DiAural: Is it for real?
jer...@my-dejanews.com
No more crossover network, "crosstalk" network is hot! What you think?
http://www.stereophile.com/shownews.cgi?416
Jazzbo8
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Arny Krüger
<jer...@my-dejanews.com> wrote in message
news:7eep82$a89$1...@nnrp1.dejanews.com...
> Check out the story at Stereophile on Kimber's new DiAural speaker
technology.
> No more crossover network, "crosstalk" network is hot! What you think?
>
> http://www.stereophile.com/shownews.cgi?416
>
Doppler distortion in speakers can be real. One class of speaker that is
very susceptable to this kind of problem are speakers that have drivers
that cover a wide frequency range. Perhaps the best documented and most
widely publicized examples of speakers with doppler distortion were
provided by Klipsch's tests of the Bose 901.
It turns out that Doppler distortion is a form of FM distortion, a form of
distortion that is very common and often audible in analog tape and vinyl.
It was pointed out at the time that while Bose 901's do have substantial
doppler distortion when pushed, at "normal" listening levels, the Doppler
distortion in 901's is about the same as that in analog tape and vinyl.
The levels of FM distortion in many forms of analog tape and most vinyl is
measurable, and is measured to be at levels that can be reliably heard.
This is one of those things that show up very nicely in DBT's. Given that
piano music is one of my faves, and piano music is a form of program
material that shows up FM distortion, anything that reduces FM distortion
to inaudible levels is something that interests me. Hence my attraction to
digital.
http://www.stereophile.com/shownews.cgi?416 says:
"Five weeks ago we reported that Ray Kimber, of Kimber Kable fame, and his
financial partner, Bruce Bastion, were in the process of bringing a new
loudspeaker technology to market. DiAural, as they've named the technique,
is claimed to eliminate what Kimber calls Doppler-encoding distortion---the
modulation of high frequencies by low frequencies of higher amplitude."
This roughly corresponds to the classic understanding of Doppler
distortion.
W/R/T classical understandings of how to reduce FM distortion in speakers,
this article seems to make some false claims, such as
"According to Kimber, this encoding takes place in the microphone, survives
intact throughout the recording and playback chain, and, if left undecoded,
seriously detracts from any audio system's verisimilitude."
Actually, Doppler distortion is due to velocity and changes in velocity.
Speakers like the Bose 901 are specially exposed to it because their
woofers and tweeters are the same driver. When the speaker's cones travel
at high velocities to produce bass, treble is FM modulated by the Doppler
effect. If treble is reproduced by a different cone, or if the speaker cone
is prevented from attaining a high velocity, then doppler is greatly
reduced. Thus, the classic two-way speaker system has greater resistance to
Doppler distortion than a single-way. Also, this means that one of the
benefits of having a subwoofer is reduced Doppler distortion.
As a rule, microphone diaphragms don't move fast enough to cause doppler
distortion. Ever see a microphone diaphragm stroke like a small woofer
being driven with heavy bass? Nope - most mics would be physically damaged
if they tried to do that! Therefore, the "coding" mentioned here is
unlikely to happen.
"The DiAural technique, developed by designer Eric Alexander, performs
decoding at the speakers themselves, using a couple of common parts
configured in a novel way that enables a woofer and tweeter (or woofer,
midrange, and tweeter) to "talk back" to each other, freeing them to
jointly produce sounds more like those that originally impinged on the
microphone."
Sounds like low slope crossovers or some other technique that reduces the
isolation betwen the drivers. This is old news, and generally found to be
the exactly wrong thing to do if the goal is to minimize Doppler
distortion.
"The DiAural circuit replaces the traditional crossover network, which
means the drivers are connected directly to the amplifier. All of which
made for an interesting telephone discussion I had with Kimber before
writing this story."
If the speakers are "connected directly to the amplifier", then there is no
crossover network. This is actually the worst case for creating Doppler
distortion.
Joseph Lee Plaziak
>
> Check out the story at Stereophile on Kimber's new DiAural speaker technology.
> No more crossover network, "crosstalk" network is hot! What you think?
>
> http://www.stereophile.com/shownews.cgi?416
>
>
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
I'd sure like to get more info on the specifics of how this is supposed
to work.
My intuition tells me it could be "a good thing". But, as Arny sez,
microphone element excursion is so small that suseptability to doppler
distortion seems unlikely. However, the mic IS picking up low f
information, so it would seem that even if the mic element excursion is
small, the doppler effect would still be present, albiet on a smaller
scale.
I can't figure how a compensatory system would be implemented to cancel
this type of distortion. I think it could be done by varying tweeter
phase relative to woofer position, but the control would have to be
dynamic.
It seems like to do this the system would have to be implemented
actively. Variably retarding tweeter phase at low audio frequencies
shouldn't be too hard to do.
