5121cdd@houxm.UUCP (C.DORY) (05/30/85)
C'mon Phil -- back to old ad hominem arguments again, huh? Why don't
you guys just give up and find an alley behind your lab and settle this
the way scientists used to (e.g., Galois).
There is one classification of audiophobe that Phil neglected to mention,
that is, the armchair audio engineer. Yes, our old friend has once again
appeared. The armchair audio engineer (AAE for short) is generally a
scientist (or worse, engineer) in his first life with a less-than-professional
(I don't mean attitude, I mean where do you make your money) interest in audio.
The problem arises when these AAEs, given that they are generally very
intelligent, high achievers, make a stab at analyzing audio in the same
vain they do their job (with the same values, etc.). The problem lies
in the fact that audio engineering is not a pure science, but a mixture
of science, art, AND EXPERIENCE! One cannot argue science -- science is
fact. It is, however, possible to argue the artistic aspects of audio.
(Experience tends to be the way by which we build our values.)
Okay, now, what is happening is that arguments are turning into flames
about science. Something's funky here. Remember, we can't argue science --
it's fact. What we can discuss (notice that I'm trying to civilize the net --
a noble goal, huh?) is the artistic application of science.
And herin lies the inner conflict of the AAE. The AAE has trouble identifying
with the "nebulous" world of art and the even more fuzzy realm of applying
science and art together. Because, here we must transverse the world
of experience (now we all know why they're called "armchair" audio engineers).
There are greater problems, even. That is when the AAE decides to move
into audio professionally (I mean make money at it). What we have now
is a knob twiddler -- a pro engineer who chases specs. I'm sure we can
all name record companies that employ engineers of this classification --
the problem here is that they tend to be the large, well-established labels
with the top artists. The problem is that they produce unmusical releases.
Enough divergence, back to Phil. Those "nebulous" terms: imaging and
dimensionality are quite definable. Let's try two approaches: the first,
using a well known metric (your ears) and second, the literature.
1) Imaging and dimensionality are two of the most noticable aspects
missing from DGG recordings. You see, music and its acoustic are
inseparable (except for rock etc.), however some of our friends,
the knob twiddlers, have made it to important places and try, for
some unknown reason, to separate the music and the acoustic thru
a technique called "multi-miking". This stuff is bad medicine
because what you have to do is create the acoustic electronically.
As well, the natural timbres of the instruments have been altered.
What happens is that the orchestra (music and acoustic) is under
the direct influence of the knob twiddler - NOT THE CONDUCTOR
WHERE IT SHOULD BE!!
2) I will cite a reference that should interest several people:
"Stereo Microphone Techniques: Are the Purists Wrong"
by Stanley P. Lipshitz (AES preprint 2261)
In this paper, Stanley defines several of these "nebulous" terms
that we all see floating around. This, however, is only a side
topic to his main premise. He sets out to show (and does a
magnificent job) that imaging and dimensionality are not obtainable
through spaced omni techniques (ala Telarc) but only through
coincident mic techniques using bi-directional or directional mics.
(Well, afterall, omnis are just sloppy cardioids anyway.)
Flame on.
