|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
Can anyone explain how Von Schweikert and Spica are able to get phase coherence in their high order cross-overs?Most people claim the only way to get phase coherence is with a first order crossover.
Thanks,
Mike
Follow Ups:
Here are some questions which I posed to Albert a couple of years ago when I was looking for speakers, along with his answers.Hope this helps.
Does your GAIN crossover resemble a first order crossover in a way which
enables the speakers to reproduce the "naturalness" of the sound, as is the
claim of
speaker designers which use first order crossovers?
> > > > > > > > > > > > > > Yes, my circuit consists of multiple first order crossovers
staggered over several octaves. The purity of sound is similar to a simple
first order but eliminates the problems of first order networks.
What is the difference between phase coherence and phase consistency, and
how does this factor into your designs?
> > > > > > > > > > Phase coherence in it's classic sense means that the drivers all
produce a similar pressure wave at the same exact moment in time, which is
obviously a good thing. However, this accuracy can only be held over a very
narrow window, usually about 5 to 15 degrees within the tweeter's direct axis.
Since there is so much driver overlap, the pressure waves do not sum to unity
when the distance between the drivers and the measurement mic/ear becomes non
symmetrical. The resulting dips and peaks in the amplitude pressure response
can vary up to 18dB (!) with corresponding lack of transient and phase
accuracy. When measuring any speaker with a first order crossover, the phase
is not coherent at locations beyond a 30 degree horizontal axis. This is due
to uneven path lengths from each driver to the mic (or ear) which do not sum
to unity. In contrast, I have chosen a 2ms window of time coherence between
the drivers which does not vary over a very wide window. This 2ms figure came
from studies in psychoacoustics which indicates that the ear can not
discriminate between rapidily repeating pulses which occur faster than the 2ms
between pulses. For this reason, optimizing the transient responses of the
different drivers to occur simultaneously has no audible advantage.
(Psychoacoustics, Zwicker and Fastl, Springer-Verlag, Berlin, 1990).For this reason, the fourth order acoustic response is quite superior in phase
response, as the interdriver delay is kept within the 2ms tolerance range.
Since the
frequency response is now very uniform over a very wide arc, both horizontally
and vertically, the Hilbert transform states that the phase will also be
consistent over the same arc. Now, why is phase consistency such a big deal?
Well, you can hear a three dimensional sound stage anywhere in the back of the
room, just as with a live performance. Even if you sit in the "sweet spot,"
the reflections from the room boundaries must be as flat in frequency response
and with the same phase as the direct signal, so that the room's reverb
matches that of the speaker itself. Only in this manner can the ear/brain
hearing mechanism "believe" that the sound could be live.
Mike,Several years ago John Dunlavy and Albert V. crossed swords in public regarding their loudspeaker designs. I believe it started with John Dunlavy challenging the technical claims made on the Von Schweikert web site (the original web site, around the time of the VR-4/VR-4.5/VR-6 models). Here is a snippet of their rather long on-line debate, pertaining to your question:
John Dunlavy said:
".... Albert has frequently and emphatically claimed that his loudspeakers "exhibit phase-coherent performance using a 4th-order network". Not so, as several posters with excellent technical/professional credentials have explained. Phase-coherent (pulse-coherent) performance, on-axis, cannot be achieved using any type of passive, 4th-order crossover network. Period!
Nor can a passive 4th-order crossover yield accurate impulse, step and
phase responses. It is a priori impossible!"Albert V. replied:
"John, later technology refutes this, I'm afraid, such as in Baekgaard's paper on using "filler" drivers with 2nd, 3rd, and 4th order crossovers in order to achieve perfect pulse coherency. (See his paper in the Journal of the Audio Engineering Society, May 1977.) Also, I have never claimed that my crossover design replicates a pulse with first order behavior. Please reread my statement on this for clarification. I said that there are benefits of the 4th order crossover which are superior to a first order, and these benefits
are: 1) superior vertical off axis measurements, 2) superior power handling and 3) clarity due to lack of excessive driver interference in their stop band.Obviously, first order crossovers will pass a square wave, while 4th order networks exhibit approximately a 0.5ms delay in processing the pulse. I am not debating you further on common knowledge, and I am growing weary of your statements that I don't know anything about engineering.
You believe that a perfectly reproduced pulse is audible, and that this makes your design "automatically" superior. However, I and many other engineer who use steep order slopes have found that this delay is not audible and that the advantage of a perfect impulse response does not guarantee listening superiority...."
Etc.
Rich H.
Where can I find the whole debate?
John,When the debate was raging, I saved one of Albert's responses to John Dunlavy. Far too long to post here. In the response, Albert reprint's Dunlavy's accusations and queries, responding to each one
sequentially. So, essentially, most of the debate can be gleaned from Albert's post.Email me (through the Asylum, click on the "R" by my name) and I'll send it to you.
Rich H.
Tell me too! Yes and what is the "ambience recovery" the Vons use; just rear tweets or something more mystereous?
This post is made possible by the generous support of people like you and our sponsors: