Does Time alignment and Phase coherency make for a better loudspeaker?

But the thread is about the time domain performance of a speaker.

One cannot change that too much with DSP. 

With my DSP controlled active speakers I can set them to linear phase or low latency the former is delayed 30ms to align the drivers the later turns off linear phase for a 3ms delay. I can't tell a difference but my room isn't an anechoic chamber. 

Y’all get the crappy lossy cabinet output shows up in the impulse response, and more importantly the waterfall...right ? All easy to hear.

Mijo in 1983 I was using an FFT program on a Commadore PET in the design of loudspeakers.... I feel like Lew... hackles and all

...my statement refers to aiming for a perfectly flat measured speaker response as a sonic end point

But the thread is about the time domain performance of a speaker.

One cannot change that too much with DSP. The impulse response can be cleaned up somewhat. But in the edges of the crossover bands it is not really possible to do much in a passive system compared to an active one.

Once the non-time/phase aligned system and the time/phase align systems are in place, then either/both can have the frequency response dialed into what ever is asked for.

And the question then that the OP sort of asked is, "will they sound different?"

A wise old uncle once told me,  "If you can't dazzle "em with brilliance, baffle  'em wth bulls__t."

The human ear brain processing can differentiate between the initial primary sound and the reflected sounds so long as there is enough time between those sounds. Something that microphones are challenged to do.

So true. I once challenged a dipole mic, which sound came first? Said it could go either way. The condenser mic would only say, "No." The omni of course gave a much longer answer, "Yes." Tried a tube mic, pretended at first not to hear me then started glowing red hot. So I gave up, and haven't challenged one since.

@jjss49 , flat is always the best place to start.

not debating that point (starting point, reference point) - my statement refers to aiming for a perfectly flat measured speaker response as a sonic end point

The human ear brain processing can differentiate between the initial primary sound and the reflected sounds so long as there is enough time between those sounds. Something that microphones are challenged to do. Which is not to say that microphones and the measurements they gather are not important. Just the opposite. The vast majority of manufacturers of time coherent loudspeakers recommend pulling their loudspeakers out from nearby walls for this very reason. Those time coherent loudspeakers that are meant to be placed in, on or near the wall are designed with an understanding of the unique predictability of the affects of this proximity that would not otherwise be available with all the differing variables of users placement and variables of users rooms, and  compensation is built in accordingly.

Mijo - measure and listen, not that difficult

you missed the essential point " rest of chain "

time and phase are just part of the magic since 1977.... 

It's harder than not so easy to have time coherence at the listener ears in a home environment unless your room is an anechoic chamber. It impossible because of reflections and the inconvenience of having two ears. 

@jjss49 , flat is always the best place to start. It is a reference point from which adjustments can be made to suit. You can tell me you like your bass boosted 3 dB. 3dB from where? Frequency response and group delays can be easily measured with a $300 microphone and computer program. What can not be measured is the speaker's radiation pattern but this can be determined exactly from the design of the speaker. It is important to remember that the room, the speaker and where the speaker is in the room have to be considered as a unit. Frequency response curves can change dramatically just by moving the speaker a few feet. Identical speakers will have different frequency response curve depending on where they are in the room. Even in what appears to be a symmetrical room they will measure differently. It seems that only people who have no experience measuring loudspeakers in rooms think good sound is hard to measure. Once you know what type of      soundstage you like and choose the appropriate type of speaker, assuming the speaker is well designed you optimize the system in your room by measurement. You can not get to the most accurate sound without it. You start with accuracy then tune it to sound good to you.  If you have a shrill recording and have to ability to punch in a Grundy Curve (BBC curve), wonderful. I have one on a preset if I need it.

a few points

flat frequency response is easily obtainable in front of speakers but if the idea is to make a loudspeaker sound good, truly flat frequency response is the wrong goal (google ’bbc dip’)

time coherence is only meaningful if delivered at the listener’s ears, which is not so easy in a domestic listening environment

phase response/inversion has rarely made a major difference in the sound in my own experience (many hifi units allow for remote switching of phase, so it can be very easy to test for oneself)

lots of the things about speakers and what makes them sound good are hard to measure, so as in many things in hifi audio, measurement based criteria are necessary but insufficient conditions for a pleasing result

@mijostyn ^that^ is all well put.

@tomic601 , An engineer does not need ears to develop a phase coherent loudspeaker, just a measurement microphone.

