Howard Ferstler wrote:
John Atkinson wrote:
Howard Ferstler wrote:
The original reviews of the IC-20 systems by both High
Fidelity Magazine and Stereo Review commented upon their
ultra-wide dispersion.

But what measurements did those magazines publish to
support that contention? And how where those measurements
performed? Perhaps I am mistaken, but I can't remember
the late Julian Hirsch performing either polar plots[,] or
the waterfall dispersion plots that I perform[,] in Stereo
Review. And again if I remember correctly, High Fidelity
only performed sound _power_ measurements.

No response from Howard Ferstler. Where _are_ the
measurements that support his comments about the Allison
tweeter's dispersion?

Even if the systems had only one forward-facing panel they
would have those Allison drivers dispersing widely clear out
to 90 degrees off axis.

But without measured support for this statement, you have
no basis to make it, unless you label it as being your
_opinion_.

Again, no substantiation for his statements from Howard Ferstler.

However, they have dual panels, each angled out 45 degrees
from dead ahead, meaning that they have extremely wide
dispersion to each side and out to 135 degrees off axis.

No, this results in _narrower_ dispersion on the primary
frontal axis, with complex lobing apparent.

Normally, I would not post a second reply to a message here
(particularly one initially posted by Mr. Atkinson as a
grandstanding move) and would wait for a counterpoint from
my adversary.

I am not "grandstanding," Mr. Ferstler, merely asking you to
provide some measured substantiation for your comments about the
objective performance of the Allison tweeter and the speakers
that use it. Particularly as your _own_ in-room measurements
reveal that even with the Allison speakers, the balance at your
listening seat is dominated by the direct sound in the region
covered by the tweeters.

However, Mr. Atkinson's comments about the radiation pattern
of the angled-panel Allison speakers gave me an incentive to
contact Roy Allison, with a question regarding the
difference between horizontally spaced tweeter and midrange
driver pairs on a single, forward-facing panel and a
situation where each of the units are on a separate panel,
angled to either side of straight ahead.

Okay.

From Allison himself:
"At frequencies above about a half wavelength a baffle
effectively limits radiation to a maximum angle of 90
degrees around a driver mounted symmetrically in its
smallest dimension - in other words, to 2 pi steradians.

My own measurements indicate that this is too simplistic a
view. If, for example, you look at fig.3 at
http://www.stereophile.com/loudspeak...61/index6.html,
which shows the 360-degree radiation pattern of a BBC LS3/5a,
you can see that even with the recessed tweeter in this design,
there is still significant radiation at some frequencies in
the tweeter's passband more than 90 degrees to the side, even
though the baffle is larger than a half wavelength at these
frequencies.

For the woofer in a Model One or IC-20 this doesn't apply, but
the distance between the woofers isn't great enough for
cancellation to occur within their range.

If I remember correctly, the centers of the woofers and midrange
units in these designs are separated by around 12", meaning that
the cancellation will occur for frequencies where this difference
results in a half-wavelength path difference. For someone
listening on-axis to one of the drivers, there will be
cancellation at 600Hz, which may be above the woofers' passband,
but is certainly in the midrange units' passband.

The tweeter radiation can't go around a right-angle corner at
all...

Not true, as I showed above with the BBC design. There will be
diffraction at some frequencies but not others, meaning that
it is not true that the drivers on each of the Allison speaker's
baffles are not in each other's acoustic environments. Again,
Mr. Ferstler where are your _own_ measurements that support
what you say.

and for the mids there is the slight beginning of interference
at the bottom of their range.

A half-wavelength path difference would result in significant,
not "slight" interference. See for example, the horizontal
dispersion of the popular center-channel speakers that use
spaced midrange units or woofers. (A subject on which I agree
with Tom Nousaine, BTW.)

However, the crossover and the mid's inherent Q value are
designed to compensate for that minor effect."

I fail to grasp how adjusting the on-axis behavior compensates
in any significant way for arbitrary and irregular dispersion
effects. Please offer more detail, Mr. Ferstler.

The ball is in your court, Mr. Atkinson.

As I said, Mr. Ferstler, use of multiple, spaced drivers
results in a less-wide, less-even dispersion pattern,
owing to the interference between the spaced drivers
producing a complex pattern of lobing. Yes, I will allow
that by angling the baffles on which the spaced drivers are
placed, you will reduce the interference, and there _will_
now be lobes aimed at wider off-axis angles than with a single
drive-unit. But the price paid to achieve the existence of
those lobes is a significant reduction in the smoothness of
the radiation pattern, which Toole's and Olive's work has shown
to be undesirable, subjectively.

Your arguments have all been circular, Mr. Ferstler. That you
like the sound of the Allison speakers goes without saying. But
without measured data to support that opinion, your projection
about the technical reasons for your preference are mere
masturbation.

John Atkinson
Editor, Stereophile

John Atkinson wrote:

From Allison himself:
"At frequencies above about a half wavelength a baffle
effectively limits radiation to a maximum angle of 90
degrees around a driver mounted symmetrically in its
smallest dimension - in other words, to 2 pi steradians.

My own measurements indicate that this is too simplistic a
view. If, for example, you look at fig.3 at
http://www.stereophile.com/loudspeak...61/index6.html,
which shows the 360-degree radiation pattern of a BBC LS3/5a,
you can see that even with the recessed tweeter in this design,
there is still significant radiation at some frequencies in
the tweeter's passband more than 90 degrees to the side, even
though the baffle is larger than a half wavelength at these
frequencies.


Fig 3 is an ugly beast of a 3-D plot to try and decipher... too much
information, if you can't slide a cursor along to see values one loses
ability to correctly interpret the amplitude value. At least my tired
eyes do.
In any case... I have a question.

The fig3 plot says "Fig.3 Rogers LS3/5a, 1978 sample, horizontal
response family at 50", normalized to response on tweeter axis, from
back to front: differences in response 180 degrees-5 degrees off-axis;
reference response; differences in response 5 degrees-180 degrees
off-axis. "

So the 0 degree plot should either agree with the on axis (fig 4 which
is also at 50") or if truly normalized to on axis then I think the 0
degree plot should be flat. I can't see either situation in Fig 3.

TIA for helping me understand,

ScottW

"Arny Krueger" wrote in message
...
Clyde Slick wrote:
"Arny Krueger" wrote in message
...

Howard, I was very polite with you about your errors in those other
posts. If you want to see vituperation, look at the corresponding
posts from that well-known technical ignoramous, Scott Wheeler.


If irony would wash your mouth with feces.

Art, I prefer to watch your example rather than following it.


Thanks for admitting that you solicit for scat fetish videos.



----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

ScottW wrote:
The fig3 plot says "Fig.3 Rogers LS3/5a, 1978 sample, horizontal
response family at 50", normalized to response on tweeter axis, from
back to front: differences in response 180 degrees-5 degrees
off-axis;
reference response; differences in response 5 degrees-180 degrees
off-axis."

Yes, the plot show the LS3/5a's output plotted in a complete circle at
5 degree intervals.

So the 0 degree plot should either agree with the on axis (fig 4
which
is also at 50") or if truly normalized to on axis then I think the 0
degree plot should be flat. I can't see either situation in Fig 3.

The fig.4 response was subtracted from every response in fig.3. meaning
that the 0-degree trace is indeed a flat, straight line.

But the notable point is that the tweeter output doesn't stop dead at
90 degrees off-axis. Below 12kHz, there is still output apparent above
-30dB floor of this graph, particularly at some frequencies.

John Atkinson
Editor, Stereophile

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

John Stone is a driver salesman. I replied at length to his
commentary elsewhere. Go find it.

You're quick to dismiss experts, aren't you? Wasn't Allison a speaker
salesman?

Allison designed both systems and drivers and published
numerous articles on system design in the JAES. He even
published articles in pop-audio magazines at one time. I
doubt if Stone is a Fellow, or even a member, of the Audio
Engineering Society.

