Phase coherence matters!

AndrewH · « **on:** 24 Mar 2005, 10:30 pm »

Some people argue that you can't hear it. Some people say it's just marketing.

I say, darn it, you *can* hear relative phase coherence in loudspeakers!

I even created a website dedicated to the phenomenon. Check it out:

http://phasecoherent.com

The aim is to help educate audiophiles on the web about phase coherence, and serve as a shopping resource for phase coherent loudspeakers.

Any comments or questions appreciated.

-Andrew

warnerwh · « **Reply #1 on:** 24 Mar 2005, 11:04 pm »

You're right about that. I've found I like phase coherent speakers. It's interesting that some of the best sounding speakers are listed on your site. You should also list VMPS(Veritone Minimum Phase Speakers). VMPS speakers are of the best values in high end audio that exist. Thanks for the info.

AndrewH · « **Reply #2 on:** 24 Mar 2005, 11:23 pm »

Thanks, I'll look into VMPS. I've come across the brand before but don't recall any explicit description of their crossover design. The flat baffle (lack of aligning of acoustic centers) doesn't seem indicative of a commitment to phase coherence. Maybe the wide range ribbon makes it unnecessary.

NealH · « **Reply #3 on:** 24 Mar 2005, 11:27 pm »

There is phase coherence, and there is phase and time coherence. The Vandersteen, Thiel and Meadowlarks are phase and time coherent. Feed them a step ramp and a microphone will pick it up looking very similar to the original signal. With the Magnepan, Soundlab and others there will be a pre-blip then the stepped ramp. This indicates the drivers are connected in proper polarity but, not perfectly phase and time coherent.

The other area that better speakers address is the wavefront polar pattern. It is best to select a crossover point where the lower driver's pattern has narrowed to the point that equals the upper driver's pattern. This should be the selected crossover point. Speakers that employ this idealogy are Meyer, Geddes, SP Technologies, Genelec, and perhaps a couple other pro-sound monitors.

_scotty_ · « **Reply #4 on:** 24 Mar 2005, 11:42 pm »

This is a rather simple parameter to define. The speaker either reproduces a 1kHz squarewave under anechoic conditions or it doesn't. If no actual evidence of this ability is available from a manufacturer or an independent test facility, than any manufacturer claiming to have a true phase coherent, transient perfect loudspeaker should be suspect.
While this is an important parameter,it by itself will not guarantee accurate reproduction of musical waveforms.

Read the information about time aligned speakers here.
http://www.geocities.com/kreskovs/TimeAligned1.html
http://www.geocities.com/kreskovs/TimeAligned2.html

Scotty

AndrewH · « **Reply #5 on:** 25 Mar 2005, 12:27 am »

rnhood - Wow, I didn't know that about Magnepan (and SoundLab). I assumed their panel drivers plus first order crossovers would have to be time and phase coherent. That kind of throws a wrench in my design. Should I only include the perfectly time/phase coherent, or should I have some kind of graded system.

Is there any source you can cite me that demonstrates Magnepan's less-than-perfect step response?

AndrewH · « **Reply #6 on:** 25 Mar 2005, 11:18 pm »

Stereophile has step response graphs for Magnepan, and they don't look very good. I've taken the maggie pages offline for the moment. (Sound labs probably should go too).

Scotty - are you saying a 1 kHz square wave under anechoic conditions is a perfect indicator of phase coherence? Is it both necessary and sufficient for phase coherence?

If so that would be a nice objective metric to apply to all the speakers I'm putatively recommending. I could show their square wave response, some derived metric, and present the rankings to the readers.

Thanks.