In order to be accurate (I did not say sound good) a speaker has to start with a flat frequency response and be time coherent which as unsounds relates assures phase coherence. Unfortunatly, for speaker designers this has to include the room the speakers are set up in. For state of the art accuracy one has to be able to adjust frequency response so that the speakers are flat from 20 Hz to 20 kHz. This also assures that both channels have exactly the same frequency response which is very important for imaging. If a subwoofer is added one has to make sure they are time coherent with the main speakers. The only way all of this can be done is with digital signal processing. Afterwards the frequency response can be adjusted so the system sounds good to the owner. I like mine up 5 or 6 dB at 20 Hz and down 6 dB at 20 kHz. Done this way the difference between speakers is based on the way they radiate into the room and what kind of source they are point vs line.

Generally, speakers with controlled radiation patterns sound better because they cause less room interaction. Horns, dipoles and line arrays come to mind. Omnidirectional speakers require much more room treatment and produce a smaller soundstage like sitting at the back of the hall. This can sound quite natural depending on the recording. 

I learned this past year how much diffuser panels helped with imaging.

So many have spoken out about room treatment, and I was convinced to try some out for myself. I finally manged to complete and install some quadratic diffusers from plans, and I am surely glad I did.

With absorption and diffusion, (and I have a lot of scope to dial it in much better) the presentation is giving me a better understanding of the advice and benefits shared here in Agon

actually phase accuracy in the rest of the chain is difficult and requires an engineer with ears deeply involved in the listening and design process... Charlie at Ayre comes to mind... No global negative feedback.... think thru what NFB is, how it works and what positive and negative aspects it can have.... Of course we owe Dr Otalla a debt of thanks for derailing the THD crowd.... The T in TIM is all about time... ( a quiz of sorts ). I have a variable NFB amp ( RIP Roger, you were a genius )... Assuming level matching, the changes to stereo image across various NFB is apparent.

Carry on ! enjoy the music.

I agree with Millercarbon. Phase and tme alignment is critically important, but when it compromises other aspects, then you will get a less than desirable effect.

BTW, absolute phase and time alignment is IMPOSSIBLE to achive for the entire spectrum of 20Hz-20kHz. Phase correct means only that the crossover regions are not out of phase.

Time alignment also applies only to limited frequency ranges... 

Nevbertheless, when a manufacturer is aware of these parameters, and stirves not to make blatant errors in these areas, it does make for vastly better speakers compared to those where the manufacturer is unaware of these issues.

;

To my ears, I've never heard a better sounding pair of speakers than my Dunlavy SC-Vs. I started out with IIIs, then IVs, and finally my Vs. Once you hear proper time-aligned, phase-coherent speakers, and get used to them, you'll probably never want to go back to anything else.

I'd owned B&W 802 S3s for 18 years before I bought my SC-IIIs in late-summer 2019, and it took a little while for my ears/brain to get used to what I was hearing on the Dunlavy's. Once I got used to the Dunlavy 'sound' I couldn't go back to the B&Ws. The IVs were even better than the IIIs, especially in the bass, and the Vs are an bigger improvement over the IVs, than the IVs were over the IIIs. 

The Dunlavy's are incredibly accurate speakers; but they can be brutally honest on bad recordings. They don't have any colorations, so if you're used to a certain type of sound that isn't incredibly neutral then you may not like them.

 

It seemes to me, that if all else was equal, but one speaker was not time and phase alligned, and the other was, the one that was would sound better.

And I even have a bit of pretty strong evidence that this is the case.

The late Jeff Bagby, well respected speaker designer (professionally, and among DIY’s), designed 2 versions of the same speaker using the great SB Acoustics Sartori drivers.

The Kairos, which is time and phase alligned, and the Adelphos for those that are intimidated by building a cabinet with a slaneted baffle. There are also crossover differences, too, in order to compemsate for the flat vs slanted baffle.

Here are the kits and write ups:

Kairos

Adelphos

My friend and I built a set of both speakers, and compared. No question, the Kairos imaged better and created a bigger soundstage than the Adelphos. Other than those parameters, other aspects sounded, to our ears, identical.

And let me add, as with most well designed DIY speakers, the end result with both the Kairos and the Adelphos, are speakers that end up sounding as good as commercially available speakers at several times the money spent.

^The subject of DSP is perhaps for another thread. But, as pertains to the subject at hand; DSP can be make available time and phase coherence to those that wouldn't otherwise have the considerable chop's necessary to accomplish it, and make it much less time consuming to develop and considerably reduce the labor costs to implement it.