From The Encyclopedia of Recorded Sound:

Allison became involved with electronic matters during U.S.
Navy service and later at the University of Connecticut's
school of Electrical Engineering. Early in his audio career,
Allison was editor of Radio Communication, TV and Radio
Engineering, and Communication Engineering magazines, Audio
Editor of High Fidelity magazine and Editor of Audiocraft
magazine.

In 1959 he joined Acoustic Research, Inc. He became chief
engineer in 1962 and became plant manager in 1964. In 1967,
he became vice president in charge of research and
development. In 1974, after leaving AR, he helped to found
Allison Acoustics, RDL, and RA Labs. He has published
numerous professional-level and hobby-related articles and
wrote a book, High Fidelity Systems. He has been a Fellow of
the Audio Engineering Society since 1973.

His most important later articles, published in both
technical and consumer-oriented journals, deal with
speaker/room interactions. He is responsible for defining
the "Allison Effect," which involves mid-bass cancellation
artifacts between loudspeaker systems and room boundaries,
and is also responsible for some highly regarded speaker
driver and loudspeaker system designs.

End of excerpt.

PS: there is no entry for John Stone in The Encyclopedia.

Howard Ferstler

John Atkinson wrote:

Howard Ferstler wrote:
John Atkinson wrote:
Howard Ferstler wrote:
The original reviews of the IC-20 systems by both High
Fidelity Magazine and Stereo Review commented upon their
ultra-wide dispersion.

But what measurements did those magazines publish to
support that contention? And how where those measurements
performed? Perhaps I am mistaken, but I can't remember
the late Julian Hirsch performing either polar plots[,] or
the waterfall dispersion plots that I perform[,] in Stereo
Review. And again if I remember correctly, High Fidelity
only performed sound _power_ measurements.

No response from Howard Ferstler. Where _are_ the
measurements that support his comments about the Allison
tweeter's dispersion?

See pages 46 and 47 of my 1991 book, High Fidelity
Audio/Video Systems (McFarland publishing). Hey, John, I
believe that there was a review of it by Cory Greenberg in
your magazine a long while back. Don't you remember those
driver curves?

Even if the systems had only one forward-facing panel they
would have those Allison drivers dispersing widely clear out
to 90 degrees off axis.

But without measured support for this statement, you have
no basis to make it, unless you label it as being your
_opinion_.

Again, no substantiation for his statements from Howard Ferstler.

See the book. Also, see Allison's quote about how the
angled-panel systems do not have problems with horizontal
cancellation artifacts in another post I submitted in
response to your earlier comments. Oops, I see up ahead that
you managed to incorporate them right here. You then went on
to do an amateur-hour analysis of them.

However, they have dual panels, each angled out 45 degrees
from dead ahead, meaning that they have extremely wide
dispersion to each side and out to 135 degrees off axis.

No, this results in _narrower_ dispersion on the primary
frontal axis, with complex lobing apparent.

Normally, I would not post a second reply to a message here
(particularly one initially posted by Mr. Atkinson as a
grandstanding move) and would wait for a counterpoint from
my adversary.

I am not "grandstanding," Mr. Ferstler, merely asking you to
provide some measured substantiation for your comments about the
objective performance of the Allison tweeter and the speakers
that use it. Particularly as your _own_ in-room measurements
reveal that even with the Allison speakers, the balance at your
listening seat is dominated by the direct sound in the region
covered by the tweeters.

This is preposterous. At a listening distance of 15 feet (to
each speaker, I would be well into the reverberant field
even with fairly directional speakers.

John, you have to remember that I am a "subjective" speaker
reviewer. I am not quite as subjective as some other
reviewers, but I am certainly more subjective than you are
if we are talking about measurements. In any case, the best
I can offer are Allison's own measurements on the drivers
and systems. Of course, you may think that he is either
fudging the facts or incompetent.

However, Mr. Atkinson's comments about the radiation pattern
of the angled-panel Allison speakers gave me an incentive to
contact Roy Allison, with a question regarding the
difference between horizontally spaced tweeter and midrange
driver pairs on a single, forward-facing panel and a
situation where each of the units are on a separate panel,
angled to either side of straight ahead.

Okay.

From Allison himself:
"At frequencies above about a half wavelength a baffle
effectively limits radiation to a maximum angle of 90
degrees around a driver mounted symmetrically in its
smallest dimension - in other words, to 2 pi steradians.

My own measurements indicate that this is too simplistic a
view.

John Atkinson trumps Allison. John, you are indeed so
complex and subtle an individual.

If, for example, you look at fig.3 at
http://www.stereophile.com/loudspeak...61/index6.html,
which shows the 360-degree radiation pattern of a BBC LS3/5a,
you can see that even with the recessed tweeter in this design,
there is still significant radiation at some frequencies in
the tweeter's passband more than 90 degrees to the side, even
though the baffle is larger than a half wavelength at these
frequencies.

For the woofer in a Model One or IC-20 this doesn't apply, but
the distance between the woofers isn't great enough for
cancellation to occur within their range.

If I remember correctly, the centers of the woofers and midrange
units in these designs are separated by around 12",

About 11 inches with the IC-20. About 9 inches with the
Model One.

meaning that
the cancellation will occur for frequencies where this difference
results in a half-wavelength path difference. For someone
listening on-axis to one of the drivers, there will be
cancellation at 600Hz, which may be above the woofers' passband,
but is certainly in the midrange units' passband.

Allison indicated that the crossover arrangement compensated
for this in the reverberant field. In any case, I certainly
have never measured an artifact at that frequency. Of
course, all I measure is room response and not direct-field
response. As you know, I do not put much stock in the
direct-field response, anyway.

The tweeter radiation can't go around a right-angle corner at
all...

Not true, as I showed above with the BBC design.

John Atkinson has managed to disprove Roy Allison without
really understanding what he was attempting to say.
Remember, John, Allison's interest involved getting not only
a smooth radiation pattern, but also involved having
flat-power input to the listening room. An absolutely smooth
first-arrival, direct-field signal was not high on his
priority list.

There will be
diffraction at some frequencies but not others, meaning that
it is not true that the drivers on each of the Allison speaker's
baffles are not in each other's acoustic environments. Again,
Mr. Ferstler where are your _own_ measurements that support
what you say.

John, unlike you I am not about to second guess Allison. And
you very well know that the only measurements I do for my
reviews are room curves. However, I also realize that guys
like you often measure to death and still cannot correlate
those measurements with what people hear. I certainly have
seen this happen in the past with some of your amp
measurements when they were contrasted with the subjective
impressions of your reviewers.

and for the mids there is the slight beginning of interference
at the bottom of their range.

A half-wavelength path difference would result in significant,
not "slight" interference. See for example, the horizontal
dispersion of the popular center-channel speakers that use
spaced midrange units or woofers. (A subject on which I agree
with Tom Nousaine, BTW.)

You are on a roll, John. But the spaced-driver situation we
have with typical center speakers involves interference
effects over a broad frequency range. Even you admit that
the interference effects with the Allison design limits the
bandwidth covered by the artifacts.

However, the crossover and the mid's inherent Q value are
designed to compensate for that minor effect."

I fail to grasp how adjusting the on-axis behavior compensates
in any significant way for arbitrary and irregular dispersion
effects. Please offer more detail, Mr. Ferstler.

John, just for kicks I will send your comments to Roy to
look over. I will get back to you.

The ball is in your court, Mr. Atkinson.

As I said, Mr. Ferstler, use of multiple, spaced drivers
results in a less-wide, less-even dispersion pattern,
owing to the interference between the spaced drivers
producing a complex pattern of lobing. Yes, I will allow
that by angling the baffles on which the spaced drivers are
placed, you will reduce the interference, and there _will_
now be lobes aimed at wider off-axis angles than with a single
drive-unit. But the price paid to achieve the existence of
those lobes is a significant reduction in the smoothness of
the radiation pattern, which Toole's and Olive's work has shown
to be undesirable, subjectively.

This is taste related, and involves imaging issues related
to the impact of the first-arrival, direct-field signals.
Yes, that can mean a lot to some individuals, but I have
never felt that it was significant when auditioning speakers
for realistic sound.

Your arguments have all been circular, Mr. Ferstler. That you
like the sound of the Allison speakers goes without saying. But
without measured data to support that opinion, your projection
about the technical reasons for your preference are mere
masturbation.

Well, I think you are a dirt bag, too, John. Unlike you, I
really believe what I believe, whereas you are a
turf-protecting (and probably job protecting) tweako
journalist who grandstands.

Howard Ferstler

John Atkinson wrote:

ScottW wrote:
The fig3 plot says "Fig.3 Rogers LS3/5a, 1978 sample, horizontal
response family at 50", normalized to response on tweeter axis, from
back to front: differences in response 180 degrees-5 degrees
off-axis;
reference response; differences in response 5 degrees-180 degrees
off-axis."

Yes, the plot show the LS3/5a's output plotted in a complete circle at
5 degree intervals.

So the 0 degree plot should either agree with the on axis (fig 4
which
is also at 50") or if truly normalized to on axis then I think the 0
degree plot should be flat. I can't see either situation in Fig 3.

The fig.4 response was subtracted from every response in fig.3. meaning
that the 0-degree trace is indeed a flat, straight line.

But the notable point is that the tweeter output doesn't stop dead at
90 degrees off-axis. Below 12kHz, there is still output apparent above
-30dB floor of this graph, particularly at some frequencies.

John Atkinson
Editor, Stereophile

I have a comment, too.

As best I can tell, this horizontal radiation pattern
readout was done at just one height above the ground plane.
Looks like measuring from a slightly different height would
result in a different set of curves, given the interference
effects and different radiation patterns from each driver in
the vertical plane. How can one set of multiple polar curves
be worth anything if a slight measurement-height change will
change the shape of most of them?

In order to make direct-field measurements like this
meaningful (if we want to see the energy smoothness of the
system into a total room space), not only would you have to
do a full series at one height, you would also have to do a
multiple series of horizontal RP readouts at a lot of other
heights. So, as comprehensive as this information looks, it
does have serious information gaps.

Well, I suppose you could measure at the seated ear height
and then the info would be good for the direct-field impact.
However, I continue to believe that the direct field, which
is only a small fraction of the total sound output from a
speaker, does not tell us much about how a speaker sounds in
a real-world listening room, at least if that real-world
room is not padded and the listener is not sitting almost up
against the speakers.

Ironically, a room curve that stresses sound power would
give us a better idea of the total power input to a room
than even many fixed-height measurements in the direct
field.

Howard Ferstler

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

Pal, you would have to use a LOT of wire to effect a proper
delay line for that concentric-ring design. You would
practically have to make it several miles long.

At last! Yes, Quad use a LOT of wire, plus some phase tricks.

http://www.stereophile.com/loudspeak...16/index6.html

"These rings were fed by delay lines (employing some 11 miles of wire!)

Holy cow. And you guys split hairs and get into a
sound-quality twist about a few feet of speaker cable
between an amp and some speakers!
Howard Ferstler

Arny Krueger wrote:

MINe 109 wrote:
In article ,
"Arny Krueger" wrote:

Howard Ferstler wrote:
Arny Krueger wrote:

MINe 109 wrote:
In article ,
Howard Ferstler wrote:

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

I have only mentioned what another rather influential and
knowledgeable individual wrote about it (well, said about it
during a lecture). For all I know, the damned thing hung the
moon. But in principle the system has characteristics that
should make some purists a bit apprehensive.

You haven't said what those are.

The esoteric circuitry that allows for broad-bandwidth
delays between each concentric ring.

It's called 'wire'. You don't think it changes the sound.

Wrong. It's a delay line composed of inductors and capacitors,
and
they are each well-known to change the sound.

Finally, some support from you instead of vituperation.

It helps when you are correct about things, Howard.

Howard figured it out.

Prove it.

Yeah. Frankly, if I "figured it out" the result still went
over my head.

Howard Ferstler

dave weil wrote:

On Sun, 17 Apr 2005 19:54:14 -0400, Howard Ferstler
wrote:

I think that
laser interferometer analysis of cone or dome behavior
probably begins to become a bit obsessive

Cue irony loop...

I'm still chuckling about the charge of serious comb-filtering on the
Quad.

There would have to be on the 57 version. On the one that
uses concentric delays with the rings there would not be.
However, the broad-bandwidth delay line that has to work
with those rings has to be causing its own problems.

Howard Ferstler

Howard Ferstler wrote:

Yeah. Frankly, if I "figured it out" the result still went
over my head.

Stephen was just trying to be cute. Instead, he exposed his ignorance.

On Tue, 19 Apr 2005 13:42:40 -0400, Howard Ferstler
wrote:

dave weil wrote:

On Sun, 17 Apr 2005 19:54:14 -0400, Howard Ferstler
wrote:

I think that
laser interferometer analysis of cone or dome behavior
probably begins to become a bit obsessive

Cue irony loop...

I'm still chuckling about the charge of serious comb-filtering on the
Quad.

There would have to be on the 57 version.

No more than any other multi-driver speaker of any reasonable size.

On the one that
uses concentric delays with the rings there would not be.
However, the broad-bandwidth delay line that has to work
with those rings has to be causing its own problems.

No more than most crossovers with their miles of wires (since you want
to compare apples and zebras).

In article ,
Howard Ferstler wrote:

Arny Krueger wrote:

MINe 109 wrote:
In article ,
"Arny Krueger" wrote:

Howard Ferstler wrote:
Arny Krueger wrote:

MINe 109 wrote:
In article ,
Howard Ferstler wrote:

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

I have only mentioned what another rather influential and
knowledgeable individual wrote about it (well, said about it
during a lecture). For all I know, the damned thing hung the
moon. But in principle the system has characteristics that
should make some purists a bit apprehensive.

You haven't said what those are.

The esoteric circuitry that allows for broad-bandwidth
delays between each concentric ring.

It's called 'wire'. You don't think it changes the sound.

Wrong. It's a delay line composed of inductors and capacitors,
and
they are each well-known to change the sound.

Finally, some support from you instead of vituperation.

It helps when you are correct about things, Howard.

Howard figured it out.

Prove it.

Yeah. Frankly, if I "figured it out" the result still went
over my head.

You figured out that the delay depends on lots of sonically benign wire.

Stephen

In article ,
"Arny Krueger" wrote:

Howard Ferstler wrote:

Yeah. Frankly, if I "figured it out" the result still went
over my head.

Stephen was just trying to be cute. Instead, he exposed his ignorance.

Says the man who was willing to bet there wasn't a mile of wire in a
Quad...

Stephen

In article ,
Howard Ferstler wrote:

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

Pal, you would have to use a LOT of wire to effect a proper
delay line for that concentric-ring design. You would
practically have to make it several miles long.

At last! Yes, Quad use a LOT of wire, plus some phase tricks.

http://www.stereophile.com/loudspeak...16/index6.html

"These rings were fed by delay lines (employing some 11 miles of wire!)

Holy cow. And you guys split hairs and get into a
sound-quality twist about a few feet of speaker cable
between an amp and some speakers!

Isn't the point that I *don't* get into a twist about wire?

Stephen

In article ,
Howard Ferstler wrote:

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

John Stone is a driver salesman. I replied at length to his
commentary elsewhere. Go find it.

You're quick to dismiss experts, aren't you? Wasn't Allison a speaker
salesman?

Allison designed both systems and drivers and published
numerous articles on system design in the JAES. He even
published articles in pop-audio magazines at one time. I
doubt if Stone is a Fellow, or even a member, of the Audio
Engineering Society.

Didn't he manufacture and sell speakers at one point?

From The Encyclopedia of Recorded Sound:

snip

"In 1974, after leaving AR, he helped to found
Allison Acoustics, RDL, and RA Labs..."

