Time coherence - how important and what speakers?

This topic comes up every few months. Designing for phase coherence may be intellectually gratifying, but here are my questions for the Phase Coherence Uber Alles crowd.

1. What does time coherence actually sound like? How could you tell a time coherent design from a non-coherent one without
looking at the measurements? Why does almost every online list of "coherent" models include some that are not?

2. If all time coherent speakers are correct, then why do they sound so different from one another?

3. If this characteristic is so important, then why do only a few companies embrace it? and why aren't those companies dominating the marketplace?

4. What about OTHER factors such as distortion, wave interference, off-axis lobing, compression etc.!? Wouldn't these have much more influence over the quality of sound?

5. If this approach is the One True Way to superior sound, then why do people vote us Best Sound At Show?

In response:

Unsound: Please allow me to expand on your comments one by one:

1. I've heard this said before regarding imaging and soundstaging. But I must tell you that as I've examined different topologies, I've found that other things such as
the dispersion characteristics, smoothness of power response, and clean decay corellate more powerfully with imaging and driver integration. Indeed, our designs are almost always singled out for praise in these areas. A very steep cutoff has real advantages because the overlap between the drivers is better controlled, so you avoid the large dips and peaks in response that upset the smoothness of the power response in slow slope crossovers. If you allow the drivers to overlap over a broad range, they may add in phase at a single point in front of the system (and this can be a very very tiny point!) but they will have sharp nulls and peaks
at different points in space. This is why such systems have trouble in three dimensional space. These interdriver cancellations actually make the transition between drivers
easier to spot, at least in my experience.

2. Which leads to my larger point. There are SO many variables in loudspeaker design, from drivers to boxes to
crossover slopes to veneers . Change any of these, and the sound changes, (okay, maybe not the veneer.) Why harp on crossover design as the focal point? It's hopelessy myopic since as you admit, the speakers within the group are all different from one another.

3. Do you seriously believe that? I suppose Dr Floyd Toole, now at Harman International, formerly at Canada's NRC, must lack the resources or talent to create low order crossover based designs because his work corellates well with ours. You can read about the Harman intl facility in Stereophile, and other places.

4. Perhaps yes!

5. I didn't say everyone voted us Best Sound. You're entitled to vote for whatever system you truly loved.
However, more seasoned audiophiles voted for us at that show
than any other system.

Your belief in low order crossovers is just that. And once you've embraced a certain mode of thought, it's difficult to
consider a different one. There's a lack of supporting evidence as to the audibility of the small degree of delay associated with steep slope designs. In fact, the research seems to indicate that it needs to be 3 times the delay that exists in high order networks before it can be barely detected.
But once you believe that you can hear, it will tend to influence your opinion of whatever speakers you listen to.
And thus the belief itself can become the overiding factor in your assesment of a given system's sound.

In response:

Karl,

It's possible for higher order crossover systems to sound "natural". This is one of the litmus tests for me when I'm working on a design. When you've created a system you must be hypercritical of it, and uncover any potential flaw in its sound before it goes out to the rest of the world. When I can sit in front of one of our creations, and forget about the balance, the crossover transitions, the boxes, and just IMMERSE myself in the music, I know we've done it. That, for me is the magic of this endeavor- it's like an audio seance!
You're bringing back some very talented ghosts. That's what I miss on most modern recordings - no ghosts. everything is
synthetic and processed with a plastic sheen to the sound.
Put on a simply miked recording of an orchestra with Heifitz
or a jazz ensemble with Miles or Pops, and the ghosts come back to life! It ain't the time domain, it's the whole reanimation that matters!

I feel the need to try to set the record straight regarding our designs.

The Infinite Slope design we employ is absent of any ringing.
None. Nada. Don't believe me?
Go look at the review of the RM33si in Stereophile, see the waterfall plot of the RM33si. (it's online now)

Yet, the impedance curve is also well-behaved - not a "hard load" for an amplifier at all. Richard Modafferi is a stickler for excellent impedance curves, but then again he has the knowledge of filter theory necessary to properly implement this crossover.

The irony of all this is that initially, both Richard and I believed that phase linearity was an important factor in loudspeaker sound. Richard had concocted many experiments to examine loudspeaker phase. Richard's problem was that he had a stubborn professor who effectively debunked Richard's attempts to prove that it was significant. Richard finally had to come around to the viewpoint that it was not a serious problem. If phase is not such a big deal, then what is? Frequency response, and radiation patterns.

The off axis comb filtering of low order crossovers is a serious problem, which is why companies that employ that method often require the listener to tilt or "focus" the main lobe towards the firmly dictated listening position.

The broad overlap between the non-coincident drivers will create large peaks and dips in response if one moves from the optimium spot. This flaw in the radiation pattern of
slow slope crossover designs may be what certain listeners are accustomed to hearing from a stereo system.
That energy does not disappear into the sunset, but rather is bounced around the room, scrambling the magic "phase coherence" that the designer so carefully
struggled for.

Once I had lived with speakers that are free from this artifact, I found it impossible to go back. Whenever I listen to slow slope crossover designs, I find that I'm easily distracted by the large changes in tonality observed with small shifts in listening position. Which listening axis is correct? And if you can contort yourself into that position, can you stay there and refocus your energies on the music? And if you should stand up, the stark balance shift reminds you that what you're hearing is fake and unnatural.

While the idea of filtering bad recordings through such a speaker may have merit, it's more likely to be dynamic compression and radiation pattern that is taking the edge off the ol' Janis Joplin records, not superior phase response. Remember you're talking about a multi track phase scrambled recording played into a phase scrambled listening room. It's too bad that filter is not defeatable. Anyone who's visited us at shows knows that our speakers are equally at home with Louis Armstrong, Frankie Laine, Bill Evans, Nirvana, Norah Jones, Cassandra Wilson, Elvis Costello & the Soggy Bottom Boys.

Happy Tunes to you!
Jeff

Bigtee,

I agree that there are trade offs in every design, but for me the trade off's in low order crossover designs are too great to allow me to enjoy listening through them. I should mention one exception, however- the Quad ESL 63 with Crosby mod's was
within its limitations a very satisfying speaker. Just don't turn them up too loud, and place them very carefully.

One needs to be able to interpret the test data correctly, in order to identify what is significant and what is not. If you are willing to ignore peaks and dips of 10-14 dB off axis, I suppose that's your perogative.

I'm not laying claim to "PERFECT SOUND". Just "TRUER SOUND".
The problem with "if it sounds good, stick with it" is that
I couldn't find a speaker that I was completely satified with. Each system pointed out flaws in the others. That's why I went into the loudspeaker business to produce Richard Modafferi's Infinite Slope designs. These systems managed to
get more things right to my ears than any other system. And the measurements happen to bear that out.

Jeff

I actually like some of my competitors! (even if I have differing ideas regarding design execution) I've met Roy many years ago at the Florida show and he struck me as a very nice fellow. Same for Pat McGinty.
David & Sheryl Lee Wilson came by and sat through my presentation during the last show, and David stayed and shook my hand afterward. Peter McGrath dropped by shortly thereafter. I've called upon guys like Ken Kantor and Michael Kelly for their advice, and they've always been gracious and accomodating. While the product is very important, a speaker company must be more than the product. The most sucessful companies know that you must take care of the customer after the sale. Ya gotta admire the people at Vandersteen, Thiel & B&W who take care of their consumers and do right by them. It's no small task, and they have set a standard of service that we try to emulate every day.