of flame ... Best wishes for the holidays to all and to all who like
good music, let's hope that the implementation of digital home recordings via
CD's lives up to its theoretical promise ....
Bill Mitchell
Bell Laboratories - Whippany
(harpo!wjm)