Time Phase/Coherence-Vandy 5A vs. SL M-1

There are some very interesting and informative threads in the archives. Type in "phase coherent" and "time coherent" in the search archive. I'm a believer in time and phase coherent speakers,assuming all other parameters are properly dealt with. Still, what ultimately matters is what speakers sound best in your system and in your particular listening room.

Disclaimer - I peddle Sound Labs.

Okay I'm wading into deep waters here, not being an engineer. But maybe that's okay - I'm a dealer after all, fin on my back, perpetual toothy grin, hmmmm... maybe these deep waters will provide an opportunity or two. A tasty exchange of ideas, if you will. Food for thought, perhaps. Cue the theme from "Jaws"...

I think I found the quote Fgm4272 was referring to. In an interview published in Soundstage online magazine (http://www.soundstage.com/interviews/int07part2.htm), Richard Vandersteen said:

"Panels tend to be large and the distances from the panel to the listener’s ear are different for each point you measure. A sound radiating from a panel has many points of origin on the panel and those all have different distances to travel to get to the listener’s ear. So panels introduce quite a bit of time smear for the most part."

Richard Vandersteen's assertion would be true if panels did not radiate as coherent line sources or planar sources (which they do, as can be demonstrated with fairly simple measurements).

If Richard's position was correct - that is, if we heard the sound from each point on a planar diaphragm at a different instant (requiring that the diaphragm not behave as a line or planar source) - then the output from such a panel would be so smeared that intelligibility would be significantly degraded (see Helmut Haas's landmark paper entitled "The influence of a single echo on the audibility of speech", JAES March 1972; and James M. Kates' "A perceptual criterion for loudspeaker evaluation", JAES December 1984). This smearing would be easily revealed by an oscilloscope trace of the speaker attempting to reproduce a square wave. I recall a very detailed brochure publised by DCM years ago demonstrating how close their speakers came to reproducing a square wave - bettered only by the (drum roll please....) flat panel Quad ESL!! (Now commonly known as the "57".)

Roy Johnson of Green Mountain Audio, another designer of well-respected time and phase coherent loudspeakers, recently wrote:

"Full-range single panel speakers are/can be minimum phase - and IMO why they are enjoyable, even at the expense of sitting dead on center because of that large panel vs. the short treble wavelengths."

Here's the link (scroll about 3/4 of the way down the first page to the last paragraph of Roy's post): http://www.audiocircle.com/circles/viewtopic.php?t=9652&postdays=0&postorder=asc&start=0

[Note that the faceted-curved panel of the Sound Labs avoids the small sweet spot problem Roy mentions.]

Now despite all the foregoing devastatingly convincing arguments (saying that tongue in cheek, folks), I think what really matters most is what Fgm4272's ears tell him. And if it's a tie, then I'll readily concede that the Vandies kill the SLabs in the WAF department.

Cheers,

Duke

Typo - should have been "Fmg4275". Sorry.

Planars do have problems that are measurable. However, all speakers are a trade off in one way or another.
Refering to the Soundlab, If you look at the transfer function and impulse response graph, it is fairly easy to see what Richard was talking about. These speakers do smear the signal over an extended time frame. They also artificially produce spatial effects by the dipole radiation pattern(they store and release energy from the stretched edge clamped diaphragm) with energy bouncing of the front wall behind the speaker. This is why a lot of people like planars, they here more "Ambience" but it's not real.
I am not a speaker guru, but I can read and interpret measurements.
As I said, every speaker has problems and the idea is to not create more but solve the ones that exist. Ultimately, it is what sounds best to you.
I certainly would not want to get into a dispute with Richard Vandersteen or any "Real" designer for that matter. They know the weakness and strengths of all designs and each chooses a path. When Vandersteen started out, he wanted to mimic a Quad because back in the 70's, the ESL63 was the speaker that defined midrange. He wanted to create a "Full" range speaker with that same majic in the mids without the drawbacks. Times change and technologies move forward.
My one last statement would be that if, I "Had" to pick a planar, the Soundlab would be at the top of my list. In my opinion, it gets more right than the others.

