[PaulFolbrecht]

I am interersted in hearing from speaker designers (& others) on how important they think this characteristic is in a speaker.

The two speakers that impressed me the most at RMAF - the new Thiel 3.7 and the Green Mountain C3 - are t&p aligned and tout this highly.  Both of these speakers had a cohorence I don't think I've experienced before, to that degree (and I didn't know the GMA speakers were t&p aligned until after listening the first time).  You could say I am a convert (maybe).

It seems strange to me that many speaker makes claim it just isn't important.  Is this just glossing over the fact that it's too complex for many to handle competently?

[zybar]

Paul,

I have heard speakers sound good AND bad that are time and phase aligned.

Like everything, I think it comes down to design and implementation.   aa

I didn't hear the new Thiels at the show (previous models weren't my cup of tea), but I definitely enjoyed the GMA's with the Flying Mole amps.  I agree that there was a nice top to bottom coherence with this combo.

I do know one area that phase and time coherence is important is in the bass.  I can change these parameters with my TacT 2.2x and actually hear how being out of phase and time smears the sound and messes things up.

George

[andyr]

Hi Paul,

As zybar has commented, there's a lot more to making a speaker sound good than just time alignment and phase coherence.

As I understand it, in a multi-driver speaker, having the acoustic centre of all the drivers in the same vertical plane means it should be time-aligned ... that might be one reason that Magneplanars sound good!   

Phase coherence is more problematical ... certainly my 3-way Maggie with their 12 or 18dB crossovers and inverted mid-panel are completely stuffed, in this respect!   

You might like to read some articles on "transient perfect crossovers" by John Keskovsky at:
http://www.geocities.com/kreskovs/John1.html

He knows what he's talkin bout!!   

Regards,

Andy

[Daygloworange]

Interesting topic. I've often thought about this as well.
I'd like to hear some of the speaker manufacturers here at AC give us their thoughts on this. I'm real curious as to what their thoughts are on this topic.

Cheers

[EProvenzano]

Interesting topic. I've often thought about this as well.
I'd like to hear some of the speaker manufacturers here at AC give us their thoughts on this. I'm real curious as to what their thoughts are on this topic.

Cheers

I know Dave Ellis touches on this subject in this thread:
http://www.audiocircle.com/index.php?topic=20868.0

I believe he discusses sloped baffle and phase coherence near the middle to end of the thread.

[Duke]

While I'm not a speaker designer in anywhere near the same league as Roy Johnson or Jim Thiel, I'll toss out my $.02.

I think that time & phase coherence is gravy; and getting the basics of good frequency response and low coloration are the meat & potatoes. 

Roy and Jim did their homework and got the basics right - otherwise their efforts on time & phase coherence would probably be superfluous.   I admire their work.

Duke

[andyr]

I think that time & phase coherence is gravy; and getting the basics of good frequency response and low coloration are the meat & potatoes. 

Absoloootely!!   

But that shouldn't stop us from wanting a tasty meal!!    And roast beef without gravy is like ... strawberries without cream, tonic without gin, steak without mustard .... 

And a speaker without an even frequency response or with high colouration is ... like a MacDonalds' hamburger!   

For me, I would dearly love to have speakers which would enable me to easily pick whether the recording was playing back in correct absolute polarity or inverted (so I could then change polarity, to get it as it should be).  Unfortunately, as I said above, much as I love my 3-way Maggies, that's impossible!   

Regards,

Andy

[NewBuyer]

Is speaker design the only system element that impacts ultimate time/phase frequency alignment? I wonder if system electronics or even speaker/ic cable material (silver, copper, etc) may also have an impact on this aspect...

[JLM]

As much as they get slammed, single driver speakers are inherently time and phase aligned.  You just can't beat their coherency.

Physical (vertical positioning) of multiple drivers can only help time alignment in a single horizontal plane that is located somewhere between the driver mounting heights.

Phase alignment must be carried through all power amplifiers, crossovers, and physical positioning of the drivers.  One advantage of Linkwitz's Orion is the use of a 6 channel power amp, so the signal going to each driver is treated the same.

Even with all this properly done, the sound at crossover frequencies coming from two different drivers (brands, types, materials, etc.) is bound to sound different. 

After living with single driver speakers, others sound like a jumbled mess.  And don't even try to listen to "them" nearfield.  Midrange (80 - 5,000 Hz) is supremely important.  Fancy designs can't replace having a single high quality driver covering those 6 octaves.  Around here, besides Ed Schilling and Louis Chochos, big B does it best with his 166 - 10,000 Hz single midrange ribbon based VMPS designs.

IMO this is the biggest, most glaring flaw of nearly all speaker designs (not using a quality extended range driver to cover the mid-range).  All the rest (while important to me as an audiophile) is secondary. 