Joe
tubeguy.com
Gary C. Eickmeier
> Doppler distortion in speakers can be real. One class of speaker that is
> very susceptable to this kind of problem are speakers that have drivers
> that cover a wide frequency range. Perhaps the best documented and most
> widely publicized examples of speakers with doppler distortion were
> provided by Klipsch's tests of the Bose 901.
I have never placed much stock in tales of doppler distortion. I base
this not on experience or testing, because I am not an engineer, but
just thinking about it... The microphone diaphragm is a single element,
so it would seem that the most accurate way to reproduce its movements
would be to have a single element driver covering the full range. I
mean, if the story is that you don't want to have these high freqs
riding on the low freqs, that is exactly what happens in the microphone,
so it would seem to me that using multiple drivers and crossovers would
fail to decode, or reproduce, the same effect of the sound waves
impinging on the diaphragm.
> It turns out that Doppler distortion is a form of FM distortion, a form of
> distortion that is very common and often audible in analog tape and vinyl.
> It was pointed out at the time that while Bose 901's do have substantial
> doppler distortion when pushed, at "normal" listening levels, the Doppler
> distortion in 901's is about the same as that in analog tape and vinyl.
So what you are saying is that at the larger excursions of a speaker
compared to a microphone, the movements don't translate one for one? Of
course, there are many other effects happening with greater excursions
(cf Dave Clark's work), so maybe we're not doing a pure experiment here.
I use subs with my 901s simply to relieve them of doing the very low
frequencies, so that the higher freqs can be cleaner. One observation
about the sound of my 901s is that they have a LOT of slam, or dynamic
impact, and I have certainly never heard any distortion from them. I
mean, how much distortion have you heard from Quad ESL-63s?
Barry's story sounds exciting, but I am suspicious, and I sure hope for
his sake that this is not another subjectivist tale, or even a hoax.
> As a rule, microphone diaphragms don't move fast enough to cause doppler
> distortion. Ever see a microphone diaphragm stroke like a small woofer
> being driven with heavy bass? Nope - most mics would be physically damaged
> if they tried to do that! Therefore, the "coding" mentioned here is
> unlikely to happen.
In other words, there is no "distortion", doppler or otherwise, as long
as the mike stays within its linear operating range. I would think the
same applies to loudspeakers. Maybe their circuit is nothing more than
an infinite slope crossover and a limiter. Might it work like Velodyne's
sub feedback mechanism?
Gary Eickmeier
Paul Wagner
>... If treble is reproduced by a different cone, or if the speaker cone
> is prevented from attaining a high velocity, then doppler is greatly
> reduced. Thus, the classic two-way speaker system has greater resistance to
> Doppler distortion than a single-way. Also, this means that one of the
> benefits of having a subwoofer is reduced Doppler distortion.
> distortion....>
My first thought when reading this was: bi-wiring.
That, though, seems to be an ineffective solution.
I can't figure out why, though... can you explain?
--PW--
jer...@my-dejanews.com
quality even if it is indeed an issue. Most recording these days are multi-
mic'd and processed in the studio. How can a simple circuit possibly correct
multiple mic'd sources with different instruments (thus frequencies).
According to the article, Kimber's circuit can be inserted anywhere in the
playback chain and have the same effect. So it acts like a complex tone
control that shapes the playback to make it sound more "realistic". Can you
patent a tone control?!
Arny Krüger
Paul Wagner <paulw...@earthlink.net> wrote in message
news:370B25...@earthlink.net...
> Arny Krüger wrote:
>
> >... If treble is reproduced by a different cone, or if the speaker cone
> > is prevented from attaining a high velocity, then doppler is greatly
> > reduced. Thus, the classic two-way speaker system has greater
resistance to
> > Doppler distortion than a single-way. Also, this means that one of the
> > benefits of having a subwoofer is reduced Doppler distortion.
> > distortion....>
>
> My first thought when reading this was: bi-wiring.
I was hoping for something effective... ;-(
> That, though, seems to be an ineffective solution.
Agreed.
> I can't figure out why, though... can you explain?
http://www.pcavtech.com/techtalk/biwire/biwire.gif shows some equations
that might shed a little light...
Arny Krüger
Gary C. Eickmeier <image...@worldnet.att.net> wrote in message
news:370B789D...@worldnet.att.net...
> Arny Krüger wrote:
>
> > Doppler distortion in speakers can be real. One class of speaker that
is
> > very susceptable to this kind of problem are speakers that have drivers
> > that cover a wide frequency range. Perhaps the best documented and most
> > widely publicized examples of speakers with doppler distortion were
> > provided by Klipsch's tests of the Bose 901.
>
> I have never placed much stock in tales of doppler distortion.