@audition__audio , that is your own bias at play. It is just about numbers, ones and zeros. The very first step in the modern recording process is turning the music into those numbers. Once you are in numbers you can do just about anything without any added distortion. It is all about the programming which has improved over the past 30 years but the basics were well known 30 years ago. 20 years ago some very sophisticated processors were available but the mentality of us audiophiles shunned any added complexity. Our culture was used to the problems of analog devices and I think we generalized those problems to digital devices.

The improvements that can be made with digital processing far outweigh any downside. I digitize my phono stage to run it through a digital preamp/processor. You can go back and forth between the analog turntable and the 192/24 digitized one all day long and you can not hear the difference. The two major advatages of digital processing are being able to adjust frequency response, matching channels precisely and digital bass management with time and phase correction. IMHO you can not get to state of the art sound without them. Another way of looking at it would be, you can make any system sound better with digital processing. Also you can not get to state of the art sound by listening. You have to measure.

Past attempts at DSP were abysmal failures, but perhaps significant gains have been made in the past few years. Funny I still dont consider DSP a viable option which is probably not wise. Just dont like the idea of DSP at a fundamental level. 

^It might be imprudent to overgeneralize conclusions from this research. A loudspeaker that is phase coherent might not be time coherent, though a loudspeaker that is time coherent will be phase coherent. That there was some evidence reported (though deemed staticaly irrelevant)  of discernment of phase coherent recognition, then asks an unanswered question; were the correct responses consistently from consistent respondents?

From years on this forum, if there’s anything that I’ve learned, is that it would appear that while we all hear somewhat similarly,  many of us seem to listen differently. For what ever reasons whether consciously or subconsciously we often times prioritize different things when listening, and are more sensitive to different aspects of sound. I believe that some are more sensitive to time and phase coherence (perhaps at the expense of other attributes) than others. For those that are, I suspect that time coherence allows for a greater appreciation of transient detail, and a quicker neurological processing of soundstage and imaging cues as they are presented in real time.

A proper step response (in an appropriate space, at an appropriate distance) is perhaps the best indication of a time correct loudspeaker. John Dunlavy had said that once a a proper step response is achieved, everything else starts to fall in place. This suggests to me that time coherence could have advantages beyond the specific goal at hand.

Some have argued that time coherence only works for given listeners in specific locations. Adherents would argue that they work just as well as non- time coherent loudspeakers for listeners in less than ideal locations, but non- time coherent loudspeakers can never really work for any listeners at any locations.

In the past, making time coherent loudspeakers was a daunting project, that not many were capable of and/or were not deemed worth the effort. I suspect that as DSP and more integrated loudspeakers become more prevalent, so will time correct loudspeakers become more prevalent.

Every other component in the chain can maintain time cohernence without corruption, why not loudspeakers?

No, these things don't on their own, make for a better speaker. It's not like this hasn't been studied for years. 

https://www.audioholics.com/room-acoustics/human-hearing-phase-distortion-audibility-part-2

Room treatments do not add to coherency.  Room treatments subtract.  The speaker must produce a time or phase coherent sound field in order to produce that astonishing three dimensional soundstage.  The reflections- both airborne and the mechanical interaction between the speakers and the floor (as well as to the amps and other components) though small they may be smear the sound.  The room treatments remove those reflections to restore coherency and detail as well as bring more clarity to the bass.  That is my experience.  

i have modest specimens of 3 time-and-phase-aligned speakers -- one by magnapan, one by thiel, one by vandersteen, and they all image equally well [IOW when fed the right material they all seem to "disappear" leaving just the stereo image] to my ears but each has different requirements for listening room space and placement. the only ones that actually work in a small room, are the thiel and the maggie speakers. the vandy needs LOTS of space, no less than about a 300 sq/ft, in smaller rooms they get shouty at the crossover point [2.8kc] and totally lose their magical stereo imaging. i also have a mirage omnipolar and although i don’t guess they are time/phase-aligned, they image even better than the other 3 in my experience, again only if you give them plenty of space around them even though you can listen to them in the nearfield and get quite a spectacularly stable 3D stereo image that way with a wide sweet spot.

It all depends on how the designer achieves T/P coherency. 

You cannot achieve T/P coherency with room treatments. 

sometimes

I learned this past year how much diffuser panels helped with imaging.  Diffusers on the back wall brought the imaging into much sharper focus.  The next step was putting my speakers on springs.  That not only improved the bass but also increased detail and focused the soundstage even more.  Controlling room reflections has a big impact on imaging.  The springs isolate the speakers from the floor and reduce speaker cabinet ringing.   I saw the difference with accelerometer FFT plots.  As minor as that seems it makes an audible difference.  And remember, the Thiel speakers have a mineral front baffle that is very rigid but still benefitted with isolation.

