The double edged sword of loudspeaker accuracy.
rjs...@calcube.com
threads, I'm inclined to challenge the whole notion of "accuracy."
I'll confine this response to loudspeakers only, since they
ultimately represent the tangible link to sound reproduction. My
perspective on this subject is one initiated from being a loudspeaker
designer and also a music aficionado. Biased perspective? Maybe.
The term accuracy means exactness, correctness, or freedom from
error. In the context of further discussion, this is the definition
of accuracy I'm going to use as a *frame of reference*. This
definition represents a readily understood concept with universal
acceptance.
According to Miriam-Webster (dictionary), the term loudspeaker means:
"a device that changes electrical signals into sounds loud enough to
be heard at a distance." I think that definition is pretty
straightforward and unless one resorts to semantics, most individuals
of a technological discipline would find this definition to be
accurate and universally accepted. I'm buying into this definition
for further discussion.
So we've now established a *frame of reference* regarding the terms
"loudspeaker" and "accuracy." We understand the terms, accept the
terms, and even apply them when discussing audio or music. THEN WHY
IS THERE SO MUCH CONFUSION WHEN WE COMBINE THE TERMS?! Why is it that
two words with concrete definitions lose all significance when
combined? What is it about the term "loudspeaker accuracy" that has
so many people confused about deriving a universally accepted
definition? Well, I'll tell you: THERE IS NO SUCH THING AS
LOUDSPEAKER ACCURACY! That's right, there's no such thing as
loudspeaker accuracy because the *frame of reference* for loudspeaker
accuracy is a dynamic one, not a static one. Loudspeaker accuracy is
a chameleon that changes colors between measurements and realism.
Measurements being the objective color while realism the subjective
color. Or is this assumption correct? Depending upon the environment,
either of these two colors may be accurate....or not. Let's take a
closer look.
Measurements. As a loudspeaker designer, when I think of
measurements I think of objectivity. I think of measurements as an
objective frame of reference to consistently improve upon existing
iterations leading one to theoretical perfection. With the exception
of a healthy newborn child, there is no real perfection. Measurements
are simply experiments performed in controlled environments. So why
are loudspeaker measurements performed in controlled environments
rather than the 'real world' environments that we listen to them in?
Let's go back to the original definition of loudspeaker: "a device
that changes electrical signals into sounds loud enough to be heard
at a distance." The operative phrase in this definition is "changes
electrical signals into sounds." Loudspeaker accuracy in the context
of measurements means designing "a device" that does not alter the
characteristics of the "electrical signals." Is this possible? Not
yet. Can we get close? Maybe, but it must be realized that every
piece of material used to build such "a device" deviates from the
theoretical perfection initially calculated, thereby never realizing
real world accuracy (defined as freedom from error). It is this
deviation from perfection that has allowed so many interpretations of
loudspeaker design to proliferate. I.E., every loudspeaker designer
thinks they have *the answer*. The digital age may prove different,
but with state of the art as it is today, there is no perfect
"loudspeaker accuracy" as it pertains to measurements, only close
approximations.
Realism. So what is realism? Without getting too detailed, or
conversely too vague, realism in loudspeakers is simply the
subjective illusion that somehow one is transported in time and space
to the *original event*. Which original event? A poetry reading at a
coffee shop? A vocal booth in a recording studio? A concert at
Carnegie Hall? A night of rock at the Roxy in L.A.? No matter what
the venue, we demand that our loudspeakers transform our listening
rooms into the real event, albeit with a little imagination (close
your eyes for reality). This imagined participation with the
*original event* is the other definition of accuracy. Realism. The
difference between this accuracy and the accuracy derived from
measurements, is that accuracy in the form of realism takes on many
different personalities that conform to the eye/ear of the beholder.
Realism can be timbre, ambient expanse, presence, localization, etc.
Is it possible for any loudspeaker system to conformally approach
everyone's definition of realism? No, but some seem to come closer
than others. Is it possible for multiples of loudspeakers (more than
2 channel stereo) to recreate realism? No cigar yet, but getting
better. Again, the digital age may yield breakthroughs towards an end
of electronic tranparency, but the fact remains that perfection is
elusive in the real world.
So where are we now? If there is no perfection, and the term
loudspeaker accuracy can be loosely translated to fit multiple
definitions, where do we draw the line for a *frame of reference* of
good or bad? Do these multiple definitions render magazine reviews
worthless or priceless? How as industry professionals or consumers
can we delineate what is the correct *frame of reference* for product
judgement? Are realism and meeasurements mutually exclusive entities,
or are they tightly interwoven?
'Realism' and 'measurements' have often spawned debate within the
audio community as to the possible notion that both theories
represent separate camps of ideology. 'Realism' has been associated
with a "subjectivist" camp while 'measurements' have been relegated
to an "objectivist" one. Both camps posture that their way of judging
sound reproduction is the more correct methodology over the other.
Like most other bastions of fact or opinion such as politics,
religion, or sports, there's a scale of balance in loudspeaker
accuracy with varying degrees of deviation from equilibrium.
Equilibrium defined as that point in which realism and measurements
carry equal weight (in the mind of an individual) in the
determination of loudspeaker accuracy. On one side of the scale we
have 'measurements' (conservative, objective?), and on the other side
we have 'realism' (liberal, subjectivist?) It is this scale that we
use to judge a loudspeaker's performance. For the consumer, other
factors such as company reputation, price value perception, word of
mouth, etc. come into play, but ultimately the quantifiable and
qualifiable traits of a loudspeaker will be the final arbiter in the
decision making process.
How viable are magazine reviews in determining loudspeaker accuracy?
Well, most magazine reviews do present an objective and subjective
profile of a loudspeaker system and they employ competent individuals
to effectuate the review. Competent in the respect that the reviewers
have been exposed to a much greater cross section of products than
the average individual, so they do have a pretty good feel for a good
or bad product. And for those reviewers that perform measurements,
they're knowledgable about their measuring equipment and the
interpretation of their results. So is a magazine review the last
word in loudspeaker accuracy? Nope. A magazine review represents a
floating point frame of reference for product performance on two
fronts:
A) Measurements. Quite frequently a magazine will have different test
equipment and a different measurement venue than a manufacturer. This
can and does lead to a difference in results between both parties, as
sometimes witnessed by a follow up correspondence from the mfr. in
which they include their own in-house measurements. Does this negate
the integrity of the review? No, because all loudspeakers coming
through the door of the magazine are subject to the same scrutiny on
the same equipment.
B) Realism. No two individuals are exactly alike, so don't expect two
reviewers to have the same opinion of a loudspeaker. A reviewers
biases, hearing, listening room, associated equipment, etc. all
affect the review of a loudspeaker. Does this negate the integrity of
the review? No, because the reviewer will usually maintain the same
character from review to review.
So while reviews may be a valuable tool in the evaluation of a
loudspeaker, realize that the floating point frame of reference can
only accurately compare product to product, not product to benchmark.
And opinions, no matter how qualified, are like..........well, you
know the rest.
In closing, if you've never been to a recording studio during mixdown
or mastering, you should go. Seeing the process that creates the
software we listen to will speak volumes about realism. When one is
exposed to the process, their opinions about what is "real" will be
greatly skewed. Quite frequently, the voice you hear in your
listening room and the one in the studio vocal booth don't exactly
match. And if you ever get the chance to visit an anechoic chamber,
you'll appreciate why loudspeaker designers get so nitpicky about
measurements. So if we can't trust measurements or realism, what do
we trust? Our OWN ears. After the opinion and/or after the math, we
(as designers or consumers) ultimately choose what we like or deem to
be 'right.' The question of loudspeaker accuracy will continue to be
a hotly debated topic with facts and opinions eclipsing the magnitude
of the audio frequency spectrum, and as long as people are different,
loudspeakers will be too. Fortunately, this freedom of choice and
opinion advances the art to higher levels in which we all participate
and enjoy. That's it, I'm out of breath.
--
Robert J. Salvi, Ambiance Acoustics
http://www.calcube.com
San Diego, CA USA
619-485-7514
SDuraybito
>So we've now established a *frame of reference* regarding the terms
>"loudspeaker" and "accuracy." We understand the terms, accept the
>terms, and even apply them when discussing audio or music. THEN WHY
>IS THERE SO MUCH CONFUSION WHEN WE COMBINE THE TERMS?! Why is it that
>two words with concrete definitions lose all significance when
>combined? What is it about the term "loudspeaker accuracy" that has
>so many people confused about deriving a universally accepted
>definition? Well, I'll tell you: THERE IS NO SUCH THING AS
>LOUDSPEAKER ACCURACY!
It is instructive to note that our audio forefathers coined the
phrase "high-fidelity" as opposed to "high-accuracy." They wisely
understood that, unless reproduced sound could be determined to be
indistnguishable from live, ALL audio is inaccurate. Hence the use
of the word "fidelity," which puts the burden upon the listener to
believe or "have faith" in the ability of reproduced music to
resemble live music as much as he or she can believe.
In short, loudspeakers are not accurate but they do have fidelity.
Siegfried
-- Opinion: a belief stronger than impression and less strong than positive
knowledge --
Bob Trosper
that you seem a little overexcited. As I understood them your main
points were that :
1) Accuracy is defined by some agreed upon transformation of the
electrical input signal to sound and is measurable objectively
2) Realism is defined by some impossible to agree upon transformation
of the sound produced to a musical experience for the listener and is
not measurable objectively
3) Reviewers are self consistent and if you find one you agree with
you can use their experience as a reliable subjective measurement 4)
Listeners should choose what they want
5) Production decisions in the control room or afterwards are based
on what the person making the decisions wants
OK. Sounds right to me. It seems to explain observable facts in the
audiophile world and doesn't seem to have any obvious internal
contradictions.
-- BT
rjs...@calcube.com
> -- BT
Well Bob, I do tend to get a bit emotional about what I do......<g>
You've pretty much captured the essence of the post, although I might
reword #1 to read:
"Objective accuracy is defined by some agreed upon benchmark of the
electrical input signal to sound and is measurable. The benchmark
being theoretical perfection, an unattainable quantifiable frame of
reference."
Other than that... I think next time I get long winded I'll send the
post to you first for editing;) You know how some people like to hear
themselves talk............or write.......
gins...@catskill.net
remarks take me back to the early days of what I call the 'modern
high-end era'. Early on, the so-called underground mags like The
Absolute Sound, Stereophile, International Audio Review, etc. often
spoke of an original, acoustical musical event as the only reference
to a component or systems' quality. The system (or component) honored
the music or it didn't. How it did, or didn't became the subject of
endless reviews and debate that continues to this day. My position as
an audio professional and music lover puts me squarely in-between
both the 'objectiveness' and 'subjectiveness' camps. I must measure
to make sure the obvious design problems have been dealt with. I
listen to refine my design path. For example, speaker systems can be
designed (with high quality drivers) to project sound into a given
space with so-called flat power response. That is, no frequency or
group of frequencies will be louder than any others. However, my ears
tell me it's better to have gentle transitions between the multiple
drivers that I have chosen, and my engineering background clues me
into such things as phase shift and transient response. So I use
6dB/oct. slopes (after the subwoofer low pass which is 18 dB/oct.)
even though there would me no measurable difference had I used higher
order slopes. Now, if I closely examine the last statement, I'd have
to admit that while the room power response might not show a
measurable response difference with slope change (remember, I'm using
top notch drivers crossed over well within their bandwidth limits)
there are plenty of ever-more sophisticated tests that will show the
changes, if only I could afford them......So right now, my ears are
the most cost-effective way to converge on a design that gives me
musical satisfaction.
I guess what I'm getting at is my feeling that the pure
technologists, those who believe that there are now enough
sophisticated measurements to explain everything, leave me feeling a
bit cold. It seems like all so much hubris. On the other hand, I like
the fact that there are people like Mr. Pierce who quite properly
demand to see the science behind some of the more, ahem, interesting
performance claims that some vendors or end-users make. I personally
feel that, at a minimum, the science has to be right; after that, we
will see......
I do know this; sometimes it takes the subjective listeners to alert
the objective scientists to look deeper to enable them to come up
with measurements to explain widely-held beliefs. Remember the days
when a capacitor was a capacitor? Then came that seminal Audio Mag.
series (Marsh, et.al.) and the rest is history.
--
David Ginsberg
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
John Ongtooguk
: .....THERE IS NO SUCH THING AS
: LOUDSPEAKER ACCURACY! That's right, there's no such thing as
: loudspeaker accuracy because the *frame of reference* for loudspeaker
: accuracy is a dynamic one, not a static one. Loudspeaker accuracy is
: a chameleon that changes colors between measurements and realism.
: Measurements being the objective color while realism the subjective
Why stop with loudspeakers if you're willing to tolerate any
coloration of reproduction as long as it sounds pleasing ? How
about THERE IS NO SUCH THING AS AMPLIFIER ACCURACY, CD PLAYER
ACCURACY, TURNTABLE ACCURACY, PREAMPLIFIER ACCURACY, MICROPHONE
ACCURACY, OR ANYTHING ACCURACY ? Loudspeakers are no different
than any other component when one is attempting to reproduce
audio signals as faithfully as possible. If you look at some of
work being done in real 3D audio, work that offers some of the
best models on what's important for accuracy, a basic assumption
that is used is that the recording and playback is a linear
system as one needs to preserve the impulse response(s) of
the recording space or the space that one is attempting to emulate.
