A history of Time Aligned Loudspeakers

[robh36062]

At least, its good for business! But mildly irritating from a systems engineering point of view.

Well some things will never change and if something gets settled something new is sure to upset the cart. PWK stirring the pot!

Rob

 

[kimmosto]

Both these were quite audible to me in ABX using music, so any focus on time alignment should be in the LF, and as you say should be below 0.45 cycles or even below 0.25 to ensure below audibility. AVRs should have DSP-based filters...

Did you listen with speakers or headphones? I don't have suitable loudspeaker set at the moments so ears only with headphones was the only option yesterday to test audibility thresholds. Testing with headphones and speakers have some fundamental differences. Experiences with speakers have indicated for a long time that headphones are not adequate for testing transient energy/pressure losses, and possible effects to soundstage and sound brightening with multi-ways.

 

[Thomas_A]

Did you listen with speakers or headphones? I don't have suitable loudspeaker set at the moments so ears only with headphones was the only option yesterday to test audibility thresholds. Testing with headphones and speakers have some fundamental differences. Experiences with speakers have indicated for a long time that headphones are not adequate for testing transient energy/pressure losses, and possible effects to soundstage and sound brightening with multi-ways.

I've done both, but the specific experiment in that linked thread was with headphones. My easiest way to hear phase/GD is shifts in bass region below 300 Hz and asymmetrical signals, e.g. bass guitars etc, which cause timbral changes (same results with speakers and the headphones I have).

 

[gnarly]

Good job! Perfect reconstruction!! But unfortunately not applicable in a passive crossover.

True that. A passive's impulse and step falls apart past first order.
Tis quite the shame passive is hamstrung such...

But pretty easy active....Here's the acoustic measurements of a 3-way active main.

From this horn speaker shown with tractix secondary flares on right, and with simple mouth round over on primary horn section on left..

But frankly, I haven't bothered with looking at step function for maybe 6 years.
Staightforward Mag and phase curves are better to analyze and tune with ime.
Get them right and you know step and impulse are right too. Here's the mag and phase that goes with the step and impulse from above.

IMO, Flat mag and phase is all we really need to know / work for........the rest is just audio techno-fluff-jargon.

 

[sergeauckland]

May I be allowed to relate an anecdote?

Back in the mid 1980s, when I had a HiFi shop, (spectacularly unsuccessful). At the time, there was a lot of stuff in the magazines about time alignment, so having plenty of time on our hands, I decided to experiment a bit. I removed the tweeter from a loudspeaker (don't ask me which one it was over 40 years ago), blocked up the hole in the baffle, and put the tweeter on top on a simple frame so it could be moved back and forth. Moving it some 15cm either way from the 'normal' position made absolutely no difference to the sound, as heard by those of us in the shop at the time.

I can imagine that if the tweeter had been a kilometer behind the woofer, then (disregarding the attenuation) the delay would be noticeable, but a +-15cm (6") variation didn't seem to make any difference. From that I concluded that time alignment within reason was a non-issue, and have disregarded it since then.

S

 

[Thomas_A]

May I be allowed to relate an anecdote?

Back in the mid 1980s, when I had a HiFi shop, (spectacularly unsuccessful). At the time, there was a lot of stuff in the magazines about time alignment, so having plenty of time on our hands, I decided to experiment a bit. I removed the tweeter from a loudspeaker (don't ask me which one it was over 40 years ago), blocked up the hole in the baffle, and put the tweeter on top on a simple frame so it could be moved back and forth. Moving it some 15cm either way from the 'normal' position made absolutely no difference to the sound, as heard by those of us in the shop at the time.

I can imagine that if the tweeter had been a kilometer behind the woofer, then (disregarding the attenuation) the delay would be noticeable, but a +-15cm (6") variation didn't seem to make any difference. From that I concluded that time alignment within reason was a non-issue, and have disregarded it since then.

S

I have not bothered either, but I have a distant plan to skip my 100 Hz AVR crossover for my two subs acting as speaker stands and get something better. I could even raise it to 300 Hz to get some more headroom for my tiny 5.5 inch mid woofers.

For the mid to tweeter, I will not go further, I down-sized my equipment long time ago from active to passive crossover.

 

[Trifonov Audio]

True that. A passive's impulse and step falls apart past first order.
Tis quite the shame passive is hamstrung such...

But pretty easy active....Here's the acoustic measurements of a 3-way active main.
View attachment 534788
From this horn speaker shown with tractix secondary flares on right, and with simple mouth round over on primary horn section on left..
View attachment 534790



But frankly, I haven't bothered with looking at step function for maybe 6 years.
Staightforward Mag and phase curves are better to analyze and tune with ime.
Get them right and you know step and impulse are right too. Here's the mag and phase that goes with the step and impulse from above.

IMO, Flat mag and phase is all we really need to know / work for........the rest is just audio techno-fluff-jargon.

View attachment 534799

Very good, almost perfect for an active system! When the project is "under the hat" of a subtractive crossover or DSP processing - then phase-time problems disappear! You can also see large horns - which is a perfect solution! What are the drivers called and where are the slope frequencies?

 

[Trifonov Audio]

May I be allowed to relate an anecdote?

Back in the mid 1980s, when I had a HiFi shop, (spectacularly unsuccessful). At the time, there was a lot of stuff in the magazines about time alignment, so having plenty of time on our hands, I decided to experiment a bit. I removed the tweeter from a loudspeaker (don't ask me which one it was over 40 years ago), blocked up the hole in the baffle, and put the tweeter on top on a simple frame so it could be moved back and forth. Moving it some 15cm either way from the 'normal' position made absolutely no difference to the sound, as heard by those of us in the shop at the time.