What does ^that^ have to do with time/phase alignment in a speaker?

It is almost implying that a speaker which is not time/phase correct only need springs and wall treatments.

Or another spruiking of the springs for the Nth time.

Even Levon Helm's ( RIP ) brilliantly damped kick drum has harmonics and reflections that allow the human ear to locate it and ascertain the acoustic reverberant space it is located in. Time and phase information in the waveform that makes it to the ear ... or not..  

@tomic601 , You are looking at it the wrong way tomic. It has nothing to do with musicians on stage. Each instrument is a singularity. Lets take a bass drum. When the pedal hits the drum there is an attack followed by the primary resonant frequency of the drum. Some drummers dampen this which I think is a shame. The attack is at higher frequency and can get caught in the crossover frequency of subwoofers. If the sound of that attack reaches you at different times from the subwoofer and main speakers you first smear the attack and if the delay is long enough you get an echo. Improperly timed subwoofers numb bass transients. Also remember that time shifts phase. Everyone knows what happens if you shift phase 180 degrees. If you don't, wire one of your speakers backwards and listen to what happens. The effect is much less obvious at higher frequencies but it is there. listen to what happens to the image. In order to shift phase 180 degrees at 100 hz you would have to move a main speaker 5 feet forward or backward. This can easily happen with some set ups. Using a swarm system is a way around his problem. 

I learned this past year how much diffuser panels helped with imaging.  Diffusers on the back wall brought the imaging into much sharper focus.  The next step was putting my speakers on springs.  That not only improved the bass but also increased detail and focused the soundstage even more.  Controlling room reflections has a big impact on imaging.  The springs isolate the speakers from the floor and reduce speaker cabinet ringing.   I saw the difference with accelerometer FFT plots.  As minor as that seems it makes an audible difference.  And remember, the Thiel speakers have a mineral front baffle that is very rigid but still benefitted with isolation.  

For Thiel (and Vandersteen) time and phase coherency is their marketing hook but in both cases their designers combined a range of good practices to make their speakers sound great.

He list I heard was: Dunlavy, Quad, Spika, Thiel, Vandersteen

And we pretty much have the majority of posters here talking about their T/P aligned speakers.

https://www.stereophile.com/content/measuring-loudspeakers-part-two-page-3

I well-remember the magic of listening to Dunlavy's in Colorado Springs years ago in The Sound Shop, at a friend's home with IVs, and listening with John Dunlavy and meeting him (a wild guy) in his CS facility. They were very special but a tiny magic zone. For generally superb musical quality but one-person listening, they might get on my list.

As for knowing if I can really hear time and phase coherency, I'm not real sure of my ears unless someone told me what to listen for. I might be able to with a bit of learning.

I do know I am very sensitive to frequency adjustments and much prefer the sound of a room that is bass-corrected and very minimally DSP processed to yield a flat frquency curve with care to time and frequency. A 'repaired' room always sounds cleaner and better defined with no loss of bass intended to be there originally.

I suspect Mr Vegas is a reincarnation of another fellow we all 

know and miss. Not.

I agree with 8th-note.  I have the Thiel CS6 speakers too.   On some recordings I feel like I could get the singer's autograph.  The top Vandersteens are the only other speaker I have heard that does that and sounds a bit better.  All it takes is money.

I have a pair of Thiel CS6 speakers for which time and phase coherence were prime design considerations. This was Jim Thiel's raison d' etre. The CS6 even had the tweeter in the center of the midrange cone to maximize the coherence of the speaker.

As for sound, I went to the 2018 AXPONA and was pleasantly surprised that my speakers and system holds up to pretty much anything I heard that cost less than 6 figures. I think my speakers sound very good for a variety of design reasons but the time and phase coherence have a significant effect.

As mentioned in previous posts there is a tradeoff for any design philosophy and in the case of Thiel speakers it was the need for high current amplification. My speakers have low sensitivity combined with a brutal impedance curve. I run them with a Krell KSA 300S amp that will put out 2400 watts @ 1 ohm and weighs 185 lbs. If you like tube amps you are pretty much SOL.