Typically compromises are made due to complexity and/or
computational difficulty, but the ideal is still one of preserving
impulse responses as that is what is used to define a specific
listening environment. While stereo is not 3D audio the same
principles apply, at least if one is interested in faithful
reproduction, which is preserving signal integrity to the
best of one's ability. To accomplish that it's obvious that
phase coherent speakers are required.
John Ongtooguk (jo...@vcd.hp.com)
David Wareing
>rjs...@calcube.com wrote:
>: .....THERE IS NO SUCH THING AS
>: LOUDSPEAKER ACCURACY! That's right, there's no such thing as
>: loudspeaker accuracy because the *frame of reference* for loudspeaker
>: accuracy is a dynamic one, not a static one. Loudspeaker accuracy is
>: a chameleon that changes colors between measurements and realism.
>: Measurements being the objective color while realism the subjective
>: color....
> Why stop with loudspeakers if you're willing to tolerate any
> coloration of reproduction as long as it sounds pleasing ? How
> about THERE IS NO SUCH THING AS AMPLIFIER ACCURACY, CD PLAYER
> ACCURACY, TURNTABLE ACCURACY, PREAMPLIFIER ACCURACY, MICROPHONE
> ACCURACY, OR ANYTHING ACCURACY ? Loudspeakers are no different
> than any other component when one is attempting to reproduce
> audio signals as faithfully as possible. If you look at some of
> work being done in real 3D audio, work that offers some of the
> best models on what's important for accuracy, a basic assumption
> that is used is that the recording and playback is a linear
> system as one needs to preserve the impulse response(s) of
> the recording space or the space that one is attempting to emulate.
> Typically compromises are made due to complexity and/or
> computational difficulty, but the ideal is still one of preserving
> impulse responses as that is what is used to define a specific
> listening environment. While stereo is not 3D audio the same
> principles apply, at least if one is interested in faithful
> reproduction, which is preserving signal integrity to the
> best of one's ability. To accomplish that it's obvious that
> phase coherent speakers are required.
Many, perhaps most, things in audio are non-obvious or
counter-intuitive. It is quite common to use minimum-phase versions
of measured head related impulse responses. See for instance the book
by D. Begualt "3-D Sound for virtual reality and multimedia" Page
161.The well known MIT Kemar Compact data is an example of this. It
may not be ideal but it is not thought to matter greatly.
Of course in the case of stereo, when the "phantom" image is in the
centre, the combination of ideal impulse responses _l_ from two
speakers at each ear yields one impulse from the nearest speaker to
an ear followed by a second, ( filtered by headshadowing) a fraction
of a millisecond later, from the speaker farthest from the same ear
_l_i_. The combination double pulse has a phase which is a function of
frequency and a magnitude response with peaks and notches you could
drive a bus through./\/\/\/\
This applies to a listener who is exactly central. If you turn your
head or move to the side, the path lengths change and since the
wavelength can go down an inch or so, you don't have to move far to
change the frequency reponse of the direct sound and it will then be
different for each ear..
Perhaps it is lucky that the direct sound FR is filled in to some
extent by the diffuse field FR from the sound that bounces around the
room which is of more or less random phase.
Accuracy?
David.
RLSpeakers
>"Objective accuracy is defined by some agreed upon benchmark of the
>electrical input signal to sound and is measurable. The benchmark
>being theoretical perfection, an unattainable quantifiable frame of
>reference."
It doesn't have to be perfect, only free of AUDIBLE distortion. The ear is
forgiving. If you get below a certain level of audibility, then the reproduced
signal is indiscernable from the same signal being reproduced by a theoreticaly
perfect speaker.
I don't know of any, although Velodyne claims one.
I think most folks would be agahst at how much distortion there is in
loudspeakers.
Mike McKelvy
Real Life Sound
http://members.aol.com/RLSpeakers/rlsindex.html
John Busenitz
> Why stop with loudspeakers if you're willing to tolerate any
> coloration of reproduction as long as it sounds pleasing ?
"Sounds pleasing" does not have to be an issue. If nonlinear
phase is not audibly unfaithful, then it is effectively NOT a
coloration.
> How
> about THERE IS NO SUCH THING AS AMPLIFIER ACCURACY, CD PLAYER
> ACCURACY, TURNTABLE ACCURACY, PREAMPLIFIER ACCURACY, MICROPHONE
> ACCURACY, OR ANYTHING ACCURACY ? Loudspeakers are no different
> than any other component when one is attempting to reproduce
> audio signals as faithfully as possible.
Nope. Loudspeakers have a 3-dimensional output, whereas electronics
operate only in one dimension. What then is the perfect loudspeaker?
Is the power response flat, or the magnitude response? You can't have
both. What about the radiation pattern? The low-frequency alignment?
Spherical or cylindrical radiation?
> If you look at some of
> work being done in real 3D audio, work that offers some of the
> best models on what's important for accuracy, a basic assumption
> that is used is that the recording and playback is a linear
> system as one needs to preserve the impulse response(s) of
> the recording space or the space that one is attempting to emulate.
Well, the frequency response more than the impulse response. And
a loudspeaker is NOT linear, whether it has linear phase or not!
> Typically compromises are made due to complexity and/or
> computational difficulty, but the ideal is still one of preserving
> impulse responses as that is what is used to define a specific
> listening environment. While stereo is not 3D audio the same
> principles apply, at least if one is interested in faithful
> reproduction, which is preserving signal integrity to the
> best of one's ability. To accomplish that it's obvious that
> phase coherent speakers are required.
No, it is preserving the audibly perception of signal integrity.
"Phase-coherent" speakers may or may not be a part of that. What
if, by sacrificing phase response, one can obtain better linearity,
better polar response, better frequency, response, etc? There is
no free lunch. By espousing linear-phase response, you are inherently
naming other factors as less important, factors which have been proven
to make a difference. Linear-phase response has not be established as
necessary for audible fidelity.
Howard Ferstler
> After following the "lesson in sound.." and "phase coherence"
> threads, I'm inclined to challenge the whole notion of "accuracy."
> I'll confine this response to loudspeakers only, since they
> ultimately represent the tangible link to sound reproduction. My
> perspective on this subject is one initiated from being a loudspeaker
> designer and also a music aficionado. Biased perspective? Maybe.
(Remainder of this post deleted to save space.)