Odd, this is almost the same question on E'stats versus dynamics, but without Sean here to talk about himself. (So, now that I have you, let's talk about me, HA) Just a background here.)
Ok, I worked for Jim THIEL as Vice President of Sales for some time, so I know of him and also, through shared stories, about Richard, who is, by all accounts a good guy, who makes an honest product.
Thiel started in 1976 or 1977 depending on whether you're talking to Jim or Kathy. HA!(Think crows feet)
Anyway, Jim and Richard (Vandersteen), sort of grew up together (businesswise) in terms of their product development.
Jim immediately after building his first "box" speaker which was equalized, began to realize the advantages of time and phase coherence. To wit: the speakers are CS designated, which reperesents a shortened version of that name, Coherent Source which is trade marked by them.
Without going crazy on the technical side, which I am afraid I am about to here, I would first ask YOU to buy with your ears, (as someone else here did) not your eyes, and brain. However,in all fairness, listening tests generally are kind to Phase and Time Alignment, based on many other parameters, with the result that Time Alignment and Phase coherence, produce results which add to the sense of realism with the following caveats.
Below, we will discuss at the beginner level, the phase/time coherent issue. Then we will discuss briefly SL's as to whether the SL is Time and Phase coherent, and or the issues that Panel's face.
Time alignment usually causes little enough disagreement, so I will begin there.
Thiel noticed, that the tweeter when sitting normally, several feet from a speaker, is simply closer to the listener. No, higher frequencies don't travel faster than low frequencies, a common thought process, though I am willing to listen to a compelling argument on any subject)
So, Thiel simpy slanted the speaker on a rearward, cant, allowing that the distance from the drivers (to your ear), at each of their stations would be about the same. (Remember that the tweeter is a convex rather than concave device also.)
That should take care of that. Now the real juice.
THIEL's goal (and Vandersteen's) (simplified a million times over) is to have the drivers positively pulse, (in accordance with the input signal) at the same time to create the signal. Example, on a piano, Concert A below Concert C is 440hz.(cycles per second). Some loudspeakers find this note (think three way) in the middle of a crossover. The woofer is asked to play the funamental (dominant tone of this note,is centered at 440 cycles persecond) and the midrange, tweeter drivers, have to play the harmonics of that tone. A harmonic is the first (and second) octave(s) above that, or double the frequency, or in this case 880hz, the second harmonic 1760hz.(This basic thought keeps extending into the netherregions) Then as is the case with all but perfectly generated sine waves, the second harmonic is double that again. Each time you go up eight notes on the (Occidental not oriental scale, say Western versus Asian), that being doe, re, me, fa, so, la, ti, doe, you have doubled the frequency (one time) of the original note. Harmonics are there on all the musical notes, (again except for computer generated sine waves): harmonic structure, is what gives the note of a piano its own unique, characteristic sound, versus, say a trumpet or clarinet, or simply put, that is how we recognize them from one another.The goal of the THIEL speaker, and Vandersteen, is to play that dominant, and first and second and third harmonic (and so on), BACK as it was originally produced. To do that, without changing the basic sound, one must play all that data, back in the same time function, ie, all (call them secondary sound) harmonics start at the same time, and traveling together kinda. (Think of waves from a pond traveling outward).
WHEW, if I haven't lost you due to boredom...
This sounds easy to accomplish, but its not. The drivers play the dominant, but then when it comes to the harmonic, there is a time delay, (phase shift) causing the note to be replicated with the dominant, to be reproduced, but without the harmonics, happening at exactly the same time.
Now, from our engineering we learn that any change from the original to the replicated product, (the actual note we are speaking of) is a form of distortion, or in this case, phase smear, or whatever the people in charge choose to call it this year.
Jim Thiel and Richard V. and many others design their speakers, with what is known as first order crossovers, in order to induce minimum phase 'shift'. . Wow, again with the boredom.Sorry, but you asked.
