"Audio Empire" wrote in message
...
On Tue, 6 Mar 2012 16:28:58 -0800, Arny Krueger wrote
(in article ):
I have come across several references to an article on the topic of
so-called hi rez audio files
recently. It found it interesting - as is the site whence it came.
"24/192 Music Downloads ...and why they make no sense" is he
http://people.xiph.org/~xiphmont/demo/neil-young.html
Well written article and there's a lot of truth to it. However in his zeal
to
make his point, the author has made a few assumptions that I don't think
are
really in evidence.
1) In spite of the author's premise, I really don't think that ANYONE
expects
a digital system frequency response beyond 22.5KHz to actually have any
(or
at least not very much) actual program content.
I just answered a post from a fairly widely respected member of a pro audio
forum who was bragging about how much 20 KHz content he was seeing in
recordings.
(a) Few microphones have any response much above 20 KHz and most large
capsule mikes of the type generally used for recording have a huge
resonance
peak somewhere between about 8 and 16 KHz and drop off rapidly above that.
He was rebutting this exact issue. FWIW I agree with you.
(b) Those in the know realize that sampling rates of 88.2, 96, 176.4, 192,
and 384 KHz have merit simply because they move the Nyquist frequency WAY
out
of the audio passband and that 88.2KHz is probably quite far enough with
96
and above merely being overkill.
Let's face it, digital filtering has improved to the point where such
humungeous (2 cotave) guard bands serve no purpose at all.
2) His characterization of "golden ears" shows a basic lack of
understanding
of the actual meaning of the term. He's right that nobody has the kind of
super hearing that he characterizes as being the definition of the
golden-eared audiophile. But it's a strawman argument because "super
hearing"
is not what being "golden-eared" is all about. Most audiophiles these days
are over fifty. They certainly do not have super hearing. Many can't hear
much over 12 KHz, and if they have been exposed (either in their careers
or,
more likely, being exposed to loud electronic rock-n-roll in their youth)
they may not hear that wide a frequency response, and yet they still
posses
"golden ears". How is that possible, I hear some of you ask?
My answer is: denial. Case in point. A few weeks ago I was sitting next to a
reasonably well-known *name* in high end audio at a listening session. I was
complaining vigorously about the audible hum and noise. A number of people
around us shared my concern but Mr. name said that he heard nothing wrong.
Because all the
term "golden ears" means is that some sound enthusiasts care enough about
the
sound of music to have trained themselves to listen for artifacts in
reproduced music and the equipment used to reproduce it and to identify
and
quantify those artifacts. Things like ragged frequency response, different
types of distortion and their origin, problems in recordings and the
ability
to tell a real stereo recording from a multi-channeled mono one. That's
all
it is.
I'm on both sides of this argument. To some degree effectiveness at audio is
about both raw ability but it is also, and perhaps more preeminently as you
seem to be saying, about identifying the various sounds in what you hear.
However, through the magic of sighted evaluations there is also a large
subset of audiophiles and professionals whose hearing seems to be mostly in
their wallets. Any time you want to you can expose them with a blind test,
whether stealthy or in the open. Many have become too crafty to trap that
way. It's not a homogenious world out there.
When my friend J. Gordon Holt was in his late 70's, he could still
listen to a stereo system and tell you exactly what was wrong with it!
Things
like: the phono cartridge is mis-tracking, the speakers have boomy bass,
the
room has standing waves, The amplifier has high-frequency distortion, the
speakers are out of phase (a common one), etc., etc., etc.
Of course, and us blind testing advocates agree that a lot of problems like
those don't need blind tests to identify or prove. These can all be faults
and artifacts that are well above the well-known thresholds of audibility,
or not.
3) He is right about 24-bit. It is a much better CAPTURE format than is
16-bit simply because it allows the recordist more headroom. 32-bit
floating-point recording is even better. The ideal LPCM capture format
would
be 32-bit/88.2KHz. It moves the Nyquist frequency well outside the
passband,
and it gives the engineer lots of headroom.
On balance, if you know what you are doing you can make great-sounding
recordings with 16/44. 24 bits gets you 144 dB dynamic range, but in fact
audio gear that performs at the even *just* 20 bit level is still not sold
on every street corner. There are few live venues and recording studios that
have even 13 bits of acosutical dynamic range.
4) What he really should be attacking is the high-resolution audio content
provider's' insistence on using FLAC to deliver the files. FLAC is so
flawed
that the way that these sites use it, it's anything but lossless!
I don't know about that. In my tests and usage FLAC has proven itself to be
bit-perfect and sonically ideal. However, I have to admit that its
relatively gentle approximately 2:1 compression is not why I like it or use
it. What I like about FLAC is its support for tagging that thoroughly
eclipses what legacy .wav files support.
Otherwise, the article is good reading and should be de riguer for anyone
contemplating purchasing high-res downloads.
Especially when combined what is known about 50% of extant so-called hi rez
recordings having been through low-rez (typically analog recording)
production steps that limit their as-delivered performance to something like
12 bit resolution and CD format bandpass, but with more response variations.