[andyr]

As much as they get slammed, single driver speakers are inherently time and phase aligned.  You just can't beat their coherency.

Absof'inglutely!!  No crossover is the only way to go!!    Trouble is ... single-driver speakers can't deliver 20-30KHz.  So what d'ya do??   

Phase alignment must be carried through all power amplifiers, crossovers, and physical positioning of the drivers.  One advantage of Linkwitz's Orion is the use of a 6 channel power amp, so the signal going to each driver is treated the same.

I'm not sure I follow you about SL's recommended (but very average quality, sonic-wise) 6-channel power amp??  It's merely a useful "package" ... 6 channels and not relatively expensive.

I suggest SL Orions would sound better if they were driven by 6 high-quality monoblocks ... but, yes, from the same mfr, so they have a similar sonic imprint.

Regards,

Andy

[Russell Dawkins]

I wonder if system electronics

barely possible, I think, if very different amps are used for different drivers

"or even speaker/ic cable material (silver, copper, etc) may also have an impact on this aspect..."
I very seriously doubt that.

[GHM]

Yes I also agree there's something to the T&PA thing..just not sure why completely? My progression through several pairs of speakers leaving out a few. Were Maggies, GMA's back to Maggies. Once I got back to Maggies I heard a lack of what others describe here with T&PA speakers..although the sound was more open with the Maggies. As much as I loved the Maggies and still do,the GMA's ruined me in the way they produce music. From there.. I went to SDs and found that certain quality I heard in the GMA's. I fault the GMA's for taking me on this path, since they were the first T&PA speakers I've had at home. They are still one of my all time favorite speakers. I kick myself everyday for selling them!

We all listen differently and look for different qualities to match our taste. I've always thought the T&PA thing was a bunch of BS until I learned how to listen for it...There ain't no going back now! As mentioned before..I agree there's more to the speaker than T&PA but it sure doesn't hurt to have it at the dinner table.

This idea of T&PA along with no crossovers in the important part of the midrange was re-enforced recently after auditioning a big name brand (that shall remain nameless)... it sounded like a Louisiana Gumbo!!

[dwk]

I've always been intrigued by 'T&PA' (has this thread coined a new acronym? Haven't seen that one before) as well, and it's been a driving factor behind my interest in DSP xovers, coax drivers, and most recently in looking at Danley's Unity design.

I certainly agree with the idea that it's 'gravy' on top of a well designed speaker, but I also expect that there are only certain frequency ranges where phase is truly audible. Plus, sensitivity probably varies from person to person as well, even leaving aside the training question. Makes for a landscape that is tricky to navigate.

The obvious problem with 'T&PA' speakers implemented conventionally is that you are making some significant compromises in many other areas to achieve the 'T&PA' characteristics. Both first-order acoustic designs and single-driver designs have some serious weaknesses when compared to 'roughly equivalent' multi-way speakers. Sometimes this is OK, sometimes not.

In my case, since I'm a ostensibly DIYer I'm drawn to designs that are interesting to experiment with as well as listen to. For example single-driver speakers just don't really interest me; this is somewhat surprising since I'm largely a 'folkie' and so much of my music material probably is reasonably well suited to that approach.  Also, while I've heard some good first-order designs (I'd count Dunlavy and Theil in that group), others leave me nonplussed (the GMA's I heard at RMAF were HORRIBLE in my opinion, although show conditions could have been part of that). So, what I seem to find interesting is unconventional approaches that attempt to capture what is good about T&PA while mitigating the problems - in my view the big ones are dynamic cabability and power response.

So, recently I've been really interested in Danleys Unity designs - IMHO these seem to be getting close to the 'holy grail' of speaker design - very high output capability, very low distortion, uniform power response and phase coherence (although only his recent SH-50 is truly linear phase and capable of passing a square wave over a wide bandwidth).  The SH-50 is way out of my price range, but I just took delivery of a pair of Yorkville U15's. Being PA cabinets they may be a little rough out-of-the-box, but they hold the promise of being phase-coherent from 300 on up, and with a subtractive-delay DSP xover to the woofer should be pretty good over the audible band. (that's the plan, anyway. stay tuned)

[PaulHilgeman]

I'll put a few cents in on this one

The first and most important thing to remember with time and phase aligned multi-driver arrays is that they are made to sum properly at a certain point(s).

If this T&PA speaker is a 2-way design, there will be a single plane that slices through the speaker at which any point on this plane will be equidistant from both drivers, this is the plane of proper summation.

If this T&PA speaker is a 3-way design, there will be a single ring of points around the speaker that is a certain distance from the speaker.  This only holds true if the acoustic centers of each driver are in a single line.  If they are not in a single line, i.e. the midrange is farther back or forward, there will be a singular point where the output from all drivers sum properly.