_scotty_ · « **Reply #7 on:** 26 Mar 2005, 05:11 am »

Andrew, The squarewave test defines phase coherency.
Remember that amplitude error and phase error are equivalent for the purpose of this discussion. An impulse test will not show amplitude error
and neither will a step function. A speaker can pass these tests and still fail
to reproduce a 1kHz squarewave. As you are well aware terminology relating to phase accuracy or time alignment has been used for quite some time as marketing buzzwords and the precise technical definition has been obscured as a result. In order to reproduce a squarewave properly no compromise in design or execution is allowable if this goal is to be met.
In recent times only the loudspeakers from John Dunlavey have met this criteria that I am aware of. The Thiel 3.6 has a claimed phase shift of +or- 10degrees, this is low phase shift but allows for a total deviation of 20degrees which would not look good on an oscilloscope screen shot displaying its squarewave performance. I wish you good luck with your website but I believe that you may find it almost impossible to wade through the marketing hyperbole regarding this subject to arrive at a list of loudspeaker manufacturers who actually build speakers that can pass a squarewave
test. If you relax your standards to include manufacturers who make unverifiable claims of "low phase error" without the testing results to substantiate their claim you have wasted your time publishing meaningless
information on your website. Good luck getting at the truth.
Scotty

LordCloud · « **Reply #8 on:** 29 Mar 2005, 05:28 pm »

A speaker being time and ohase coherent is the first thing I look for in a speaker. All you need to do is hear a properly designed and engineered time and phase coherent speaker PROPERLY set up with carefully chosen equipment and you'll never go back. Most speakers are nowhere near set up to show off their capabilities, a time and phase coherent speaker is even more difficult to set up properly. It has taken me quite some time to get the best out of my speakers, but when it's done properly it is amazing!

jeffreybehr · « **Reply #9 on:** 29 Mar 2005, 05:49 pm »

I agree heartily with everyone raving about the wonderful sound of a phase- and time-coherent speaker. I thought my Kindel PLS-As sounded good until I heard a pair of Quad 989s a couple times. I bought a pair last fall and have never heard better sound in my system.

The goodness of their sound is indeed difficult to describe--I talk about their 'togetherness', their 'wholeness'--but my descriptions never do justice to the quality of the reproduced music.

I believe (but could be wrong) that the Quad 63s, 988s, and 989s are indeed phase- and time-coherent; they are single-diaghram, point-source, crossoverless designs.

_scotty_ · « **Reply #10 on:** 29 Mar 2005, 11:21 pm »

LordCloud, Did Meadowlark supply measurement data with your loudspeakers substantiating their phasecoherent performance.
No such data in the form of actual measurements is available on their website. Screen shots available online show the Shearwater to have a
an excellent step function. The screen shot from the stereophile review of the Meadowlark Swift shows a minor misalignment which may be a error in microphone positioning for the measurement. I have been unable to find any screen shots of squarewaves reproduced by Meadowlark loudspeakers available online. Scotty

JoshK · « **Reply #11 on:** 30 Mar 2005, 12:55 am »

Where should the mic be positioned in order to avoid such errors? Does this suggest that the phase alignment is meant for a single point in space?

rosconey · « **Reply #12 on:** 30 Mar 2005, 01:02 am »

i'm barely coherent

and have many phases

Davey · « **Reply #13 on:** 30 Mar 2005, 01:53 am »

Quote from: JoshK

Where should the mic be positioned in order to avoid such errors? Does this suggest that the phase alignment is meant for a single point in space?

Josh,

Yes, just a single point in space for speakers like the Swift. In my view these are not time-coherent speakers.

I think a person could make a pretty good argument that there aren't any speakers other than single-driver full-range or maybe some coaxial drivers with coincident tweeters that really satisfy this criteria. And then only if other factors are adhered to as well. Maybe the Dunlavy speakers because of their symetrical layout. Nah, not even those really.

However, I did get a laugh out of the "discussion" of time coherence on the Meadowlark site. I didn't learn anything.....except about "how the nature connects the stars to my soul." I feel like a beer.

Cheers,

Davey.