Interesting post. I think that some basic, measurement-oriented
standards are required. For example, a good speaker or speaker
and subwoofer combination should have a room-measured bandwidth
that is able to handle the musical program material the listener will
be dealing with. Also, distortion levels have to be down to certain
levels, and I also believe that the overall power/room response of a
system needs to be flat enough to insure decent accuracy.
That leaves us with imaging, soundstaging, etc. and the way that a
system's radiation pattern will influence those aspects of performance.
I think this is where subjectivity comes into play. Some recordings,
when played back on certain kinds of systems ("kinds" referring to
their radiation patterns) in certain rooms, sound great. Other
recordings, played back on different kinds of systems in other rooms
sound great, also. Certain kinds of music (and recordings of that
music) sound better on narrow-dispersion, imaging-oriented systems
than they do on broad-soundstage, omni-directional radiating
systems. Other kinds of music happens to work better on speakers
with more diffuse characteristics.
In other words, I believe that there are recordings that would sound
better on something like a Dunlavy speaker or a Legacy Whisper
than on anything else. On the other hand (please sit down) there
are probably some recordings out there that would sound better
on a pair of Bose 901s than on anything else. Obviously, there are
recordings that would not sound too good on either, and there are
also systems that tend to split the difference between both of the
extremes mentioned, in terms of radiation patterns, and they might
do better on some recordings than either of them.
Once the basic, minimum standards I first mentioned are satisfied, I
believe that the evaluation of loudspeaker systems is truly a matter
of educated taste.
Howard Ferstler
fer...@ibm.net
John Ongtooguk
: Many, perhaps most, things in audio are non-obvious or
: counter-intuitive. It is quite common to use minimum-phase versions
: of measured head related impulse responses. See for instance the book
: by D. Begualt "3-D Sound for virtual reality and multimedia" Page
: 161.The well known MIT Kemar Compact data is an example of this. It
: may not be ideal but it is not thought to matter greatly.
Some do think that it does matter, as noted below, and even the
documentation for the Kemar data notes:
***************************************************************
ftp://sound.media.mit.edu/pub/Data/KEMAR/KEMAR-FAQ.txt
..... More precisely, the compact data is
equalized using a minimum phase inverse filter derived from
the Neumann measurement of the speaker (see Q15 and Q16), so
the compact data may contain phase aberrations of the
measurement system that are not necessarily part of the HRTFs.
**************************************************************
They also include the inverse impulse response of the speaker
and headphones for other corrections that some may want to do.
From a couple of thesis:
***********************************************************
http://alumni.caltech.edu/~franko/thesis/thesis.html
Frank Filipanits Jr., "DESIGN AND IMPLEMENTATION OF
AN AURALIZATION SYSTEM WITH A SPECTRUM-BASED TEMPORAL
PROCESSING OPTIMIZATION", Master's Research Project
Alternately, since convolution in the time domain is equivalent
to multiplication in the frequency domain, an FFT may be
performed, followed by multiplication by the desired filter
response (Fig. 2), followed by an inverse FFT [27][28]. It is
important, however, to realize the role of temporal and phase
cues in the HRTF; it is not accurately represented by a
magnitude (real-valued) filter [22]. The phase information
must be maintained through the use of complex FFTs and a
complex frequency domain filter.
http://www.isr.umd.edu/~cpbrown/thesis.pdf
C. Phillip Brown, "Modeling the Elevation Characteristics
of the Head-Related Impulse Response,"
... The Snapshot system also performs a minimum-phase
reconstruction of the measured HRIR. During the minimum-phase
reconstruction, the data is time-aligned and equalized for the
low-frequency roll-off that occurs in the sound source (speaker).
Unfortunately, while it does simplify some computations, the
minimum-phase reconstruction changes the HRIR wave form, and
might disrupt the timing cues. The data is finally stored as
a stereo, 44.1 kHz, 16-bit, 128 sample representation of the
HRIR.
...the software was changed to eliminate the
minimum-phase reconstruction, at the expense of having no
compensation for the low-frequency roll-off of the speaker.
...Early in the modeling process it was decided to base the
work in the time-domain. We consider the detailed timing
information presented in the HRIR to be critical in
developing an accurate model. A modeling approach in the
frequency-domain may yield an apparently close match in the
magnitude response, but reproducing the phase response is
difficult.
**********************************************************
: Of course in the case of stereo, when the "phantom" image is in the
: centre, the combination of ideal impulse responses _l_ from two
: speakers at each ear yields one impulse from the nearest speaker to
: an ear followed by a second, ( filtered by headshadowing) a fraction
: of a millisecond later, from the speaker farthest from the same ear
: _l_i_. The combination double pulse has a phase which is a function of
: frequency and a magnitude response with peaks and notches you could
: drive a bus through./\/\/\/\.......
:
: Accuracy?
Binaural reproduction using speakers seems to usually provide
attempts at crosstalk cancellation, which minimizes the comb
filtering that you describe. While the comb filter effect is a
problem with stereo it seems that it's minimized by using
careful recording techniques, 'minimalist's' is one term,
and by using phase coherent speakers, as otherwise you just
compound the problem.
John Ongtooguk (jo...@vcd.hp.com)
John Ongtooguk
: John Ongtooguk wrote in message <75me7r$1fk$1...@agate.berkeley.edu>...
: > Why stop with loudspeakers if you're willing to tolerate any
: > coloration of reproduction as long as it sounds pleasing ?
: "Sounds pleasing" does not have to be an issue. If nonlinear
: phase is not audibly unfaithful, then it is effectively NOT a
: coloration.
'Audibly unfaithful' is obviously subjective and will vary with
listening requirements. It appears that at least some cues that
are important for spatial hearing have reflections that need to
be reproduced faithfully down to a range of at least 50 us,
per some of the work noted below and in another reply, which
means that speakers would need to be 'phase coherent', either
as made or equalized as such.
*************************************************************
http://www.isr.umd.edu/~cpbrown/thesis.pdf
C. Phillip Brown, "Modeling the Elevation Characteristics
of the Head-Related Impulse Response,"
....A second peak occurs at tý50 us, which
varies somewhat with elevation, especially near f=ţ80ł. A
third peak (actually a "valley"), is very prominent. As
the sound source is moved from f=+80ł to f=-80ł, the
valley's position in time (with respect to the non-varying
initial peak) increases nearly-monotonically from tý100 us
to 300 us. A fourth peak (the third echo) is also apparent,
and it follows the same path as the valley (2nd echo),
offset by tý50 us to 100 us. Note the similarity of the
echoes in the near and far ears now that the head-shadow
and timing differences are removed.
