trb@drux3.UUCP (Buckley) (02/08/84)
Dear Mr. Johnston,
This is a long letter, but I hope it will explain why I hear what I
hear, what I listen to, and why I believe the majority of people in
this discussion never understood what Phil Rastocny was trying to
say in the first place.
I'm not going to challenge your technical qualifications. The very
fact that you're an engineer with Bell Labs should say enough. I do
fail to see, however, what your telephone digital signal processing
work and high-fidelity dsp have in common in terms of this
discussion. And even so, my belief is that it doesn't take an
engineering whiz to evaluate high-end audio gear, specifically the
differences between analog and digital. All it takes is a good ear,
and a knowledge of what to listen for.
I used to be a trumpet player. I have a friend that custom-made
trumpets for a living. They cost about $1200.00 each, and were
mechanically perfect. He knew an amazing amount of physics of how a
trumpet worked, what metals made what sounds, and what method of
tempering the bell produced what tones. There is probably no one
any MORE knowledgeable about the trumpet technically than he.
However, he could barely play the thing. No real talent for music.
But he DID have a good ear!
Friends of mine, however, who are professional trumpet players
barely know anything about the trumpet other than how to oil the
valves. Yet, they know what SOUNDS right, and one friend tried
THIRTY-SEVEN trumpets right off the assembly line at Schilke, Inc.,
before he found the one that sounded right. All were consecutive
serial numbers, all made with the same materials and manufacturing
processes. Yet, they did sound different (I did a test like that
myself, once). Again, it didn't take a scientist to figure out what
sounded right.
You questioned whether I REALLY knew what live music sounded like.
Well, here goes. I DO regularly listen to live music. I was both a
professional and a student musician, and many times a week I get to
hear live music in many varied settings. I KNOW what a trumpet
should sound like. I KNOW what a cymbal and high-hat should sound
like, and I can tell the difference between some of the top brands
of cymbal makers. And, if you want to start throwing questionable-
relevance qualifications around, the hearing in my left ear rolls
off at 17.1 kHz, and my right ear at 19.2 kHz. I also have perfect
pitch.
Here in Denver is an old, restored movie house, called the Paramount
Theatre. It's now used mainly for concerts. The Paramount is one
of the most acoustically perfect buildings in existence; possibly as
good as Carnegie Hall. It is the only other theatre in existence,
along with Radio City Music Hall, to have the twin Wurlitzer organs
still intact and functional.
One of the world's biggest jazz impresarios, Dick Gibson, lives in
Denver and puts on a series of Gibson Jazz Concerts in the
Paramount. There are six a year, and I have been going (along with
Phil Rastocny and others) for the past four years, in the 4th row,
center seats; seats which I specifically requested after having been
to a number of various concerts there. The amplification at these
concerts is very subtle, and is the best amplification of acoustical
instruments I have ever heard.
In these concerts we have heard an incredibly wide variety of jazz
musicians, and I personally have entered into another dimension of
realization of what instruments really DO sound like.
There's nothing like the listening experience of hearing two
different sounds come out of the SAME drum set, as you compare Louis
Bellson with Shelly Manne. Phil Woods has a unique alto sound that
is recognizable anywhere, and Bill Watrous a tone quality unmatched
by any other trombone. After a while, you get an appreciation of
what each instrument should sound like, and even the differences, no
matter how subtle, between musicians of the same instrument (I'm
talking sound here, and not style).
You suddenly realize the wide array of harmonics that can come out
of trumpets and saxophones, and you suddenly hear a new musical
experience when something is played at very low volume levels, say,
for instance, Scott Hamilton in a slow part of a ballad, or Harry
"Sweets" Edison playing his trumpet barely audible with a harmon
mute and no amplification.
So, when I sit down to listen to one of my 1300 some jazz and
classical albums, I know what to expect. I can take one record of a
musician whom I am familiar with, Sarah Vaughan for example, and
play it on many systems until I hear it, that TRUE live sound. I
was never brought up on stereos and records and FM radios. I was
brought up in a musical family, and have been surrounded with live
music all my life, and I know what something SHOULD sound like, not
just what we LIKE it to sound like. I have attended more live
concerts and performances in more places in my life than many
professional musicians ever play at. FM radio and top-40 albums
have made us a generation of boom-chinkers, where all we like is
bass and treble, and to hell with the fidelity, the soundstage, and
the sonic accuracy.
Now, on to why digital has some problems, and why current state-of-
the-art analog is still FAR superior! I have heard Harry James live
many times. Harry had a full, unmistakable sound. The ONLY
recording I have EVER heard that captures the trumpet EXACTLY how it
sounds live is the Sheffield Lab "King James Version." And what
makes a trumpet sound like a trumpet is the high end harmonics. A
good trumpet player could play into a spectrum analyzer and you
would see all kinds of clean, high-amplitude harmonics. A good
trumpet player, in the proper acoustical surroundings, could play a
concert b-flat, and you could hear at least 3 octaves of b-flats
above the dominant.
Unfortunately, there wasn't a digital recorder there to record this
at the same time, so we can't make comparisons. But, once you get
the mental "picture" of what something should sound like, while
listening regularly to live music to prevent losing that reference
point, you can then evaluate ANY recording and ANY sound system as
to their accuracy.