Thought so. Someone may have mentioned it.

End of excerpt.

PS: there is no entry for John Stone in The Encyclopedia.

So what?

Stephen

John Stone wrote:

On 4/17/05 6:54 PM, in article , "Howard
Ferstler" wrote:

Hey Howard, this will be my last post for at least a week. I'm with an
engineer from Norway and we'll be seeing some of those engineer jerks I
mentioned before. Anything you want me to ask Toole? How about Linkwitz? If
this thread is still alive when I get back, I'll post the curves of the
Allison tweeter so others can see what we're talking about.

Good enough. Remember, those guys are a whole lot more
interested in having the direct-field signal be flat and
coherent than they are in sound power. However, the
direct-field signal is only a small fraction of what we hear
from any speaker, and in my opinion its only salient feature
is its ability to deliver precise imaging.

Well, let's look at the power curves that Consumer Reports
has run for years. Their reviews consistently put Allison's
lower-priced speakers ahead of most of the competition. Yes,
I know there is more to performance than power curves, but
the fact is that Allison's "dated" tweeter design, not to
mention his work with integrating that tweeter with the
woofer (often an upward-facing unit that helped to keep the
horizontal dispersion uniform over a wide area) allowed him
to keep up with and usually surpass all of those more modern
units you automatically lionize.

Irrelevant. Nobody has argued about the dispersion. Nothing in all this
drivel you have posted so proves anything about the "pulsating hemisphere
theory"

Except the design of the driver.

And don't say it has
to work that way because that's the way he designed it. That's a circular
argument.

Baloney. It worked, and continues to work as good as even
the best current designs (when the integral crossover work
is figured in) and better than most. Hey, John, I know you
have a business to protect, so I forgive you.

Against Allison? He's been out of business for years, and never made raw
drivers for anyone else.

Actually, Mark Davis asked Allison to build drivers for the
original dbx Soundfield One system. Davis (Phd from MIT and
the guy who later on helped give us Dolby AC-3) felt that
the Allison drivers were the best available. Roy toyed with
the idea and then decided that being an OEM supplier was not
his cup of tea.

When soft dome tweeters were first introduced the white papers at
the time said they had better dispersion than cone diaphragms because the
convex shape pushed the high frequencies outward. But when concave dome
tweeters came out, it was discovered that they had exactly the same
dispersion as convex domes. So much for that theory.

I am not aware that Allison had anything to do with this at
all. Certainly, his driver is different from most other
domes. In any case, Allison has told me that driver shape
does not automatically make a difference, so he is in
agreement with your latest research. Indeed, he once showed
me a curve of a 1-inch cone tweeter (used in the old AR-6
decades ago) that he said radiated as widely and smoothly as
typical dome one inchers. I printed that curve in a book way
back in 1991, by the way.

Finally. Now put that together with the coaxial cone being driven off the
edge of a 1/2" dome and you get the picture

But the Allison driver, being convex instead of concave like
the small cone tweeter, had way better dispersion. The
reason involves the nature of the shape of the driver and
the radial flexing surround.

The tweeters still
worked. The explanation was all wrong. What was discovered through physical
observation(laser interferometer) is that the diaphragm completely decouples
from the voice coil at high frequencies, and the tweeters act like ring
radiators. Concave or convex, it made no difference. The dome was
contributing virtually nothing outside a very narrow ring on either side of
the voice coil.

OK, but what we are discussing is the impact of the flexing
surround around that dome. Because it moves radially and is
far larger in relation to dome size than typical other
surrounds, it has an impact on the dispersion. You may
speculate about how it really behaves, but as best I can
tell that is all it is: speculation.

That's right Howard, speculation. On your part and on ours.

Many thanks for stating that your team is also speculating.

But ours is
informed and yours isn't.

Informed speculation? Well, I assume that Allison's
"informed speculation" is as precise as yours, particularly
considering that he designed the tweeter.

You are speculating based on no knowledge of what
paper does at high frequencies. We are speculating based on what we've seen
paper do at high frequencies in laser tests.

On drivers other than the Allison unit.

Hey, John, even he would admit that. So would I. However,
you also do not know everything he knows about drivers.

If you're talking about me, you're right. If you're talking about SEAS, our
body of knowledge on drivers far outstrips his. We've been at it much longer
and have far better tools at our disposal. Our head engineer has studied all
Allison's papers and we have measured that tweeter. So we know what he
knows and a bunch more.

Actually, I reviewed a system that used SEAS Millennium
drivers a while back. It performed admirably, but not
exceptionally. The tweeter's output into the top octave was
not a power-response match for the Allison tweeter or even a
number of conventional tweeters I have measured, such as the
one in an NHT unit I also reviewed. There was the usual
midrange-driver efficiency rolloff up near the crossover
point to the tweeter. Still a nice system, however. Just not
world beating.

As explained above there are
plenty of complex mechanical systems out there that work, but whose behavior
is not fully understood. And radiating diaphragms exhibit VERY complex
vibrational behavior, much of which is not fully understood-even today.

And in some important ways all of this work with analysis is
overkill. There is a point where going the current state of
the art one better has no meaningful impact.

Ah yes, that famous Howard Ferstler "frozen in time" point of view.

There is a lot to say for the good old days. Certainly that
would apply to both overkill engineering and the way audio
buffs behave these days.

I think that
laser interferometer analysis of cone or dome behavior
probably begins to become a bit obsessive, given what is
possible from even a perfect one-inch standard dome.

You think? What is the relevance of that given that you're operating totally
out of the dark?

Well, that system I mentioned, with its SEAS drivers did not
perform any better than my Dunlavy Cantatas when it comes to
direct-field detail, and against my Allison IC-20 systems it
was a no-contest situation when it comes to a sense of
soundstaging spaciousness and frontal depth.

I mean,
speakers were able to do live-vs-recording work that was
fantastic way back in the AR-3 era. Yes, that system is
dated in many ways (particularly concerning its rather high
woofer/midrange crossover frequency), but given its success
during those demos (where seasoned audio buffs and
journalists were even impressed) it is hard to believe that
world-shaking improvements have happened since that era.
Better systems, yes, but not as eye opening as you "we make
super-duper driver" freaks might believe.

Are your scared of the dark?

Well, those AR-3 systems sure impressed those audio
journalists.

This is why driver
design remains part science and part art. The behavior of that Allison
surround will be very complex because it isn't just a paper cone, its a
paper cone that's rigidly fixed at one end.

Actually, it is not rigidly fixed. There is a very thin foam
ring between the surround edge and the mounting plate.
According to Allison, without the ring the driver will ring
somewhat. Good try, though, John.

Doesn't change anything. The paper surround is far less rigid than the foam
ring so the ring will do very little.

Not according to Allison. Heck, he obviously tested the
thing with and without the ring. Your guys were certainly
not in a position to do that.

If the foam is only there to "somewhat
stop ringing" it isn't a key part of the moving system. Good try Howard.

Bunk. It is obviously a key part.

And
you better hope the effects of that foam gasket are subtle, because that 25
year old foam is gonna be hard and brittle.

Depends upon the kind of material. Some such stuff holds up
for a long time. The foam he used with the tweeter, as well
as the foam surround he used with the midrange, was not the
same as the foam he used with the woofers. Admittedly,
woofer-surround of foam does deteriorate. In any case, I
have Allison tweeters here 27 years old that measure the
same on axis as others also on hand that are 15 years old.
And those 15-year old jobs managed to deliver sound-power
output to a higher frequency than the Millennium tweeter in
that system I mentioned. Indeed, in terms of flat power to
the microphone in my listening room, virtually no other
tweeters I have measured were as flat between 4 kHz and 16
kHz as the Allison units.

"Greater" how? Its no greater than other 1/2incher above 5khz, so I'm
assuming you're referring to its output below 5K. Yes, that's explained by
the large surround. But at those frequencies, dispersion is not an issue.