...Early in the modeling process it was decided to base the
work in the time-domain. We consider the detailed timing
information presented in the HRIR to be critical in
developing an accurate model. A modeling approach in the
frequency-domain may yield an apparently close match in the
magnitude response, but reproducing the phase response is
difficult.
************************************************************
: ...Loudspeakers have a 3-dimensional output, whereas electronics
: operate only in one dimension. What then is the perfect loudspeaker?
: Is the power response flat, or the magnitude response? You can't have
: both. What about the radiation pattern? The low-frequency alignment?
: Spherical or cylindrical radiation?
If one wishes to record a range of useful signals at a designed
listening position, attempt to reproduce the signals, and then
re-record the signal for comparison, with the desire to have
a minimum of differences between the recorded signals, phase
coherent speakers will be required for faithful reproduction of
the waveforms.
: > Typically compromises are made due to complexity and/or
: > computational difficulty, but the ideal is still one of preserving
: > impulse responses as that is what is used to define a specific
: > listening environment. While stereo is not 3D audio the same
: > principles apply, at least if one is interested in faithful
: > reproduction, which is preserving signal integrity to the
: > best of one's ability. To accomplish that it's obvious that
: > phase coherent speakers are required.
: No, it is preserving the audibly perception of signal integrity.
: "Phase-coherent" speakers may or may not be a part of that. What
: if, by sacrificing phase response, one can obtain better linearity,
: better polar response, better frequency, response, etc? There is
: no free lunch. By espousing linear-phase response, you are inherently
: naming other factors as less important, factors which have been proven
: to make a difference. Linear-phase response has not be established as
: necessary for audible fidelity.
Phase coherent speakers or speakers that have been equalized using
an inverse impulse response are required for the faithful
reproduction of cues that we are capable of hearing. You may choose
not to reproduce such cues or to reproduce them poorly, a
compromise which is acceptable and even desired in many, probably
most cases, but it is still a compromise.
John Ongtooguk (jo...@vcd.hp.com)
rjs...@calcube.com
Sorry if this is redundant, but I'm trying for the last time to submit
this post. Severe news server problems. Thanx!
------------------------------------------------------------------------
After following the "lesson in sound.." and "phase coherence" thread,
I'm inclined to challenge the whole notion of "accuracy." I'll confine
this response to loudspeakers only, since they ultimately represent the
tangible link to sound reproduction. My perspective on this subject is
one initiated from being a loudspeaker designer and also a music
aficionado. Biased perspective? Maybe.
The term accuracy means exactness, correctness, or freedom from error.
In the context of further discussion, this is the definition of accuracy
I'm going to use as a *frame of reference*. This definition represents a
readily understood concept with universal acceptance.
According to Miriam-Webster (dictionary), the term loudspeaker means: "a
device that changes electrical signals into sounds loud enough to be
heard at a distance." I think that definition is pretty straightforward
and unless one resorts to semantics, most individuals of a technological
discipline would find this definition to be accurate and universally
accepted. I'm buying into this definition for further discussion.
So we've now established a *frame of reference* regarding the terms
"loudspeaker" and "accuracy." We understand the terms, accept the terms,
and even apply them when discussing audio or music. THEN WHY IS THERE SO
MUCH CONFUSION WHEN WE COMBINE THE TERMS?! Why is it that two words with
concrete definitions lose all significance when combined? What is it
about the term "loudspeaker accuracy" that has so many people confused
about deriving a universally accepted definition? Well, I'll tell you:
THERE IS NO SUCH THING AS LOUDSPEAKER ACCURACY! That's right, there's no
such thing as loudspeaker accuracy because the *frame of reference* for
loudspeaker accuracy is dynamic, not static. Loudspeaker accuracy is a
chameleon that changes colors between measurements and realism.
Measurements are an objective color while realism is a subjective color.
Or is that assumption correct? Depending upon the environment, either of
these two colors may be accurate....or not. Let's take a closer look.
Measurements.
As a loudspeaker designer, when I think of measurements I think of
benchmarks. I think of measurements as an objective frame of reference
to consistently improve upon existing iterations leading one to
theoretical perfection. With the exception of a healthy newborn child,
there is no real perfection. Measurements are simply experiments
performed in controlled environments. So why are loudspeaker
measurements performed in controlled environments rather than the 'real
world' environments that we listen to them in? Let's go back to the
original definition of loudspeaker: "a device that changes electrical
signals into sounds loud enough to be heard at a distance." The
operative phrase in this definition is "changes electrical signals into
sounds." Loudspeaker accuracy in the context of measurements means
designing "a device" that does not alter the characteristics of the
"electrical signals." Is this possible? Not yet. Can we get close?
Maybe, but it must be realized that every piece of material used to
build such "a device" deviates from the theoretical perfection initially
calculated, thereby never realizing real world accuracy (defined as
freedom from error). It is this deviation from perfection that has
allowed so many interpretations of loudspeaker design to proliferate.
I.E., every loudspeaker designer thinks they have *the answer*. The
digital age may prove different, but with state of the art as it is
today, there is no perfect "loudspeaker accuracy" as it pertains to
measurements, only close approximations.
Realism.
So what is realism? Without getting too detailed, or conversely too
vague, realism in loudspeakers is simply the illusion that somehow one
A happy and safe holidays to all!
Gary Eickmeier
accuracy and realism, and open up the discussion to a question: What
differences do you hear between live music and the reproduction? Take it
from the biggest difference down to the smallest.
I think all of our theoretical discussions should be put on hold unless
they can narrow these differences down.
Gary Eickmeier
JeffC 1474
>differences do you hear between live music and the reproduction? Take it
>from the biggest difference down to the smallest.
>
>I think all of our theoretical discussions should be put on hold unless
>they can narrow these differences down.
I think it depends on what you care most about in live music. You
can pick up on this all the time in stereo reviews. Someone will
gush about an amps "liquid midrange" - those people probably like the
timbre of violins etc. Someone will complain about a turntables wow
and flutter - those people probably like solo piano music.
I personally think the biggest difference between live music and
reproduction is dynamics and ease at loud volume. This shows up
for me with percussion, massed instruments in a forte section, and
live jazz/rock, especially in the lower frequencies.
I really don't care if the timbre is exactly the same
or the pitch is exactly the same because those aren't the things
I care that much about in live music either. (I'm not saying I
like things out of tune, just that if the whole thing is 2% higher
in pitch on playback I don't care.) Sure I'd like a perfectly
accurate speaker, but I wouldn't be willing to pay for it. I'd
only be willing to pay for my priorities.
Regards,
Jeff
Chris Malcolm
es
>Let's take a step back, for a moment, from the differences between
>differences do you hear between live music and the reproduction? Take it
>from the biggest difference down to the smallest.