Regular analog recordings rarely have accurate soundstage, due to
recording engineers who don't know how to place mikes in a studio or
concert hall. But digital recordings NEVER have accurate
soundstage. They're always very flat, although admittadly, some are
better than certain analog recordings, which sound like one mike was
used. The only album I have ever heard to correctly capture the
live soundstage is "For Duke" with Bill Berry and friends, a
direct-to-disc recording. All the digital recordings I have ever
heard, whether vinyl or CD, have no depth or "liveness" to them.
Someone far more technically knowledgeable than I could explain the
physics behind that; suffice it to say, I can hear the difference.
I have NEVER heard anything but a tinny, raspy high-hat and cymbal
on a digital recording. Even the best of the digitals, which are
generally far better than most analogs, have never had a proper high
hat sound. Then, try and listen to a low-amplitude section of a
digital recording. It loses ALL semblance of sonic accuracy, and
can be quite irritating to the ears. (Try listening to a digital
recording of Sarah Vaughan as she vocalizes on a soft, slow song.
Bleah!)
Jim, I have many digital records, and have heard a good amount of CD
systems. I like what I hear, generally. These recordings are
usually far better than most analog recordings of today. But,
state-of-the-art analog is still better, and will remain so for some
time.
I'm not attacking you personally at all. I'm certainly not going to
debate technical matters with you, I'm surely not as knowledgeable
about digital signal processing. All I'm trying to say is, you
don't seem to know what to listen for. I doubt that you've begun to
hear what I have, or ever will. I don't think you, or any other of
the contributors, have that mental image of what things SHOULD sound
like. I don't compare stereo to stereo, I compare the stereo to
LIVE PERFORMANCES!
I believe this is part of what Phil was trying to say. Do what you
want with your electrical specs, but your ears are the final judge.
If you have the ability to capture the mental image of sound, your
ears are all you need.
-------------------------------------------------------------------------------
Tom Buckley
AT&T Information Systems Labs
...ihnp4!drux3!trbjj@rabbit.UUCP (02/09/84)
OK. Now I've got time to be a bit more specific. 1) I'm not JUST in the business of trying to figure out if something sounds realistic, I'm also in the business of trying to find out WHY something does or does not sound realistic. That's why my digital experience matters. You're welcome to tell me what you THINK you hear, but if you can't help me find out what it is, and accept that some things have been proven NOT to be what you think you hear, then you aren't willing to help me further the state of the art. On the other hand, if you can show me that a supposedly unimportant factor is what you hear, I'll be quite interested, to put it mildly! <Phase response is one of those things that were "unimportant" for years. Hah!> 2) I damn well DO know how things are SUPPOSED to sound. I don't know why you can even begin to presume on anyone else's experience, especially if you don't know anything about it, and I can't see where you figure that you should explain your opinions to the world if you are acting on mere, and inaccurate supposition. About myself: I do NOT have perfect pitch. My HF threshholds are 16.6K in left ear <thanks to car windows, mostly> and 17.9 in my right ear. I don't know how old you are, but that alone could account for the differences in HF response, which don't matter as much as you think. Please see previous articles on intermodulation in the ear for the reasons why. <This same phenominon is a good argument for higher sampling rates, but I don't see it cropping up ANYWHERE. If you want to know more, mail me, TB.> I am also somewhat of a musician (I haven't had as much time to practice and perform as I'd like, but what else is new?) I play mostly anything in the flute family, and I have more than a passing acquaintance with pipe organ building, theory, and repair. <Mostly baroque, I really dislike the bland romantic monstrosities.> (I wanted to play brass at an early age, but one look at my dental work will show why I never did, sigh.) I go to live performances fairly often, both for enjoyment and for learning/working/helping purposes. I think that I can also tell you what's good/bad about a recording. I also have a fair amount of experience with recording and miking problems, <See my complaints about miking and mixing elsewhere. grrrrrrr> as well as acoustic problems caused by halls, stages, etc. I don't have perfect pitch at all. I have fairly good relative pitch, i.e. I can judge an octave to the point where I can hit a half beat/second, while only hearing ONE tone at once. <Not beating them.> I suspect that won't impress you, but then again, perfect pitch doesn't have anything at all to do with trained listening, either. <In fact, I can't see where it says anything about listening, other than showing that you do have a good auditory system, with a somewhat unusual hookup, so to speak.> I think you would do better to describe what you hear, represent it as your own opinion, and try to HELP people who are trying to measure what you hear. <I've done enough double blind testing to see people confuse themselves thoroughly when they couldn't find out which was the reference system. I also know a few people who are marvelously accurate. They are also almost all either recording or reinforcement types who can also tell me what they hear or musicians who can tell me when I duplicate something.> 3) I don't see why you think that insulting people <and, yes, when you're representing yourself as an expert and calling someone else, especially someone who makes his living that way, an incompetent amateur (which is what you imply) it's an insult, at least> isn't a personal attack. It is in fact a personal attack, and I certainly would like to see a retraction. Mail me, if you must continue this slander. I'd rather work for better listening. -- ENGINEERS ARE PEOPLE, TOO! (allegra,harpo,ulysses)!rabbit!jj