Oh, it happens to an audible degree. I learned all about it as the owner of
Ohm F's. Of course it was kinda hard to separate the AM distortion from the
FM (Doppler) distortion.
>I base
> this not on experience or testing, because I am not an engineer, but
> just thinking about it... The microphone diaphragm is a single element,
> so it would seem that the most accurate way to reproduce its movements
> would be to have a single element driver covering the full range.
If output levels weren't an issue, this could be true. Unfortunately we
want to hear what our speakers reproduce at decent volumes in regular
listening rooms!
>I
> mean, if the story is that you don't want to have these high freqs
> riding on the low freqs, that is exactly what happens in the microphone,
> so it would seem to me that using multiple drivers and crossovers would
> fail to decode, or reproduce, the same effect of the sound waves
> impinging on the diaphragm.
I think we want this effect to vanish below audibility.
> > It turns out that Doppler distortion is a form of FM distortion, a form
of
> > distortion that is very common and often audible in analog tape and
vinyl.
> > It was pointed out at the time that while Bose 901's do have
substantial
> > doppler distortion when pushed, at "normal" listening levels, the
Doppler
> > distortion in 901's is about the same as that in analog tape and vinyl.
> So what you are saying is that at the larger excursions of a speaker
> compared to a microphone, the movements don't translate one for one?
Agreed. Doppler distortion is caused by having the diaphragm moving with
substantial velocity. Due to the small distances moved, this is rarely an
issue with microphones.
> Of
> course, there are many other effects happening with greater excursions
> (cf Dave Clark's work), so maybe we're not doing a pure experiment here.
Agreed. But Clark, like just about everybody else, generally routes the
high and low frequencies to separate drivers. So, for the stuff he designs,
Doppler is far less of an issue.
> I use subs with my 901s simply to relieve them of doing the very low
> frequencies, so that the higher freqs can be cleaner.
This works.
> One observation
> about the sound of my 901s is that they have a LOT of slam, or dynamic
> impact, and I have certainly never heard any distortion from them. I
> mean, how much distortion have you heard from Quad ESL-63s?
I think ESL-63's protect themselves from amps producing enough power to
cause distortion in the basic speaker.
> Barry's story sounds exciting, but I am suspicious, and I sure hope for
> his sake that this is not another subjectivist tale, or even a hoax.
The speakers might sound wonderful. But again, what is actually happening
here?
> > As a rule, microphone diaphragms don't move fast enough to cause
doppler
> > distortion. Ever see a microphone diaphragm stroke like a small woofer
> > being driven with heavy bass? Nope - most mics would be physically
damaged
> > if they tried to do that! Therefore, the "coding" mentioned here is
> > unlikely to happen.
> In other words, there is no "distortion", doppler or otherwise, as long
> as the mike stays within its linear operating range.
Agreed.
> I would think the same applies to loudspeakers.
Agreed. Of course, FM distortion is a form of nonlinear distortion so we
are really making a tautology here... ;-)
> Maybe their circuit is nothing more than
> an infinite slope crossover and a limiter. Might it work like Velodyne's
> sub feedback mechanism?
Feedback needs an outside source of power.
Gary Sanford
snip...
>In other words, there is no "distortion", doppler or otherwise, as long
>as the mike stays within its linear operating range. I would think the
>same applies to loudspeakers. Maybe their circuit is nothing more than
>an infinite slope crossover and a limiter. Might it work like Velodyne's
>sub feedback mechanism?
>
>Gary Eickmeier
Remember your ears are a single full range transducer also. If
"doppler distortion" is so prevalent you think that full range
electrostatic speakers would be prone to it. Instead, they
are some of the most transparent.
~~~~~~~~~~~~~~~~
Gary Sanford
sanf...@ibm.net
Arny Krüger
Gary Sanford <sanf...@ibm.net> wrote in message
news:370bcfc4...@news3.ibm.net...
Doppler distortion in speakers comes from 2 things, both being necessary:
(1) A cone or diaphragm that moves at a relatively high velocity.
(2) Reproducing a wide range of frequencies through that cone or diaphragm.
Full range electrostats typically have negligable diaphragm motion, and
therefore they bypass condition (1), even though they fit condition (2).
The absence of condition (1) rules in this case.
If you look at the drivers in a Bose 901, they can fit both conditions,
particularly if the 901's are used as full range speakers.
Arny Krüger
<jer...@my-dejanews.com> wrote in message
news:7egjov$qvm$1...@nnrp1.dejanews.com...
> I still don't see how Doppler correction could be used to improve the
sound
> quality even if it is indeed an issue. Most recording these days are
multi-
> mic'd and processed in the studio. How can a simple circuit possibly
correct
> multiple mic'd sources with different instruments (thus frequencies).
It would be tough.