>I think all of our theoretical discussions should be put on hold unless
>they can narrow these differences down.
The most important difference is scale: the apparent image is too
close. All sorts of other defects follow from this, e.g., the sound
pressure is often raised, and the music sounds cramped and lacking in
dynamic range.
The second most important difference is boxiness, which
almost all box speakers suffer from.
The third most important difference is whatever it is in many speakers
and amplifiers that congeals massed voices into a harsh group sound
instead of easily separable distinct voices. I say "voices" because
the effect seems most noticeable in choirs, although it affects
orchestras too.
--
Chris Malcolm c...@dai.ed.ac.uk +44 (0)131 650 3085
Department of Artificial Intelligence, Edinburgh University
5 Forrest Hill, Edinburgh, EH1 2QL, UK DoD #205
http://www.dai.ed.ac.uk/daidb/people/homes/cam/
Bob Trosper
>
> Let's take a step back, for a moment, from the differences between
> accuracy and realism, and open up the discussion to a question: What
> differences do you hear between live music and the reproduction? Take it
> from the biggest difference down to the smallest.
Ah, well, the biggest difference is I have no difficulty wondering
whether my "system" is reproducing it correctly. Probably the worst
thing about audiophilia is just enjoying music as music rather than
wondering if some tweak you could make to your system would make it
better.
From here on, these are pretty much as they come to mind. Rank 'em how
you like.
Another is, as another poster has mentioned, the experience of power and
dynamics without absurdly loud volumes. Many have measured the decibels
at a seat in symphony hall vs. home playback levels, and home playback
has to be 5-10 dB higher to sound as "realistic".
Another is the experience of hall space. I've never heard a system,
surround or not, that reproduces this correctly.
Another is the experience of "detail", that is the ability to "hear into
the mix". Much more easily done live than reproduced, except for those
artificial emphases in the recording. Those, however, are an artifact of
the recording and engineering process.
Another is all those people in the hall with me. There's something about
a communal experience vs. a singular one. It may be worthy of note that
some religious ceremonies CANNOT be performed in the presence of one.
Think also of the experience of watching a move alone in a theater vs.
one full of people. Even without the obvious sounds from the audience,
just the change in breathing, body heat, mass leaning forward or back is
very effective in affecting your own mood. The same thing happens at
live musical events, of course, including classical music.
So, even if your mythical perfect sound system can get all the audible
cues right, that last one is going to be a real toughie.
>
> I think all of our theoretical discussions should be put on hold unless
> they can narrow these differences down.
Why?
-- BT
Gary Eickmeier
>
> Gary Eickmeier wrote:
> > I think all of our theoretical discussions should be put on hold unless
> > they can narrow these differences down.
>
> Why?
>
> -- BT
AHA! I thought the reason was apparent, but if not, good question!
Let's suppose some theorist says that what is important in sound
reproduction is phase response. You've got to have good time, phase, and
impulse response, along with the rest. So you do an experiment with an
all-pass network in which you have two presentations, one with perfect
impulse response and one with varying degradations from that. You could
tell not only whether it was true, but how much time misalignment was
audible. IF you found that you could have up to 2000 degrees of phase
response error with no effect, then that theory would have done nothing
to bring the reproduction closer to the original.
Another way of thinking of it is if you have a speaker with perfect time
response and no other obvious problems, like frequency response
anomolies, and it does not close the gap or certainly doesn't amaze you
in its realism, then the theory must not be addressing factors important
to perception in the problem.
But maybe the biggest thing is that you don't usually go up to a live
orchestra and say to yourself, what fantastic phase response. Especially
if there are factors MUCH more apparent, or audible, that separate the
live from the reproduced.
I think we have, in this portion of this thread at least, a phenomenon
never before seen in RAHE: we all agree on something! Even Eickmeier!
To me, the biggest one (difference) is dynamics. You stand anywhere near
the most modest of bands, and you have just got to say to yourself
there's no way we could reproduce this. It's like they said in Jaws when
they first saw the shark, "We need a bigger boat." Meaning, of course,
acoustic power.
Which agrees basically with other factors mentioned by posters, physical
size and power. I have always thought that the larger the room the more
real it sounds, because you can easily hear the size of the
presentation.
Finally, in my biggies, as Bob mentioned, I would say spatial
characteristics. It is quite difficult to get two or even three speakers
to image like 60 discrete sources on the soundstage of a good hall with
good acoustics. Especially to be able to move around, stand and sit, etc
etc as you can do live.
There are probably a few more, but basically any theory of reproduction
must address these differences and suggest how the new method can narrow
the gap in one or more of them. We must think more globally about the
problem, and not get lost in details like what material the tweeter is
made of, if we are nowhere close to the real thing yet.
Gary Eickmeier
John Rossmann
>>What
>>differences do you hear between live music and the reproduction? Take it
>>from the biggest difference down to the smallest.
>I personally think the biggest difference between live music and
>reproduction is dynamics and ease at loud volume. This shows up
>for me with percussion, massed instruments in a forte section, and
>live jazz/rock, especially in the lower frequencies.
>Sure I'd like a perfectly
>accurate speaker, but I wouldn't be willing to pay for it. I'd
>only be willing to pay for my priorities.
Although this is a bit off the "loudspeaker accuracy" topic, my experience
has been that the ability to reproduce dynamics and preserve the sense of
"ease" at high volume is probably the most obvious improvement one hears
as one moves to higher quality equipment.
Over the past 10 years, I've been able to gradually upgrade my system.
I listen primarily to classical music of just about all kinds, and large
orchestral works have always been a substantial part of my aural diet.
It wasn't until the last upgrade that the system truly reproduced very
loud large orchestral passages with all the other sonic characteristics
unchanged, i.e. no soundstage contraction, no change of timbre, no sense
of stress. All other aspects of the sound improved with each upgrade too,
such as being able to differentiate instruments and voices, understand
lyrics, and all that other stuff. But the most obvious change whenever
the power amplification or speakers were upgraded was the improvement
at high volume with complex material.
You can perhaps have dynamics without paying a lot, but to have "ease"
at loud volume means that all those other characteristics must be
preserved as well, and in my experience that is what costs. One could
perhaps say that it's the top 20db and the bottom 10Hz that destroy
one's bank account!
John
--------------------------------------------------------------------------
John C. Rossmann (jo...@synersys.com)
"Giving money and power to government is like giving whiskey and car keys
to teen-age boys." -- P. J. O'Rourke
--------------------------------------------------------------------------
JeffC 1474
>You can perhaps have dynamics without paying a lot, but to have "ease"
>at loud volume means that all those other characteristics must be
>preserved as well, and in my experience that is what costs. One could
>perhaps say that it's the top 20db and the bottom 10Hz that destroy
>one's bank account!