All of this holds true only if perfect first order crossovers are employed, extending at a minimum of one octave past the crossover point, preferably two octaves.  There are other ways to do this will 2nd order slopes and filler drivers as well, but John K's site is best for this: www.musicanddesign.com

Now, if you are not listening on this plane, on the circle, or at the point of proper summation, you will have the compounded effects of non-proper summation: 1) frequency response aberrations due to different arrival times at and around the crossover point, and 2) non perfect T&PA.

Also, every reflection off of every surface will have this effect applied to it as well.  The two most prevalent things that color the sound reproduced are the mirror image of the speaker below the floor surface and the mirror image off of the ceiling.  Lets take the floor for example.  In a 3-way system with a woofer near the floor, a midrange centered maybe 36 inches up and a tweeter centered 42 inches up, the initial arrival times at the listener can be all perfectly aligned with first order slopes, but the reflection off of the floor will be totally out of whack.  The path length for the woofers reflection will be very short, the midrange longer, and the tweeter longest.  The same goes for the ceiling reflection, but in reverse order.  The reflections will all have the colorations of the off-axis response of the driver then filtered by the reflective properties of the surface they reflect off of.

Listening to this speaker what you will hear will be: (Bar |  denotes time delay, brackets [] denote frequency response abbaration due to floor reflection and off-axis response, braces{} denote frequency response abbaration due to ceiling reflection and off-axis response)

Woofer Midrange Tweeter | [woofer] | [Midrange] | [Tweeter] | {Tweeter} | {Midrange} | {Woofer}

You get the idea....

It is my opinion, that a much more balanced, coherent, sound can be had when designing with much more attention paid to off-axis response smoothness, lack of cancellations in the off-axis, and having the same time coherence at off-axis angles.

The Ronin 1.0c does not use first order crossovers, so phase coherency is not possible, however, the disc, ring, and point analogy above can be expanded to almost any point that is less than about 70 degrees off axis with regard to proper arrival times for proper summation in the crossover region. 

This also means that all reflections arrive at your ears properly, and not skewed in the time domain like other phase and time coherent speakers.

The Ronin 1.0d with its linear phase digital crossovers is time and phase coherent on and off axis with the same foundations for even power response in the Ronin 1.0c.

Odds are pretty good that if you have really liked a T&PA speaker before it was just a plain old good crossover design, with good drivers, good cabinet etc.  The only time the T&PA really holds true when you aren't in an anechoic chamber is with coaxial drivers combined with first order slopes or linear phase digital crossovers.

Hope this helps and properly conveys my opinion on the subject since it originally asked for manufacturer input.

Regards,
Paul Hilgeman
Nomad Audio

P.S. The Gravy thing - Well Said 

[Midnite Mick]

I have been following this thread as I am very interested.  I am just getting into DIY myself (and loving it), but am not knowledgable enough to comment.  However, I am wondering where are all our local speaker designers with their comments?

Mike

edit: above was posted just before this

[JLM]

Andyr,

My (and GHM's) single driver speakers are rated in room to be flat 30 - 20,000 Hz (89 dB/w/m, 8 ohm, 27 watts continous/80 watts peak).  That's plenty good enough for me.  BTW the link is:

http://www.geocities.com/rbrines1/

But many other single driver options exist where a powered subwoofer is added to provide foundational bass.  Designs also exist where the extended range driver (usually promoted as a full range driver) is used to cover a wide midrange with woofer and tweeter added.  These have advantages, but there is no free lunch and no perfect speaker.  The mighty Fostex F200A driver has allowed me to travel this purist path.

[Occam]

Paul,

Thank you for that very excellent post. It might well explain my long term affair with my Alons - an attempt to align the acoustic centers of the drivers on a vertical axis, as well as first order crossovers morphing to 2nd order an octave out. Your own approach share many similarities with (and hopefully improvements on) the Alons dipole mid speakers.

You explanation also gives credence to the appeal of Linkwitz's Pluto, Olsher's CBAE and Danny's GR Research 'O' (omnidirectional) speakers. They all attempt a vertical alignment of the driver's acoustical centers. In the case of Olsher's and Danny's (for the 2 of his 2 1/2 way) the right angle mounting of mid and tweeter moves the acoustical centers closer laterally as well, though not as close theoretically as your own coincident mid/tweet.

[BobRex]

Paul, what you say is logical to a point, and sounds like the same criticism every designer that does not try to T&PA a speaker gives.  And yet...  The tone that I always read into this is "I/We know more than those idiots.  Obviously they've either ignored the realities of speakers/room interactions or they are working in a vacuum."  And yet, Theil, Dunlavy, Vandersteen, even our own Big B. seem to be very successful flying in the face of profound logic.  Perhaps they are the ones who are on to something.   