Worrying about doppler distortion in microphones makes little sense to me
because microphones fail to present one of the two required conditions for
Doppler distortion, and that is high-velocity diaphragm motion.
> According to the article, Kimber's circuit can be inserted anywhere in
the
> playback chain and have the same effect. So it acts like a complex tone
> control that shapes the playback to make it sound more "realistic". Can
you
> patent a tone control?!
It's been done in the past, and I don't see why not in the future. Patents
don't really require that the patented device be effective or efficient.
Gary C. Eickmeier
>
> <jer...@my-dejanews.com> wrote in message
> news:7egjov$qvm$1...@nnrp1.dejanews.com...
> > According to the article, Kimber's circuit can be inserted anywhere in
> the
> > playback chain and have the same effect. So it acts like a complex tone
> > control that shapes the playback to make it sound more "realistic". Can
> you
> > patent a tone control?!
>
> It's been done in the past, and I don't see why not in the future. Patents
> don't really require that the patented device be effective or efficient.
I am fascinated by the fact that they would be dicking with the
waveform. I mean, the audio waveform represents the exact movements of
the microphone diaphragm, right? Of course, I realize that we don't need
exact waveform reproduction for the sound to remain audibly perfect
(there can be phase delay), but what would this process be doing to the
waveform? Which has always been held sacrosanct in high end circles.
I don't think they said it could be inserted anywhere in circuit. They
said it substituted for the crossover, and that the drivers were enabled
to talk to each other in order to straighten out the doppler distortion
"problem."
Now, as you have said, there is no "problem" within the waveform itself,
so they can only be correcting for their particular drivers, if bass
frequencies attempt to carry too many high frequencies on them - or
something. I just thought that this was supposed to be taken care of by
having separate woofers and tweeters. I'm so confused.
Gary Eickmeier
Joseph Lee Plaziak
> Now, as you have said, there is no "problem" within the waveform itself,
> so they can only be correcting for their particular drivers, if bass
> frequencies attempt to carry too many high frequencies on them - or
> something. I just thought that this was supposed to be taken care of by
> having separate woofers and tweeters. I'm so confused.
>
> Gary Eickmeier
Kimber is saying that there IS a problem with the signal, that the
effect is already present in the signal, so it wouldn't matter if the
speaker had one driver or many. The biggest reason for having separate
drivers is to use a driver that is best for that particular band. I
think what Kimber is saying the circuit will do is correct for whatever
doppler effects there would have been in the signal already.
Imagine a speaker going in and out. The cone goes at a certain speed as
it travels in or out. If that cone is being asked to go in and out at
many different speeds at once, what can happen? Say the cone happens to
be moving out right now, during a strong bass note. While the cone is
moving out slowly to make the bass tone, it is at the same time moving
in and out by little bits as it travels, trying to reproduce higher
frequencies at the same time. The cone may have vibrated at the same
rate while it was moving in and out, but the higher frequency vibrations
would have been effected by the cones larger overall motion, caused by
the bass note, stretching and compressing the wavelength as it goes,
effectively changing the frequency at a rate equal to the larger signal
(the strong bass note). That's the doppler effect. Ray Kimber would have
you believe he has figured out a way to cancel out that effect with any
playback system by compensating somehow for doppler distortion already
in the signal.
The real question is the audibility of making this type of correction to
the signal.
Joe
tubeguy.com
Arny Krüger
Joseph Lee Plaziak <tub...@mailhost.chi.ameritech.net> wrote in message
news:370D1DC3...@mailhost.chi.ameritech.net...
>
> Imagine a speaker going in and out. The cone goes at a certain speed as
> it travels in or out. If that cone is being asked to go in and out at
> many different speeds at once, what can happen? Say the cone happens to
> be moving out right now, during a strong bass note. While the cone is
> moving out slowly to make the bass tone, it is at the same time moving
> in and out by little bits as it travels, trying to reproduce higher
> frequencies at the same time. The cone may have vibrated at the same
> rate while it was moving in and out, but the higher frequency vibrations
> would have been effected by the cones larger overall motion, caused by
> the bass note, stretching and compressing the wavelength as it goes,
> effectively changing the frequency at a rate equal to the larger signal
> (the strong bass note). That's the doppler effect. Ray Kimber would have
> you believe he has figured out a way to cancel out that effect with any
> playback system by compensating somehow for doppler distortion already
> in the signal.
So, Kimber presumes that the signal had doppler distortion in it, before it reached
the speaker. The means by which this could happen is not clear.
> The real question is the audibility of making this type of correction to the
signal.
Doppler distortion is FM distortion, just like jitter. Presumably, current wisdom
about the audibility of jitter would apply. Probably the most recent source of
information on that topic is AES Preprint 4826, "Theoretical and Audible Effects of
Jitter on Digital Audio Quality", availble from www.aes.org.