[Sonnova]

On Mon, 4 May 2009 09:57:13 -0700, wrote
(in article ):

On May 4, 11:00*am, wrote:
On May 3, 8:40*pm, Sonnova wrote:

On Sun, 3 May 2009 17:11:39 -0700, wrote
(in article ):

On May 3, 8:09*am, wrote:
On May 2, 2:08*pm, "Gary Eickmeier" wrote:

But it is an issue with reproduced music. The goal should be realism,
not
"accuracy."

But "realism" isn't definable, whereas "accuracy" is easy to define
and gives us something concrete to work towards.

I don't see why ease of definability should in any way affect our
aesthetic goals.

Getting recordings to sound "real" on accurate speakers should be the
business of the recording engineer, not the loudspeaker system.

What exactly are "accurate" speakers? It seems you are building an
argument on a mythological creature that is probably not something
everyone would agree on.

I would think that an accurate loudspeaker would be one which faithfully
reproduces the waveform with which it is fed. The extent to which any
loudspeaker accomplishes that goal is a measure of its accuracy.

A loudspeaker is fed an electrical signal that has only one dimension
in time. You can define any electrical signal in audio by time and
amplitude. When a speaker converts that signal to a sound wave it does
so into a three dimensional sound space. so how does one determine
which speaker has the most "accurate four dimensional wavefrom when
using a two dimensional waveform as a reference?

Sonnova's definition of accuracy is woefully
incomplete, to be sure. That such simple
definitions are inadequate is something that's
been known to acousticians and phyiscists for
a very long time.

Woefully incomplete? You mean that a speaker that reproduces the waveform
that it is fed exactly, without either adding anything or taking anything
away, or changing anything in any way wouldn't be an accurate transducer?
That's like saying that an amplifier that exhibits infinite, absolutely flat
frequency response and zero distortion, IOW, Stewart Hegeman's mythical
"straight wire with gain", wouldn't be an accurate amplifier.

But, by the same token, your objections are
based on some misunderstandings of that physics.

First, to get one step closer to a reasonable
definiiton of "electrocacoustic accuracy," you
must specify that the transducer will reproduce
an instantaneous sound level pressure that is
a direct linear function of its instantaneous
electrical input to within the specified tolerances
in a specific region of space positioned relative to
a reference axis as defined by the manufacturer of
the transducer.

IOW, precisely reproduce what it is fed.

Oh, by the way, this is, in essence, the way
agreed upon standards such as IEC 60268-10
deal with the problem.

There's clearly more to specify, but it deals with
the major objections raised.

However, those objects, as I mentioned, are
flawed, in that the reference position, while it
is in three dimensional space, to me (correct
me if I am mistaken) seems to imply there's
more information in the signal than simply its
instantaneous voltage (or, once transduced)
pressure vs time, such as directional information.

There might be. But whatever else is there, its the result of the interaction
of various electrical signals which are analogous to the complex wavefront
converted by the microphone. That the microphone is also an imperfect
transducer is not important to this discussion, because irrespective of
equipment flaws up the pipeline, the ideally accurate loudspeaker would
precisely reproduce those flaws as well. IOW, the speaker would still be
reproducing the waveform with which it is fed. The room doesn't enter into
the equation, nor does the electronics ahead of this "perfect" and "accurate"
loudspeaker.

And the problem is simply this: there is NO instrinsic
directional information of ANY kind in a stereo
recording. Microphones convert instantaneous sound
pressure into instantaneous voltage. And sound
pressure contains NO directional information: it
is simply the sound pressure ata point in time and
space. Period.

As a person who spent many years recording a major symphony orchestra in
concert, I can attest that your above statement (as I understand it) is
simply wrong. Human hearing is designed to pick up on spatial and directional
cues as a survival tool. We are very good at it. Those cues ARE, indeed,
captured by microphones in the form of phase shifts, time delays and
intensity differences. As you say, these things are the result of the
conversion of instantaneous sound pressure into analogous instantaneous
voltages. But these voltages are the result of the sum of the pressure waves
hitting the microphone diaphragm and the voltage produced as a complex
waveform contains all of those spatial cues. When played back through two
speakers, some semblance of those phase, amplitude and temporal cues are
reproduced into the listening room. On a really well miked recording, one can
literally point to each instrument in the ensemble and say "The oboe is§
right there, and the chorus sounds like it's behind the orchestra on a
platform. Some recordings are literally THAT "holographic." I have several
recordings made in the Watford Town Hall in Watford England that actually
sound as if the room is "L" shaped! I got that sense just listening. When I
finally saw a photograph of a recording session in that hall in a magazine, I
was flabberglasted to find that it was, indeed, "L" shaped! So, there is most
assuredly directional information in a stereo recording. (of course, I might
be misunderstanding what you are trying to say here.)

As a corollary, that point on the reference access,
where that sound pressure is supposed to be
reproduced to one degree of accuracy or another,
well, the sound pressure there is an equally vectorless
quantity.

Stereo does not make it much better: the very physics
of the process prevent any reasonably accurate encoding
of ANYTHING other than the instantaneous conversion
of vectorless sound pressure into vectorless voltage and,
eventually, back again.

The inability of stereo to capture any true directional
or spacial cues was demonstrated rather soundly
well over a half century ago at Bell Labs, and no
amount of protestation and vigorous claims from
the high-end audio community or from big manufacturers
living on mountaintops have demonstrated otherwise.

Stereo is one big friggin' auditory illusion, everybody
who has ANY experience in the field knows that.
It's an illusion: it's smoke and mirrors and requires
the suspension of objective sensory-driven skeptisim
to work. Get over it and move on.

Seems to work fairly well, given it's limitations. And yes, as far as the it
goes, it is an illusion. But good stereo can and often does present the
illusion of a fairly accurate soundstage.

So, I would suggest a different formulation of the
two questions in to one: How can objective
electroacoustic accuracy assist in serving up
and supporting the auditory illusion of stereo?

I think the original question was much simpler than that: How would one
define a theoretically accurate speaker system?