_scotty_ · « **Reply #14 on:** 30 Mar 2005, 02:54 am »

Davey, Dunlaveys qualify as phase coherent, they will reproduce a 1kHz squarewave. This is the very definition of phase coherent. If a speaker cannot reproduce a squarewave than it is not phase coherent. A speaker
can have a good looking step function and still have large variations in its frequency response. A single driver can still exhibit significant mechanical,and electrical phase shift.
The Ohm F might be the only phase coherent single driver speaker speaker that is phase coherent in the horizontal plane through 360 degrees as a byproduct of its radiation pattern.
Josh, Stereophile made an assumption about where to place the microphone
to measure the Swift, a very minor mislocation of the mic could produce
the observed step function shown in the review. If they had been concerned
about their result they could have tried relocating the mic and measuring again.
I will have to agree with Davey and say that there may be no speaker currently manufactured for the audiophile home consumer that is phase coherent
at this time. Scotty

John Ashman · « **Reply #15 on:** 30 Mar 2005, 03:25 am »

I've found that phase coherent speakers often give up so much that they benefits are obviously outweighed by other downsides - higher distortion, limited dynamics, poor off axis dispersion and/or other issues including very complex and parts heavy crossover to get it right. Digital speakers and DSP such as DEQX solve this, so it will probably only be a matter of time before analog phase correct are "phased" out for digital versions because so many downides of 1st order crossover design are reduced or eliminated. Of course, I'm biased here, but I don't think it makes sense to champion analog phase correct speakers without noting the obvious downsides to such a design. I've heard lots of Vandersteens, Thiels, Meadowlarks, etc and as much as I admire the engineering, I've never liked the sound much and as even proponents will tell you, they take a lot of energy to sound their best. But digital designs can easily do phase/time alignment without any downsides, so eventually we'll be able to do a big cumbaya and time/phase alighnment will be a given, not something that is only achieved through extraordinary measures.

Davey · « **Reply #16 on:** 30 Mar 2005, 03:42 am »

Scotty,

Yeah, I'm with ya on the Dunlavy's. However, what I was thinking about was moving the microphone vertically away from the design axis which is on the tweeter. I don't think you'd see very good square-wave reproduction in that case.
I guess if we accepted the definition that a system can be "time coherent" even if it only exhibits this at a single point in space then more speaker systems would qualify. (And yes John, some DEQX'd ones.) But if we were really being critical on this then I'd agree with Scotty.

Anyway, I don't think the step-response measurements used by Stereophile are very illustrative of the time coherency of a speaker system anyway because the time window is too short for a legitimate full range speaker and the listening system between our ears does such a great job of integrating time differences like these.

I'd like to see a lot less marketing-speak and BS on this subject from manufacturers/dealers.

Cheers,

Davey.

John Ashman · « **Reply #17 on:** 30 Mar 2005, 03:48 am »

Davey,
An interesting note was that Jack from NHT said that time and overall phase is so subjectively hard to pick up that the tweeter/midrange was time aligned (from memory) "just because there was no reason not to do it". Kevin Voecks and many other excellent speaker designers have simply been unable to verify that phase/time is very audible, if at all, and Voecks has done some serious looking into it. But, like I said, digital makes it easy, so why not? Driver/cone distortion, dispersion and other factors not done well by 1st order systems are a *lot* easier to notice. Phase/time is #5 out of 5 major things that define a speaker's performance, distortion, dispersion, cabinet/acoustic distortion and FR accuracy being the other major ones in no particular order.

suits_me · « **Reply #18 on:** 30 Mar 2005, 05:36 am »

I know Revel does not push time and phase coherence, nor did Snell, but Voecks' Symdex did way back when. He seemed pretty sure then....

I wish we could get Roy Johnson in here, but, alas, since he decided to get an investor and make some money he hasn't been posting at audiogon.

JoshK · « **Reply #19 on:** 30 Mar 2005, 02:31 pm »

Scotty/Davey,

Thanks for confirming. I was pretty sure it was a single point. That is why I never understood sloped baffles. Simple geometry would tell you that that will only work for a single point, which is all well and good, but a different claim then often made by the manufacturers.

Truth be told, if one is evaluating critically a speaker design then they should be in the listening sweet spot and not walking about the room. But if the speaker does its thing for a spot 8.8' away from the speakers and another listener happens to have his listening spot 9.2' away, what then?

I guess that is why digital correction is such a cool thing, because it allows flexibility in the design to match the room and setup of the user, provided the user knows what they are doing.