But output is. The surround allows the driver to disperse
like a 1/2 incher and still have considerably more output
capability below 5 kHz. Frankly, given what Hirsch measured
with that budget Allison speaker I mentioned in another
post, it has greater output above 5 kHz, as well.

You're speculating again.

Well, one thing for su the 15-year-old tweeters in my
Allison IC-20 systems generated flatter and more extended
power over the treble range than those new SEAS tweeters in
the system I reviewed. I will admit, however, that the
tweeter (and the midrange drivers, too) were sure pretty
looking items. Whatever the design results when it comes to
sound, SEAS does know how to make good-looking drivers. Of
course, looks are not what we listen to.

Why should it be magic? One look at the surround design and
one has to see that it is going to have an effect that is
related to the radial movement.

There's that precision again. "It is going to have an effect" How is the
effect achieved. That's been the question since day one. And you still
haven't answered it.

Neither have you. Your guys never did an analysis that
substantiates your opinion of the tweeter's diaphragm
behavior. You just assume that the radial movement of the
surround does not contribute to its top-octave behavior.

I'm the one
saying that the simple explanation is that above 5khz the radiation is off
the dome.

Conjecture. Did your people do a laser interferometer
measurement? If not, then you are speculating. Incidentally,
the dome is not just hanging out there. Behind it is a fiber
damping material.

If its conjecture, then what is the dome there for?

Part of the overall design. The thing was designed to
simulate a pulsating hemisphere without actually being a
pulsating hemisphere.

Baloney. You are speculating about a tweeter design that you
have not even analyzed with the laser device. You just
assume that the driver must have problems or anomalies.

We know it has problems and anomalies from the measurements. And you just
said laser devices are obsessive. Which is it?

Well, they are obsessive to you. I have done room-curve
measurements of the tweeter (installed in the IC-20 systems)
and it had a response that was both more extended and
smoother than what I measured with the system that employed
your SEAS Millennium tweeter.

You are basically saying that if I do not know the physics
involved I am unqualified to say anything about the concept.

Exactly. Because if you understood even a little, you wouldn't be making the
claims you are making.

Ditto, for you.

So, you are basically saying that a pulsating hemisphere
movement will not disperse any better than a moving
diaphragm of the same diameter?

Not mounted the way the tweeters are in your speakers, no.

OK, now are we discussing the mounting of the tweeters in
the IC-20 systems, or are we discussing the design of the
tweeter itself?

You are speculating. You have no solid proof that this
theory of your works. You just assume that the driver
behaves this way. You speculate even more than I do, John.

I understand how drivers work. Turn the Allison diaphragm around and it is a
cone tweeter with a rigid (mostly) termination. Those drivers are well
understood.

Yes, and on pages 46 and 47 of my book, High Fidelity
Audio/Video Systems there are on and off-axis curves from
the Allison convex dome and a one-inch cone tweeter. They
are almost mirror images of each other, reversed, and yet
the Allison tweeter has way, way better dispersion than the
one-inch cone.

Heck, Thiel did exactly the same thing-though far more refined-with the
coaxial mid/tweet in his 2.3 and 2.4 systems.

Far more refined? Looks like we are speculating again, John.

I've discussed it at length with Jim Thiel. He's had many years and much
better tools to refine the concept. That's what happens in the real world.

Well, I suppose any designer would consider his designs to
be more refined than what is produced by the competition.

This gave him much better low end output
(the cone could move). He uses this driver between 400 and 20khz
And guess what? He got the output and dispersion of a 1" dome with the low
frequency extension of a 3.5" cone. Surprise,Surprise!!

Good for him. Given that the Allison tweeter (and similar
Allison midrange) were designed back in 1975, I think that
Allison gets the prize for the first idea. Actually, the
tweeter in the old Kloss Advent speaker was also kind of
this way, although the shape of the thing was quite
different from the Allison tweeter. As best I can tell, the
Thiel tweeter is also shaped quite differently, and maybe
somewhat like the Kloss design.

Nope. It's just a dome. Don't you have any curiosity at all? It's all there
with a click of your mouse.

I thought you said that it had a special, oversized
surround. Make up your mind.

and all your other
measurements are referenced to these on axis measurements. Like it or not,
that's how its done.

Well, given your take on this and my comments above, it
looks like they need to widen their horizons a bit.

Bull****. The effect of wider dispersion will show up in the polar plots,
which are done on all tweeters.

Give me a break. In many cases, the spec sheets for tweeters
stop measuring at 45 degrees off axis. And there is a whole
lot more room area covered from 45 to 90 degrees off than
there is from zero to 45 degrees off. Almost 2.5 times as
much, actually. This means that the response beyond 45
degrees has a huge impact on the spectral balance of the
driver in a real-world listening room. One that does not
have padded walls, at least.

First off, your Allison tweeters are no different in dispersion capability
at 6kHz and below than a 1" dome.

I certainly hope so. So what?

So it means that the power response is the same between the 2 at those
frequencies. And those are the frequencies where crossover occurs. To splice
the mid or woof to the tweeter, the sensitivities in that range will have to
match.

The radiation patterns also have to match. Typically (as
with the speaker I reviewed that had SEAS drivers) the
midrange is so much larger than the tweeter that the
wide-off-axis response begins to sag as the midrange reaches
up towards the tweeter and then it flares outward in power
as the tweeter cuts in. This results in a choppy wide
off-axis response (the same thing may happen with the woofer
as it reaches up towards the midrange crossover point) that
I think is very detrimental to proper sound reproduction in
normal listening rooms.

Of course, many guys (ones that some consider to be
knowledgeable) do not consider that wide off-axis response
all that important, provided that the direct-field,
first-arrival signal is flat. Well, I disagree. The
direct-field signal is only a pint-sized fraction of what we
hear with speakers.

and this is where the low sensitivity of the Allison comes into
play-and why he has to double or quadruple them to get reasonable overall
output.

Baloney. The sensitivity may be low on axis, but the overall
efficiency is enhanced by the wide dispersion.

Look at it this way (again): if we have two drivers, one
wide dispersing and one narrow dispersing and measure both
on axis and both are equal, which will play louder with the
same input signal in a typical room? You probably believe
that they would be equal. However, they are not, because the
wide-dispersion job is putting more reflected,
power-response energy into the room. Only in a padded or
anechoic room would the other tweeter be equally loud.

Nobody will accept a tweeter with this sensitivity level today.

Again, the tweeter's wide dispersion makes its
power-response efficiency rather good. And when it comes to
response levels at the listening position it is the
power-response efficiency that counts and not the
direct-field, on-axis efficiency.

Put those tweeters on angled panels, like what we have with
the more upscale Allison models, and at 90 degrees off the
systems are as flat as they are on axis. Of course, I am
being unfair here, because we are discussing drivers and not
systems.

Put our tweeters on angled baffles and the gain is the same. Really Howard,
isn't the angled panel the main contributor to the Allison dispersion? How
audible is a few dB more or less above 10kHz, mostly bounced off the wall,
above 10kHz? You probably couldn't tell the difference given the state of
your hearing.

My hearing "limitations" notwithstanding, the Allison IC-20
sounds more spacious and realistic than systems with
tweeters that disperse in a more conventional manner.

Both types existed when Allison was in business and customers freely chose

Yep. This means that the critical distance between the
direct and reverberant fields is not stabilized with such
designs (changes back and forth in distance from the
speakers as the frequency climbs up the scale), and to my
way of thinking this is a much bigger problem than a lack of
tight imaging and imprecise focus.

To YOUR way of thinking? From someone who doesn't even understand the basic
mechanisms of driver diameter and dispersion? What validity does your view
of "critical distance" bring to this discussion?

If the system disperses evenly at all frequencies, the
critical distance will be stabilized. If the drivers
disperse differently at different frequencies, the critical
distance will move back and forth between the speaker and
the listener. The listener will sometimes be in a
direct-field dominated location and at other times will be
in a reverberant-field dominated location. This ain't no
good, John.