Yes. Well I'm not exactly sure what you mean by "bottom 10Hz" but I
think I got your drift. My brother has Spica TC-60s (I think) and
they sound great for a low price - except that they don't play loud
and they don't play deep!
Regards,
Jeff
John Ongtooguk
: Let's suppose some theorist says that what is important in sound
: reproduction is phase response. You've got to have good time, phase, and
: impulse response, along with the rest. So you do an experiment with an
: all-pass network in which you have two presentations, one with perfect
: impulse response and one with varying degradations from that. You could
: tell not only whether it was true, but how much time misalignment was
: audible. IF you found that you could have up to 2000 degrees of phase
: response error with no effect, then that theory would have done nothing
: to bring the reproduction closer to the original.
Use dejanews to look at some of jj's replies to similar comments about
phase and all-pass filters and such, where he notes:
*****************************************************************
Some of the old reports that phase was not important used a frequency
separation such that the two tones were not exciting the same areas of
the basilar membrane at all. While they are true for what they show,
they do not in general represent the only test possible.
It's easy to prove that phase matters. Just generate an all-pass
filter with about 10 pairs of pole/zero at 1kHz, and filter something
with it. Design said filter with maximum phase shift.
It will sound different.
*****************************************************************
Decent time and frequency domain performance is merely a common
sense starting point, as decent hardware won't make your favorite
pre-recorded Herman's Hermits cassette sound like a real concert,
it will merely allow the recording to stand on it's own merits.
John Ongtooguk (jo...@vcd.hp.com)
Dave King
sound is about half as good as live. So, system distortion is 50%.
Since nothing in the chain measures nearly that high (even in
aggregate), then there must be other problems loose in the audio
universe.
I suspect multi channel playback with DSP will bring everything quite
a bit closer sooner or later.
Having said that, have you heard the TACT digital amp yet???? Just
heard it today at a friends who has Gallo References----wow.
Interesting to note that the primary improvement is the elimination of
analogue amplification stages that collectively shouldn't even be
audible anyway by accepted wisdom on audibility of distortions.
I'm here to tell ya the differences aren't subtle. Pass an ABX no
problem.
Dave King
Gary Eickmeier <image...@worldnet.att.net> wrote:
>Let's take a step back, for a moment, from the differences between
>accuracy and realism, and open up the discussion to a question: What
>differences do you hear between live music and the reproduction? Take it
>from the biggest difference down to the smallest.
>I think all of our theoretical discussions should be put on hold unless
jj, curmudgeon and tiring philalethist
Dave King <king...@mindspring.com> wrote:
>Gary, here's how I look at it. In a good modern system the reproduced
>sound is about half as good as live. So, system distortion is 50%.
Actually, it's a lot more than 50%. There is a huge loss of (audible)
information when the signal is reduced (be it by mikes or by
processing) to two channels of information.
Once this has happened, the chance of "immersion" is effectively
remote, if not gone altogether.
>Since nothing in the chain measures nearly that high (even in
>aggregate), then there must be other problems loose in the audio
>universe.
Well, the losses in 2-channel are well known. Even in the 1939's, it
was known that for proper imaging and depth perception for the FRONT
on needed a minimum of 3 channels.
JJ
--
Copyright j...@research.att.com 1999, all rights reserved, except transmission
by USENET and like facilities granted. This notice must be included. Any
use by a provider charging in any way for the IP represented in and by this
article and any inclusion in print or other media are specifically prohibited.
Gary Eickmeier
>Gary, here's how I look at it. In a good modern system the reproduced
>sound is about half as good as live. So, system distortion is 50%.
>aggregate), then there must be other problems loose in the audio
>universe.
>
>a bit closer sooner or later.
This is an old thread, and I believe the previous traffic on this
subject pretty much covered my little question. What I was after was
a simple statement on the perceived differences between live and
reproduced.
Several contributors agreed that the biggest difference you pick up
immediately is POWER. It just ain't possible for most speaker systems
to reproduce the dynamics of live sound at all frequencies. Other
areas mentioned were spatial qualities such as imaging and physical
size.
I firmly believe that because of what you said about there not being
much distortion in the chain, the differences can be overcome with
multiple channels, lots of power, and kick-ass speakers that can take
that power. And a fairly large room.
Gary Eickmeier
Thomas Nulla
<image...@worldnet.att.net> wrote:
>Dave King wrote in message <79f9ut$e...@news01.aud.alcatel.com>...
>>Gary, here's how I look at it. In a good modern system the reproduced
>>sound is about half as good as live. So, system distortion is 50%.
>>Since nothing in the chain measures nearly that high (even in
>>aggregate), then there must be other problems loose in the audio
>>universe.
>>
I don't feel comfortable assigning percentages, but I agree with the
basic point here.
>>I suspect multi channel playback with DSP will bring everything quite
>>a bit closer sooner or later.
Closer, indeed. A lot will depend on how the audio industry
*implements* multi-channel.
>This is an old thread, and I believe the previous traffic on this
>subject pretty much covered my little question. What I was after was
>a simple statement on the perceived differences between live and
>reproduced.
>
>Several contributors agreed that the biggest difference you pick up
>immediately is POWER. It just ain't possible for most speaker systems
>to reproduce the dynamics of live sound at all frequencies. Other
>areas mentioned were spatial qualities such as imaging and physical
>size.
Agreed.
Sheer power is the (relatively!) easy one. (look at the car-audio dB
levels! It is doable! :-) Spatial characteristics are much harder,
and I have never seen a credible proposal to accurately reproduce the
spatial characteristics of a range of live performances in a
real-world listening room. I have heard pretty convincing
*simulations*, however.
We may get the most 'bang for the buck' by understanding enough
psychoacoustics to be able to present our brains with sufficient
information to make us believe we are hearing what *could* have been a
live performance somewhere. As long as we can't directly compare
live-vs-same-performance-reproduced, that might be 'good enough' to
keep us happy.
>I firmly believe that because of what you said about there not being
>much distortion in the chain, the differences can be overcome with
>multiple channels, lots of power, and kick-ass speakers that can take
>that power. And a fairly large room.
When you say 'multiple' channels, about how many do you have in mind?
>Gary Eickmeier
Thomas <now playing: Miles Davis, "Bitches Brew">
http://www.io.com/~nulla (high fidelity and miscellany)
'Bus Plunge' page, Dunlavy Audio r.a.* archive to 6 December 1998
"The Universe is the Practical Joke of the General at the Expense of
the Particular", quoth Frater Perdurabo, and laughed.
jj, curmudgeon and tiring philalethist
Gary Eickmeier <image...@worldnet.att.net> wrote:
>Several contributors agreed that the biggest difference you pick up
>immediately is POWER.