I've heard lots of good sounding speakers that didn't have time and phase coherency as a primary design goal but I think this characteristic improves imaging to some degree. The imaging in my system is almost scary. On some recordings I feel like I can reach out and touch the singer or particular instruments and the depth of the image is holographic.

Every speaker manufacturer has their "thing" that defines their unique selling proposition. For Thiel (and Vandersteen) time and phase coherency is their marketing hook but in both cases their designers combined a range of good practices to make their speakers sound great.

Ahh, John Dunlavy. What an amazing speaker designer. RIP.  loved selling his SC IV's and V's back in the day. One of the best loudspeakers I've ever heard.

I think the new Grandinote Mach speakers are the best speakers on the market 

I've owned Dunlavy SC-IIIs for about 3 years now and a key feature of their design is time/phase coherency.

I've tried a number of different speakers over the past 3 years but keep coming back to the Dunlavys. Whether or not it's the time/phase feature of their design that appeals to me over everything else, I cannot say.

Speakers which have come and gone in the last 3 years include Quad ESL57s, Merlin Music VSM BME, Sonus Faber Cremona Auditor M, Dali Mentor 6, Klipsch Heresy, Klipsch Khorns, Tyler Acoustic D2, and I'm sure a couple of others that escape me right now.

I just sold the Klipshorns this weekend and the Dunlavys are back in their rightful place.

I think my next adventure will be to try a set of the PureAudioProject speakers with a horn driver. Anyone have any thoughts on the PAP?

Wolf if time and phase don't matter, how is it your ears or the Decca tree  ( substitute in your favorite semi religious cult microphone array choice ) can discern the placement of those musicians on the stage ?

IF the answer is subtle differences in frequency response aka loudness, then matching drivers down to .25 db makes sense...

just my $1.50 , two cents adjusted for inflation...

Mijo like a said 11 bands of Analog eq below 120..... in each speaker Quattro and above, admittedly not everyone's cup of tea, java, pinot, etc.....

Can any of you change the group delay on any part of your loudspeaker system from the listening position? Can any of you change absolute phase from your listening position?

I can and these are my observations. I can not hear any difference at all changing absolute phase. Both directions sound absolutely the same. I should also add that my entire system is balanced. I have Three drivers on each channel. The main speakers are one way ESLs. Then there are the subwoofers, two on each channel. The critical timing is between the subwoofers and the ESLs. I can delay either. Once you get into 5 ms delay or more deterioration in bass impact and definition becomes obvious. What you hear depends on the crossover point and slopes. You can see the group delays with a measurement microphone. The only way you can adjust them is with digital signal processing or moving loudspeakers. This is a problem for subwoofers because they work besy only in certain locations, in corners and up against walls. Being able to put the speakers where they work best and deal with the delays digitally is a large advantage. There are many processors on the market that will do this and some are very reasonably priced. 

Maybe we can get this thread removed too if we get him excited?

Dutch and Dutch 8c are coaxial,  their measurements speak for themselves one of the best on axis and off. Linear phase,  constant directivity from 100hz >. 

Every speaker manufacturer needs a hook, a set of claims to sell speakers. Nothing new. 

A simple experiment can be conducted to testify if the time alignment affects anything audible. Take a pair of speakers that you think have a good image, tilt the front panel back and take a serious listening again to see if the imaging gets thrown off. Conduct the blind test with the help from the 2nd person if necessary to eliminate any gussing game or mental effect. To me, I need to admit I could not hear any difference. However, at the same toeing angle, tilting them back does change the soundstage slightly.

What's the deal with the usage of " time and phase" ? Aren't they the same thing or I'm thinking of the wrong thing being discussed. Anyway with the Dutch and Dutch 8c there is a linear phase mode and a reduced latency mode which turns off the former. Outside of a heavily treated monitoring room or anechoic chamber you won't hear a difference, even then it's very subtle. 

the powered KEF 50 use DSP to correct the passive KEF 50 impulse response...somebody at KEF thought that a worthwhile thing to invest effort into....hmmmmmmmmm....

All the DIRAC like systems doing phase and time correction make for a largely time and phase coherent system.

I suppose DIRAC would not exist, outside of the physics and math useage, if the majority of speakers had nice impulse response right out of the box?

Ever hear a group of unamplified musicians playing spread apart on a stage...anywhere? No phase or time allignment occurs. Hmmm...

Not true...
For instance, say a single French horn player is playing a note that has the fundamental in the woofer, and harmonics in the midrange, and even higher harmonics in the tweeter.

Do you want those harmonics to all be related to each other?

Or do you want the midrange to have the phase flipped?