Now let's look at some of your arguments:
"Now, if you are not listening on this plane, on the circle, or at the point of proper summation, you will have the compounded effects of non-proper summation: 1) frequency response aberrations due to different arrival times at and around the crossover point, and 2) non perfect T&PA."  - So how does this differ from every other speaker out there?  Seems to me that years of Sphile measurements show that all non T&PA speakers start out with these 2 points.  Or am I missing something in the impulse responses?

"Also, every reflection off of every surface will have this effect applied to it as well." - Umm, once again, other speakers don't have this problem?

"This also means that all reflections arrive at your ears properly, and not skewed in the time domain like other phase and time coherent speakers."  No they don't.  You can't control room reflections any better than anyone else.  You can't predict placement, so you cannot possibly predict how reflections behave relative to the intial source.  Since reflective surfaces can be absorptive at selective frequencies, you will not always get a perfect signal reflected.  This will, to a degree, skew the frequency domain and also affect the time domain.

"Odds are pretty good that if you have really liked a T&PA speaker before it was just a plain old good crossover design, with good drivers, good cabinet etc.  The only time the T&PA really holds true when you aren't in an anechoic chamber is with coaxial drivers combined with first order slopes or linear phase digital crossovers."  Pretty condescending don't you think? 

Maybe, just maybe, T&PA is the final spice to a good creation?  The point that takes a speaker from pretty good to magical?

[PaulHilgeman]

- So how does this differ from every other speaker out there?  Seems to me that years of Sphile measurements show that all non T&PA speakers start out with these 2 points.  Or am I missing something in the impulse responses?

It doesn't.  It is a 'feature' of all speakers that do not have coincident drivers.  By being not at the design distances from each driver 1 and 2 are the results.  The only way to get around this is by making the frequencies come from the same location.

- Umm, once again, other speakers don't have this problem?

Yes, See above.

No they don't.  You can't control room reflections any better than anyone else.  You can't predict placement, so you cannot possibly predict how reflections behave relative to the intial source.  Since reflective surfaces can be absorptive at selective frequencies, you will not always get a perfect signal reflected.  This will, to a degree, skew the frequency domain and also affect the time domain.

Yes they do.  They are skewed in the time domain with regard to the on-axis sound, but not from driver to driver as they are in non-coincident designs.  In non-coincident designs, the reflection off of nearby surfaces will have a much greater difference in impulse response than a coincident design with good power resoponse.

I can not control room reflections, but I can ensure that the time domain behavior of these reflections is very similar to the time domain behavior of the on-axis response.  With a tweeter above mid design, the floor reflection will have the mid impulse arriving prior to the tweeter reflection.  Vice Versa for the ceiling reflection.

Analyze the last set of plots in the following article:

http://www.nomad-audio.com/article_1.htm

These are taken at 50 degrees above the speaker, notice how late the midrange arrival is?  Now compare to 50 degrees above axis on the Ronin, shockingly similar to the on-axis impulse response.

"Odds are pretty good that if you have really liked a T&PA speaker before it was just a plain old good crossover design, with good drivers, good cabinet etc.  The only time the T&PA really holds true when you aren't in an anechoic chamber is with coaxial drivers combined with first order slopes or linear phase digital crossovers." 

Pretty condescending don't you think?

I didn't mean it it to be.  It was meant to bring light to the fact that there are T&PA speakers out there that sound good and there are ones that sound bad (to my ears and others') and that it is not a pre-requisite for and good design.  T&PA of multi-driver, non coincident speakers in normal listening rooms needs to be looked at from more than just the on-axis, anechoic response point of view. 

I do not feel that "I know more than these idiots"  I just presented my researched, thought-out and well formed opinions on the subject matter.   

We are all entitled to our own thoughts and opinions, but others respond better to and understand our thoughts when presented with rationale, even if the rationale is simply "I thought it sounded better because I heard...", not pure negativity.  What are yours with regard to this topic?

Regards,
Paul Hilgeman

[PaulFolbrecht]

Wow, this is great stuff, guys.  Glad this thread got going.

FWIW (ignoramus speaking here), I certainly wouldn't consider T&PA much more than "gravy", either.  I own Zu Druids, used in my 2nd system.  They are good speakers, but far from the best I've heard, and there are definitely non-T&PA speakers that better them (I'd sure hope speakers at 10x their price would better them!).

Like I said, I was stunned by both the Thiel 3.7s and the GMAs at the show.  I would also say, however, that both the WATT/Puppy 8s and the Wilson Beneech speakers in that room (with deHavilland) stunned me as well, but did not have quite the same level of 'coherence'.  (I am very surprised by the guy who thought the GMA room stunk.  I believe it was called one of the best rooms at the show by one or two webzines already.)

I really like my Hyperion 938s.  In fact, I love them, and really doubt I'll buy new speakers.  Shall we say - I 'hope' I don't!