Except that this is already moot for most listening setups.
>It just ain't possible for most speaker systems
>to reproduce the dynamics of live sound at all frequencies.
Going from 2 to 5 speakers, for instance, is an increase in level
of 3.9794 dB in output. Is this really important?
>Other
>areas mentioned were spatial qualities such as imaging and physical
>size.
Those, I think, are well established factors for which evidence
was initially gathered in the 1930's.
Gary Eickmeier
>Closer, indeed. A lot will depend on how the audio industry
>*implements* multi-channel.
Agreed. I hate it when they put instruments in the rear channels.
>Sheer power is the (relatively!) easy one. (look at the car-audio dB
>levels! It is doable! :-) Spatial characteristics are much harder,
>and I have never seen a credible proposal to accurately reproduce the
>spatial characteristics of a range of live performances in a
>real-world listening room. I have heard pretty convincing
>*simulations*, however.
I'm thinking of those live vs recorded demos they did in the fifties and
later, where the recording was nearly anechoic and it was reproduced in a
relatively large room. The room supplies most of the acoustic signature,
making it quite real, and there were just a few instruments recorded, like a
string quartet or something, so it wasn't difficult to get the dynamics and
there was almost one speaker per instrument, so the spatial and imaging
weren't too tough, either.
So what you do is close mike the instruments and use speakers with
relatively omnidirectional rad pat and play it in a large room or use
surround speakers and signal processing to simulate a large room. This
doesn't get around the time between reflections of the smaller room, but
that may be less important with the surround sound and some room treatment.
One thing that is hard to get with a small room as well is sufficient
diffusion. The surround field needs to be very even, as in not spotty. The
radiation pattern of the speakers supplies the lateral reflections
naturally, the same way they occur in a good hall. This may also be
incorporated in the surround processing for the smaller, more anechoic room.
>
>We may get the most 'bang for the buck' by understanding enough
>psychoacoustics to be able to present our brains with sufficient
>information to make us believe we are hearing what *could* have been a
>live performance somewhere. As long as we can't directly compare
>live-vs-same-performance-reproduced, that might be 'good enough' to
>keep us happy.
Agreed. This is the realism vs (relatively unobtainable) accuracy argument.
>When you say 'multiple' channels, about how many do you have in mind?
I don't know. All the rear channels have to do is provide ambience. You need
enough to be sufficiently diffuse, but they don't need to be discrete
channels of information. In front, we need at least three (in a practical
system). Maybe three channels of rear info. The two sides and the rear, for
possible directional effects. If you aren't using omni speakers in front for
lateral early reflections, you would also need two speakers to the sides
with time delay relative to the direct sound, but not as great a delay as
the surround speakers. Sort of a three tiered system of delay, whether
recorded that way or signal processed that way.
Gary Eickmeier
jj, curmudgeon and tiring philalethist
Gary Eickmeier <image...@worldnet.att.net> wrote:
>Agreed. I hate it when they put instruments in the rear channels.
Well, it depends. What about Berlioz' Requiem?
>>When you say 'multiple' channels, about how many do you have in mind?
>I don't know. All the rear channels have to do is provide ambience.
Only in a good hall. In most real halls, they must be able to
also reproduce a good impression of a specular reflection off
the back wall.
Fletcher, "Speech and Hearing in Communication". Chapter 13.
Thomas Nulla
<image...@worldnet.att.net> wrote:
>I'm thinking of those live vs recorded demos they did in the fifties and
>later, where the recording was nearly anechoic and it was reproduced in a
>relatively large room. The room supplies most of the acoustic signature,
>making it quite real, and there were just a few instruments recorded, like a
>string quartet or something, so it wasn't difficult to get the dynamics and
>there was almost one speaker per instrument, so the spatial and imaging
>weren't too tough, either.
>So what you do is close mike the instruments and use speakers with
>relatively omnidirectional rad pat and play it in a large room or use
>surround speakers and signal processing to simulate a large room. This
>doesn't get around the time between reflections of the smaller room, but
>that may be less important with the surround sound and some room treatment.
>One thing that is hard to get with a small room as well is sufficient
>diffusion. The surround field needs to be very even, as in not spotty. The
>radiation pattern of the speakers supplies the lateral reflections
>naturally, the same way they occur in a good hall. This may also be
>incorporated in the surround processing for the smaller, more anechoic room.
I strongly disagree with several of the statements above, but since we
already went through this not too long ago, anyone interested who
missed the numerous refutations of these ideas should search DejaNews
for the 'soundstaging/imaging' threads.
>>When you say 'multiple' channels, about how many do you have in mind?
>I don't know. All the rear channels have to do is provide ambience. You need
>enough to be sufficiently diffuse, but they don't need to be discrete
>channels of information. In front, we need at least three (in a practical
>system). Maybe three channels of rear info. The two sides and the rear, for
>possible directional effects. If you aren't using omni speakers in front for
>lateral early reflections, you would also need two speakers to the sides
>with time delay relative to the direct sound, but not as great a delay as
>the surround speakers. Sort of a three tiered system of delay, whether
>recorded that way or signal processed that way.
I would hope that discrete rear channels will do a better job of
revealing differences between recording venues than the matrixed
alternatives I've heard...and I would certainly want to minimize those
early room reflections for my system.
My real-world purchase decisions will have to await availability of a
significant number of very-high-quality multi-channel recordings. I
wish that was now.
More than five channels would be nice, but I think economic realities
and cheap data storage requirements will keep us limited to five
full-range...and that seems to be the home-theater standard in
practice anyway.
>Gary Eickmeier
Thomas <now playing: Fleck, "Left of Cool">
Dave King
>Several contributors agreed that the biggest difference you pick up
>immediately is POWER. It just ain't possible for most speaker systems
>to reproduce the dynamics of live sound at all frequencies. Other
>areas mentioned were spatial qualities such as imaging and physical
>size.
>I firmly believe that because of what you said about there not being
>much distortion in the chain, the differences can be overcome with
>multiple channels, lots of power, and kick-ass speakers that can take
>that power. And a fairly large room.
No question the two biggest problems are power compression and field
accuracy. But increasing power without taking care that the issue of
single channel transparency is attended to is potentially problematic
because of listener fatigue. I'm sure we've all heard systems that
get loud, but who cares when all you want to do is leave the room (or
turn down the volume)?
My point in mentioning the TACT amplifer is that it shows that
amplifying signal without audible distortion is apparently not
possible. We shall see.
Dave King