You know, jerks like Heyser,
Toole, D'Appolito, Linkwitz, Barton. What you'll never understand is that if
even if these guys had an Allison tweeter available today, they'd reject it
out of hand, for all the reasons I stated before.

Well, what these guys like or liked involves taste.
Remember, it was you who said that speaker design is as much
an art as a science, and my contention is that speaker
performance is as much a product of taste as of science.

Okay, so these guys are just know nothings with an opinion, but you are the
master?

No. I am not alone in considering imaging and tight
soundstage focus to be an overrated requirement for good
loudspeaker performance in typical listening rooms.

2) Some people like speakers to sound like speakers, rather
than like live ensembles.

Oh please, there's no mistaking your Allisons for a live ensemble.

Says you. How many top-tier Allison models have you listened
to?

Enough of the model 1 to know what it is and to not like it.

Yep, you are a tight imaging guy. I'll just bet that you are
also a pop- or rock-music freak. But rock-music people do
not even need super-quality audio systems.

Heck, remember those live-vs-recorded
concerts Villchur held decades ago with AR-3 speakers? And
he was doing that with speakers that are not in the same
class as the Allison models, although Villchur was, like
Allison, more interested in flat power response and wide
dispersion than in tight-imaging and precise instrumental
focus. Looks like Villchur knew what really mattered way
back in the 1960s.

Those live vs recorded things were parlor tricks. Everyone knows in a large
reverberant space, the acoustics of the space will dominate the sound
character.

It worked. And with speakers that had rather erratic
direct-field performance, loads of diffraction effects, and
drivers designed four decades ago.

Sure, some systems these days are superior, but my point is
that the AR-3 was still good enough in that era to come
pretty damned close to duplicating the sound of live
ensembles. If all of that laser interferometer work and
other design tricks that your team and other driver-design
teams are doing is such a big deal, there would be a huge
advance over what the AR-3 could do. However, even if the
newest designs are somewhat better, they are not all that
much better. Actually, the main advance since the AR-3
involves maybe better midrange dispersion (with the AR-3a
and AR-LST) and then later on flatter mid-bass reproduction
due to the boundary-analysis work done by Allison.

They
sound like speakers, diffuse radiating speakers. Anyway, aren't you the one
that keeps harping on the fact that 2 speakers can't possibly reproduce the
experience of live music in a room?

Sure, but for reasons unrelated to what we are discussing.

Right. So tell me again how a pair of IC 20's sounds like a live ensemble?

Close, though, or at least better than anything else I have
auditioned for magazine reports. If I add in my center
speaker and throw in a bit of good DSP surround ambiance I
am doing even better.

Howard Ferstler

wrote:

Given your opinion of those crappy Allisons in comparison to real high
end speakers like the Dunlavys I am more confident than ever that you
are quite deaf, dumb and blind.

Actually, I like both. The Cantatas image better (at least
from the sweet spot) and have the ability to deliver
somewhat greater direct-field detail. That makes them
particularly good with smaller-scale baroque and
chamber-music materials. However, the IC-20s have the
ability to deliver a better sense of soundstage depth and
envelopment than the Cantatas.

Howard Ferstler

dave weil wrote:

I'll say this about the CD8 - at levels that one would usually
consider medium loud, the speaker as a system starts to distort, so it
isn't a system that one would use to achieve "realistic" levels of
high output music. This wouldn't hold true with the IC-20, which I
have heard stay relatively "clean" at pretty impressive output levels.

Still, the CD-8 stays coherent at levels that satisfies me about 95%
of the time.

It is a good system. Measures about as flat in my big room
as the IC-20, too.

Howard Ferstler

Arny Krueger wrote:

Howard Ferstler wrote:
Arny Krueger wrote:

MINe 109 wrote:
In article ,
Howard Ferstler wrote:

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

I have only mentioned what another rather influential and
knowledgeable individual wrote about it (well, said about it
during a lecture). For all I know, the damned thing hung the
moon. But in principle the system has characteristics that
should make some purists a bit apprehensive.

You haven't said what those are.

The esoteric circuitry that allows for broad-bandwidth
delays between each concentric ring.

It's called 'wire'. You don't think it changes the sound.

Wrong. It's a delay line composed of inductors and capacitors, and
they are each well-known to change the sound.

Finally, some support from you instead of vituperation.

It helps when you are correct about things, Howard.

See my response to those comments of yours. I am going to
add a few more if I can find the thread again.

Howard, I was very polite with you about your errors in those other
posts. If you want to see vituperation, look at the corresponding
posts from that well-known technical ignoramous, Scott Wheeler.

Yeah, he is a bit of a tweap, isn't he.

Howard Ferstler

Howard Ferstler wrote:

Arny Krueger wrote:

Howard Ferstler wrote:
Arny Krueger wrote:

MINe 109 wrote:
In article ,
Howard Ferstler wrote:

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

I have only mentioned what another rather influential and
knowledgeable individual wrote about it (well, said about it
during a lecture). For all I know, the damned thing hung the
moon. But in principle the system has characteristics that
should make some purists a bit apprehensive.

You haven't said what those are.

The esoteric circuitry that allows for broad-bandwidth
delays between each concentric ring.

It's called 'wire'. You don't think it changes the sound.

Wrong. It's a delay line composed of inductors and capacitors, and
they are each well-known to change the sound.

Finally, some support from you instead of vituperation.

It helps when you are correct about things, Howard.

See my response to those comments of yours. I am going to
add a few more if I can find the thread again.

Howard, I was very polite with you about your errors in those other
posts. If you want to see vituperation, look at the corresponding
posts from that well-known technical ignoramous, Scott Wheeler.

Yeah, he is a bit of a tweap, isn't he.

Howard Ferstler

I meant twerp.

Howard Ferstler

Arny Krueger wrote:

Dome tweeters have broad dispersion, especially if they are small -
that's what they do. Furthermore, who's to say that braod, nearly
hemispherical dispersion is even a good thing?

In another response, I mentioned several factors that prove
that wide and uniform dispersion over a system's operating
range is a good thing. I'll list the five of them again and
then add one more that may be the most important of all:

1) It allows the designer to keep the critical distance (the
point between the direct and reverberant fields) between the
listener and speaker stabilized. With directional speakers,
this critical distance moves back and forth between the
speakers and listener. Depending upon the frequency, the
sound is sometimes direct-field dominated and sometimes
reverberant-field dominated. With wide-dispersion drivers
the critical distance does not move so much. If a speaker is
going to be directional, it should stay that way over the
midrange and treble frequencies, and keep directionality
under control.

2) It allows the treble frequencies to be as spacious
sounding as the midrange frequencies. With treble beaming,
this characteristic is not there.

3) It allows a speaker to soundstage more effectively in
typical listening rooms. Placement is also not as critical,
as the speaker does not have to be aimed at the listener.

4) Following number three, above, it allows anyone listening
from off-axis locations to get the same spectral balance
from the speaker as those sitting closer to the speaker's
axis.

5) With the midrange being wide dispersing (particularly in
comparison with systems that have largish midrange drivers)
it reduces the radiation-efficiency sag common at and near
the crossover frequency. Yes, certain horn designs overcome
this, but the result is a speaker that spotlights sound
instead of spreading it out along the soundstage.

OK, here is number six and it is a bit longer than the
others and basically serves as an explanation for number
five:

At wide off-axis angles a typical woofer will be
omnidirectional at lower frequencies and then (because
wavelengths are more directional as they get smaller in
relation to driver diameter) get more directional as the
frequencies it handles work their way upward towards the
midrange crossover point. When the signals go high enough
for the midrange to begin to take over the dispersion of the
whole system flares outward and so we have a flat/dip/flat
situation at wide off-axis angles over the operating range
of the woofer and midrange.