Craig Dorykarn@petrus.UUCP (05/31/85)
I think Craig has hit the crux of the problem here. Writing music and conducting an orchestra is an art; even arranging microphones and mixing master recordings (activities usually called "sound engineering") is an art. In each of these things, personal and subjective judgements (even though there may sometimes be widespread agreement) are involved. However, once the signals leave the hands of the artist (in this case, the console recording engineer) and enter the recording system, the remaining steps are a pure science! From that point on, the sound reproduction system has a precisely defined job to do: reproduce, at a later time and at a different place, sounds corresponding as closely to the original signal (except for amplitude) as possible. This is a function which can be tested, evaluated and rated in a purely objective fashion. Some of the shortcomings are not likely to be audible; if you claim that they are, it's up to you to prove it with a properly controlled test. No test, no proof. Anecdotes don't count. Once you've proven that they are in fact audible, then of course some subjectivity arises as to which differences are more or less objectionable. No problem. For example, you might claim to be willing to sacrifice 50 db of signal-to-noise ratio in order to extend frequency response to 25 khz, because you've proven to me that you can hear 25 khz in a hearing test. Fine. I'd be happy to sell you my old Teac reel-to-reel tape deck. You may even tell me that you actually prefer a system with, say, more frequency response and harmonic distortion, noise, flutter and wow (i.e., high end LP players), even without any compensating improvements in other areas. I may think you're crazy, but you're entitled to your preferences. HOWEVER, you CANNOT then turn around and tell me that your system is somehow "better" at doing the well-defined task of reproducing a signal exactly as it was recorded! Your real problem is with the "engineer" artists back at the recording studio who didn't equalize the recording to suit your tastes, or introduce enough distortion, or run enough industrial strength vacuum cleaners in the concert hall while the recording was being made. You can't yell at the designers of digital recording systems for doing their job well. You're welcome to buy an equalizer, or a variable delay line (to introduce wow), or better yet, a guitar fuzzbox and modify the sound until your heart's content. But remember that makes you an artist, not an engineer. I stick by my original assertions: 1. Any audible differences between CD players (assuming proper functioning of each, i.e., a good stream of bits being presented to each D/A converter) remain to be proven. 2. The worst CD player (assuming no outright malfunction) is far better, in every possible way, at the specific job of reproducing the exact signal at the output of the studio console mixer than even the best LP player. You may not like how the studio mixer output sounds, but that's not the CD player's fault. Phil Karn
dw@rocksvax.UUCP (Don Wegeng) (06/03/85)
I have one problem with Phil's postings on this subject. Phil assumes that
all aspects of what makes one audio component sound different (I resist
using the term "better") can be measured using current techniques and
equipment. While I certainly have no scientific evidence to prove him
wrong (and he won't let me use my personal experience to support my
argument) it stands to reason that there are characteristics of sound
reproduction/electronic equipment/anything else you want to name that
either due to technical reasons or creative reasons (ie. nobody has
thought of it yet) we do not measure. To assume that we have already
figured out how to measure, and in fact *do* measure every trait that
can have an affect on the reproduction of sound seems unrealistic to
me. If this were true then all of the researchers who are trying to
improve our audio equipment could go find more useful things to do, for
the problem would have been reduced to one more suited to practical
engineers to solve.
I'm not going to get into the old argument of whether all of these
factors are measurable, because I suspect that we still haven't
figured out what subtle effects may be caused by the characteristics
that we *can* measure today.
/Don
--
"Old age and treachery will overcome youth and skill."
arpa: Wegeng.Henr@Xerox.ARPA
csnet: Wegeng.Henr@Xerox.ARPA
ns: Wegeng:Wbst207V:Xerox
uucp: {allegra,amd,decvax!rochester,princeton}!rocksvax!dwmwm@ucbvax.ARPA (Mike (I'll be mellow when I'm dead) Meyer) (06/03/85)
In article <361@petrus.UUCP> karn@petrus.UUCP writes: >I stick by my original assertions: > >1. Any audible differences between CD players (assuming proper functioning >of each, i.e., a good stream of bits being presented to each D/A converter) >remain to be proven. Correct: remains to be proven. Not "doesn't exist". One of these days, I'm going to decide for myself >2. The worst CD player (assuming no outright malfunction) is far better, in >every possible way, at the specific job of reproducing the exact signal at >the output of the studio console mixer than even the best LP player. You >may not like how the studio mixer output sounds, but that's not the CD >player's fault. Ah, but if you apply your previous criteria (audible differences), then a good LP/tonearm combo is *as good*, if not better, than a CD player. Yes, the CD specs to 22KHz are much better, but the distortion in a good analog system is inaudible anyway (excluding the record surface, of course). Likewise, the audio system specs are *much* better above 22KHz, which may or may not be audible. <mike