First order, means that the woofer, starts to attenuate (roll off, or lessen in output volume) at a rate of 6 decibels per octave. So as it travels up in frequency, from its (cross over dictated point)its output is lessened by six db. This, introduces a minimal amount of phase "shift", or change.
Ok so let's assume that the only consideration here, is to replicate the note in its purest form, assuming that this is the only goal at least for the context of this explanation, which in a perfect world, it would be. Right?
What if, though, by making the speaker, (and all its drivers) play all these notes and harmonics, in virtually perfect phase replication, we caused other problems that were more dastardly. (GOD I've wanted to use that word in a sentence since Roosevelt used it right after Dec 7th 1941. Just kidding.
What if seriously, the result, in the listening room, was, that even though, at the speaker, it was perfect BUT...at various places within the listener's room, because of the way you've asked the drivers to reproduce the tone, (at the same time) the interaction of those reproduced frequencies at the face of the speaker, caused, a thing known as Lobing. (pronounced, Looobing, long O.) That lobing is comb filter cancellation. A little distance off the face of the speaker, (comb filter by the way, looks like a comb up down up down, when plotted on a piece of paper). For neophytes, a flat, implying unchanged, line would be considered perfect, with the comb filter, at almost the opposite end of the acceptable spectrum.
Then, the result throughout differing places within the listening room is, that this cancellation, multiplied through time/space, becomes so pronounced, that by the time it reaches YOU ten feet away, (or hits a wall) there are order of magnitude 15 to 30 decibel dips and peaks in the frequency response. Rember that no change in output is considered perfect. If the result I have indicated here is real, then the "fix" is worse than the initial problem. WHEW, again with the boredom.
OK what to do?
What if we could assume, that the lead /lag, between the tweeter and mid, woof, whatever, was soooo small, that humans, can't hear that amount of delay (call that phase shift), We again, should or could, assume, that this particular 'fix' again makes the cure unacceptable?
This very argument is why some designers say, ok, instead of a 6 db, lst order, since this causes variation to happen, "I am going to make it 12db, per octave, so the launch of the sound happens the right way, or even 18 db, per octave, or 24 per octave. 'Screw the perfect "at the speaker" idea. Let's make it best AT the listener's ear. Right? Well, maybe. This is why we have a thousand relatively speaking, good speakers on the market.
Now, enter Sound Labs, (Caveat Emptor) I may become a dealer, so read with careful eyes, "Snake Eys" as Dustin Hoffman said in "Little Big Man". Don't let me Svengali you out here...
The panel of the electrostatic is (nothing in this secton, or any of these sections is perfect so give me lattitude, all you designers out there, this is a layman's description, cringe)
The panel is phase correct, in that the input excites the entire panel, as if it is one driver (which in the larger sense, it is), into making a posigive movement forward, preproducing the sound of the input frequency. So you virtually eliminate the Phase shift.
WOW, GOD THATS PERFECT. "oops, no not really"
Why, you ask?
Well, there's the issue of the back wave (sound output) of the panel.Remember, that the panel is "see through, or I guess "hear through" is more apropos here. This design, open to the back, causes room excitations, (finally some excitement you say?) and lead lag, of the data you receive at your ear. We barely touched on time alignment, but we can do that if you sign up for my midnight course, called, slit your wrists, he's started talking again, 101, at the PT Barnum School of Sound. HA.
Whoa, so what's perfect? The bad news is, Nothing. The good news is, NOTHING.
We pick, and we have a lot of choices.
Panels are inherently correct with regard to this discussion, but with a different set of problems. And, as might be expected, they sound completely different than their dynamic, (woof /tweet) brethren.
Tune in next week; be sure to have your boredom decodor ring on so you don't die of boredom.
Seriously, if you have any questions you can privately email me. If conversation is necessary, I will give you my personal number.
I hope this was helpful. And HEY all you technogeeks, out there, this is an imperfect rendering of information for a 101 study class, so gimme a break, as I put it together n about thirty five seconds.
Best,
Larry R. Staples