However, as the midrange handles still higher frequencies,
its dispersion also narrows (just like with the woofer), and
so there is a dip at wide off-axis angles as the crossover
point is approached. The tweeter cuts in as the frequencies
continue to climb, and again, the wide off-axis response
flares outward again. So, we again get a flat/dip/flat
situation with the midrange and tweeter interface.

So, while the system may be flat on axis and even fairly
flat at moderately wide angles off axis, at extremely wide
off-axis angles (beyond 45 degrees and out to past 60 or 70
degrees) the response of the system will be choppy.

This choppiness contributes to the power response (and
therefore the room response) and therefore colors the
spectral balance of the system.

Yes, directional horns can keep the response rather narrow
and controlled at wide off-axis angles and can keep those
woofer/mid and mid/tweeter slumps under fair control. (My
Dunlavy Cantatas actually work this way, while still being
rather narrow dispersing.) However, most conventional
systems, even some well-regarded ones, do not enjoy that
feature.

Howard Ferstler

dave weil wrote:

On Tue, 19 Apr 2005 13:42:40 -0400, Howard Ferstler
wrote:

dave weil wrote:

On Sun, 17 Apr 2005 19:54:14 -0400, Howard Ferstler
wrote:

I think that
laser interferometer analysis of cone or dome behavior
probably begins to become a bit obsessive

Cue irony loop...

I'm still chuckling about the charge of serious comb-filtering on the
Quad.

There would have to be on the 57 version.

No more than any other multi-driver speaker of any reasonable size.

With multi-driver systems the spacing occurs only at and
near the crossover points, and will vary depending upon how
far apart the drivers are located. With the 57, the spacing
exists at all frequencies.

Howard Ferstler

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

Yeah. Frankly, if I "figured it out" the result still went
over my head.

You figured out that the delay depends on lots of sonically benign wire.

Sonically benign? You guys debate the merits of upscale
speaker wire leads that are six feet long and you are
telling me that 11 miles of wire is no big deal? Give me a
break.

In any case, Arny showed that 11 miles is still not enough.

Howard Ferstler

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

Pal, you would have to use a LOT of wire to effect a proper
delay line for that concentric-ring design. You would
practically have to make it several miles long.

At last! Yes, Quad use a LOT of wire, plus some phase tricks.

http://www.stereophile.com/loudspeak...16/index6.html

"These rings were fed by delay lines (employing some 11 miles of wire!)

Holy cow. And you guys split hairs and get into a
sound-quality twist about a few feet of speaker cable
between an amp and some speakers!

Isn't the point that I *don't* get into a twist about wire?

Stephen

Well, if 11 miles of wire is sonically benign, then it looks
as if a good speaker hookup involving maybe a dozen feet of
the copper would not require upscale wire to do the job.
Lamp cord would do just fine.

Howard Ferstler

MINe 109 wrote:
In article ,
"Arny Krueger" wrote:

Howard Ferstler wrote:

Yeah. Frankly, if I "figured it out" the result still went
over my head.

Stephen was just trying to be cute. Instead, he exposed his
ignorance.

Says the man who was willing to bet there wasn't a mile of wire in a
Quad...

Says the man who has zero evidence that there is a mile of wire in a
Quad.

MINe 109 wrote:
In article ,
Howard Ferstler wrote:

Arny Krueger wrote:

MINe 109 wrote:
In article ,
"Arny Krueger" wrote:

Howard Ferstler wrote:
Arny Krueger wrote:

MINe 109 wrote:
In article ,
Howard Ferstler wrote:

MINe 109 wrote:

In article ,
Howard Ferstler wrote:

I have only mentioned what another rather influential and
knowledgeable individual wrote about it (well, said about
it
during a lecture). For all I know, the damned thing hung
the
moon. But in principle the system has characteristics that
should make some purists a bit apprehensive.

You haven't said what those are.

The esoteric circuitry that allows for broad-bandwidth
delays between each concentric ring.

It's called 'wire'. You don't think it changes the sound.

Wrong. It's a delay line composed of inductors and capacitors,
and they are each well-known to change the sound.

Finally, some support from you instead of vituperation.

It helps when you are correct about things, Howard.

Howard figured it out.

Prove it.

Yeah. Frankly, if I "figured it out" the result still went
over my head.

You figured out that the delay depends on lots of sonically benign
wire.

Please define "lots" in terms of miles of wire, and show an
independent source that supports your claim.

Howard Ferstler wrote:
MINe 109 wrote:

In article ,
Howard Ferstler wrote:

Yeah. Frankly, if I "figured it out" the result still went
over my head.

You figured out that the delay depends on lots of sonically benign
wire.

Sonically benign? You guys debate the merits of upscale
speaker wire leads that are six feet long and you are
telling me that 11 miles of wire is no big deal? Give me a
break.

In any case, Arny showed that 11 miles is still not enough.

Just to clarify - if this kind of delay is implemented with a delay
line (which includes other electronic parts), then 11 miles is way
more than enough. If it is implemented by depending on just the delay
due to the length of the wire, then it isn't.

Correction:

Arny Krueger wrote:
MINe 109 wrote:
In article ,
"Arny Krueger" wrote:

Howard Ferstler wrote:

Yeah. Frankly, if I "figured it out" the result still went
over my head.

Stephen was just trying to be cute. Instead, he exposed his
ignorance.

Says the man who was willing to bet there wasn't a mile of wire in
a
Quad...

Hyperbole? My well-supported claim was that it would take something
like 166 miles of wire to make the Quad ESL 63 work the way you said
it does. I speculated that there might not be a mile of wire in the
Quad, but instead SP says that there was 11 miles of wire.

Well, 11 miles of wire is still a lot less than the 166 miles of wire
it would take to implement a useful delay based purely on the delay
due to a signal passing through a wire.

Schematics of the Quad ESL 63 clearly show a delay line composed of
inductors and capacitors, not delay due to the length of the wire.

Of course Stephen, they obviously didn't teach you how to recognize a
delay line on a schematic in your music classes, so how would you
know?

I don't know which is worse -being pecked to death by ducks or being
pecked to death by a music teacher, a waiters, and a makeup artist,
each with his own delusions of grandeur.

Arny Krueger wrote:

Just to clarify - if this kind of delay is implemented with a delay
line (which includes other electronic parts), then 11 miles is way
more than enough. If it is implemented by depending on just the delay
due to the length of the wire, then it isn't.

Just how much space would 11 miles of wire occupy? If this
wire is anything but micro-thread size the speaker would
have to be huge - and damned heavy.

The same people who ignore the problems with having 11 miles
of skinny wire inside of a speaker probably use wire between
their speakers and amp as big around as garden hose.

So goes the lunatic world of high-end audio.

Howard Ferstler

Arny Krueger wrote:

Hyperbole? My well-supported claim was that it would take something
like 166 miles of wire to make the Quad ESL 63 work the way you said
it does. I speculated that there might not be a mile of wire in the
Quad, but instead SP says that there was 11 miles of wire.

Well, 11 miles of wire is still a lot less than the 166 miles of wire
it would take to implement a useful delay based purely on the delay
due to a signal passing through a wire.

Schematics of the Quad ESL 63 clearly show a delay line composed of
inductors and capacitors, not delay due to the length of the wire.

Of course Stephen, they obviously didn't teach you how to recognize a
delay line on a schematic in your music classes, so how would you
know?

I don't know which is worse - being pecked to death by ducks or being
pecked to death by a music teacher, a waiter, and a makeup artist,
each with his own delusions of grandeur.

This is a very funny post that hits the mark.

Howard Ferstler

Here is a brief, additional reply to Mr. Atkinson, and I
will leave it at that:

There is some diffracted output at more than 90 degrees, but
it is NOT significant compared with the output of the driver
radiating on the right-angle panel.

The crossover frequency between woofers and mids is 350 Hz,
well below Mr. Atkinson's postulated 600 Hz. As for mids,
reinforcement from the front wall can be expected to
increase below 600 Hz by the same reasoning that some
cancellation is to be expected from the other mid. All you
need do is put the Model Ones or IC-20s in the recommended
position with respect to the front wall.

Wrapping it up, it has been suggested to me that it is
pointless to argue with a magazine editor. I should have
realized that at the outset.

Howard Ferstler

John Atkinson wrote:
ScottW wrote:
The fig3 plot says "Fig.3 Rogers LS3/5a, 1978 sample, horizontal
response family at 50", normalized to response on tweeter axis,
from
back to front: differences in response 180 degrees-5 degrees
off-axis;
reference response; differences in response 5 degrees-180 degrees
off-axis."

Yes, the plot show the LS3/5a's output plotted in a complete circle
at
5 degree intervals.

So the 0 degree plot should either agree with the on axis (fig 4
which
is also at 50") or if truly normalized to on axis then I think the
0
degree plot should be flat. I can't see either situation in Fig 3.

The fig.4 response was subtracted from every response in fig.3.
meaning
that the 0-degree trace is indeed a flat, straight line.

OK, I see it now.... jeez I hate 3-d plots.

But the notable point is that the tweeter output doesn't stop dead at
90 degrees off-axis. Below 12kHz, there is still output apparent
above
-30dB floor of this graph, particularly at some frequencies.

You sure this tweeter doesn't have a magic surround as well? Some
tweeko-freeko types are claiming surrounds can have an effect on high
frequency radiation patterns and dispersion .

ScottW

"Arny Krueger" wrote in message
...

I don't know which is worse -being pecked to death by ducks or being
pecked to death by a music teacher, a waiters, and a makeup artist,
each with his own delusions of grandeur.


It's hard to tell what might annoy a disabled ashtray enginer.



----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Clyde Slick said:

It's hard to tell what might annoy a disabled ashtray enginer.

Getting tossed out of the Promise Breakers, his pisant little Baptist
church, and RAH-E certainly took their toll.

"Howard Ferstler" wrote in message
...
Wrapping it up, it has been suggested to me that it is
pointless to argue with a magazine editor. I should have
realized that at the outset.


You should have learned that when you
submitted plagiarized material.



----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

In article ,
"Arny Krueger" wrote:

Please define "lots" in terms of miles of wire, and show an
independent source that supports your claim.

It's part 'L63CO1A'. Order one and unwind it.

Stephen

Howard Ferstler wrote:
John Atkinson wrote:
the plot show the LS3/5a's output plotted in a complete circle
at 5 degree intervals...the notable point is that the tweeter
output doesn't stop dead at 90 degrees off-axis. Below 12kHz,
there is still output apparent above -30dB floor of this graph,
particularly at some frequencies.

As best I can tell, this horizontal radiation pattern
readout was done at just one height above the ground plane.

That's correct. It's the horizontal dispersion measured on the
tweeter axis with the speaker several feet _above_ the "ground
plane." (Why don't you just say "ground" or is that not
sufficiently technical-sounding?)

Looks like measuring from a slightly different height would
result in a different set of curves, given the interference
effects and different radiation patterns from each driver in
the vertical plane. How can one set of multiple polar curves
be worth anything if a slight measurement-height change will
change the shape of most of them?

Spew!!! Now I've got my milk coming out of my nose! Mr. Ferstler,
did you ever read even _one_ book on loudspeaker engineering?
This graph shows that, contrary to what you claimed Roy Allison
had told you, the tweeter's output doesn't stop 90 degrees
off-axis. End of story, I would have thought. but no, I expect
you will go running to Roy for _another_ explanation only to have
it garbled by the filter of your lack of understanding. :-)

In order to make direct-field measurements like this
meaningful (if we want to see the energy smoothness of the
system into a total room space), not only would you have to
do a full series at one height, you would also have to do a
multiple series of horizontal RP readouts at a lot of other
heights.

No. All you need do is do what I did in the review to which you
and I have been referring: perform radiation pattern measurements
in two orthogonal planes. The speaker's overall radiation can be
inferred from the two sets of measurements. If you had studied
mathematics instead of philosophy at school, Mr. Ferstler, you
wouldn't embarrass yourself in this manner. :-)

John Atkinson
Editor, Stereophile

Clyde Slick wrote:

"Howard Ferstler" wrote in message
...
Wrapping it up, it has been suggested to me that it is
pointless to argue with a magazine editor. I should have
realized that at the outset.

You should have learned that when you
submitted plagiarized material.

I assume that you are a very influential individual within
the industry. In that case, if my "plagiarism" is all that
big a deal for you, I suggest that you contact somebody who
can fix my ass and put me in my place. If I am a real
problem for you as a published writer, grow some gonads and
do something about it.

If you cannot do that, you loutish, forked-tongue,
sockpuppet misfit, I suggest that you just give it a rest.

Howard Ferstler

John Atkinson wrote:

Howard Ferstler wrote:
John Atkinson wrote:
the plot show the LS3/5a's output plotted in a complete circle
at 5 degree intervals...the notable point is that the tweeter
output doesn't stop dead at 90 degrees off-axis. Below 12kHz,
there is still output apparent above -30dB floor of this graph,
particularly at some frequencies.

As best I can tell, this horizontal radiation pattern
readout was done at just one height above the ground plane.

That's correct. It's the horizontal dispersion measured on the
tweeter axis with the speaker several feet _above_ the "ground
plane." (Why don't you just say "ground" or is that not
sufficiently technical-sounding?)

Maybe you are measuring above a concrete patio out in your
back yard. That would not be the ground. I used ground plane
to make sure you knew what I meant. Actually, I was just
being cute.

In any case, so what we have is one measurement in the
horizontal plane, with no reference to horizontal
measurements at other heights. One slice through the sphere
and nothing more.

Looks like measuring from a slightly different height would
result in a different set of curves, given the interference
effects and different radiation patterns from each driver in
the vertical plane. How can one set of multiple polar curves
be worth anything if a slight measurement-height change will
change the shape of most of them?

Spew!!! Now I've got my milk coming out of my nose! Mr. Ferstler,
did you ever read even _one_ book on loudspeaker engineering?
This graph shows that, contrary to what you claimed Roy Allison
had told you, the tweeter's output doesn't stop 90 degrees
off-axis.

Well, this is a subject change, given what I said in the
previous paragraph. I was referring to how the horizontal
readout will change if you move the microphone up or down
even a small distance from your initial location.

End of story, I would have thought. but no, I expect
you will go running to Roy for _another_ explanation only to have
it garbled by the filter of your lack of understanding. :-)

No, he is as tired of you as I am.

Whatever I do, I will not mislead readers by spouting the
kind of claptrap you do - and even more to the point, that
your magazine's contributors do with your editorial
blessing. Yes, John, I realize (even Dave Rich says this)
that you certainly do know how to measure speakers better
than most other reviewers. And as I have said before, you
certainly do know how to make good recordings. The paradox
here is that in spite of this you edit a magazine that is
mostly mumbo jumbo. I suppose it could be said that you are
just stuck with the job.

Say, have you visited Floyd Toole and Sean Olive yet? I am
curious about how your discussion of Ferstler went. Big, bad
Ferstler, picking on all the tweako freakos.

In order to make direct-field measurements like this
meaningful (if we want to see the energy smoothness of the
system into a total room space), not only would you have to
do a full series at one height, you would also have to do a
multiple series of horizontal RP readouts at a lot of other
heights.

No. All you need do is do what I did in the review to which you
and I have been referring: perform radiation pattern measurements
in two orthogonal planes. The speaker's overall radiation can be
inferred from the two sets of measurements.

Inferred? And your readers will understand this? I do not
think that two measurements in two planes will deliver
meaningful data, if what you are aiming for is a useful (to
readers) indication of the overall radiation pattern and
spectral balance.

If you had studied
mathematics instead of philosophy at school, Mr. Ferstler, you
wouldn't embarrass yourself in this manner. :-)

Well, one thing that the study of philosophy taught me,
John, is how to tell the ethical difference between right
and wrong.

Howard Ferstler