I can imagine that if the tweeter had been a kilometer behind the woofer, then (disregarding the attenuation) the delay would be noticeable, but a +-15cm (6") variation didn't seem to make any difference. From that I concluded that time alignment within reason was a non-issue, and have disregarded it since then.

S

Very interesting and "almost" inexplicable anecdote! We had similar cases years ago. I share the then answer (comment) from me and other like-minded people.
1) Drivers, components (for filter) and the whole speaker -without proper quality and potential for improvement.
2) In principle - the result of proper phase-time optimization is always is complex from a Special crossover together with Accurate Shifting of the upper driver back!
What - you have "experimented" (I apologize!), is wrong and incorrect.......without a preliminary special project design(Tеchnology), simulations and measurements!!

 

[LSPhil]

Very interesting and "almost" inexplicable anecdote! We had similar cases years ago. I share the then answer (comment) from me and other like-minded people.
1) Drivers, components (for filter) and the whole speaker -without proper quality and potential for improvement.
2) In principle - the result of proper phase-time optimization is always is complex from a Special crossover together with Accurate Shifting of the upper driver back!
What - you have "experimented" (I apologize!), is wrong and incorrect.......without a preliminary special project design(Tеchnology), simulations and measurements!!

Honestly, it's a bit unfair to question someone's practical experiments in such a cynical way, especially when they actually checked something with their own ears instead of just repeating theory from simulations. I did a very similar experiment myself, shifting a tweeter relative to a woofer, and I had practically the same observations.

I didn't notice any major tonal changes with shifts of a few to several centimeters. Only in one specific alignment did I get better spatial coherence, that feeling that the sound 'detaches' from the speakers and comes from a single point. But those were subtle differences, not suddenly a 'completely different speaker'.

And that's why I don't really understand what 'special simulations' the commenter is referring to. Of course, you can design crossovers for specific offsets of the drivers' acoustic centers and optimize phase around the crossover region, that's standard engineering. But claiming that any listening without such 'technology' is by definition flawed sounds a bit like audiophile dogmatism.

At most, it's known that larger offsets cause interference and change the directional characteristics. But those effects are often much less audible than internet discussions would suggest. Especially in a normal listening room, with reflections and the listener's head movement.

So in my opinion, the original poster did a perfectly sensible practical experiment, and his conclusions are not 'wrong' at all. They just didn't confirm a theory that someone really wants to believe in.

 

[Canuck57]

Hummm, what do you think about the DCM Time Windows approach?
Peter had fun with these.

https://akdatabase.com/AKview/albums/userpics/10007/TW_Audio_Critic_Reviews.pdf

A big deal in the 70s

I bought a used pair of those exact speakers (if memory serves me right) and thought they sounded quite good. How they would measure is another story. Would be interesting if Amir could find a pair to measure. Thanks for posting

 

[norman bates]

I had a 10" whizzered full range (ish) driver run wide open (that rolled off above 1.5kh) ran with a single capped tweeter.

When the tweet was slid up close to the front of the baffle, I could not hear any difference flipping the polarity of the tweeter.
But you could hear the dip (flipping the polarity on the tweeter) once you slid the tweeter back a few inches so the acoustic centers were close.

I believe time alignment is more audible on "sharp" sounds, such as woodblock or Snare drum, jingling key, rain, etc.

 

[Cbdb2]

When a Speaker has inconsistent Dispersion (ie wide dispersion in the bass, but narrow dispersion in the treble), the off-axis sound reflecting off the walls will be missing high frequencies. Because these reflected waves bounce around and reach the ear later than the direct sound, your brain perceives this imbalance as a "Smearing" of Time and Space, which degrades the perceived stereo Image and Timing.

Reflections are never frequency flat.

 

[gberchin]

But you could hear the dip (flipping the polarity on the tweeter) once you slid the tweeter back a few inches so the acoustic centers were close.

Time-misalignment has three potential effects that come to my mind (there may be others):

1. Corruption of transient signals due to time-of-arrival differences.
2. Corruption of frequency response due to interference and cancellation. See post #41.
3. Corruption of timbre due to harmonic phase differences.

Evidently #1 is inaudible for all but the most egregious misalignment. Perhaps #2 and #3 are more readily audible. In all cases, the effects of a reverberant room probably swamp the problem.

EDIT to add: 4. Corruption of stereo image.

 

[sam_adams]

How they would measure is another story.

These plots were done almost fourteen years ago and I don't remember if this was before or after I rebuilt the crossovers—can't seem to find all the measurements I'm sure I made at the time. Probably on a HD somewhere in the archives.

 

[Trifonov Audio]

Honestly, it's a bit unfair to question someone's practical experiments in such a cynical way, especially when they actually checked something with their own ears instead of just repeating theory from simulations. I did a very similar experiment myself, shifting a tweeter relative to a woofer, and I had practically the same observations.

I didn't notice any major tonal changes with shifts of a few to several centimeters. Only in one specific alignment did I get better spatial coherence, that feeling that the sound 'detaches' from the speakers and comes from a single point. But those were subtle differences, not suddenly a 'completely different speaker'.

And that's why I don't really understand what 'special simulations' the commenter is referring to. Of course, you can design crossovers for specific offsets of the drivers' acoustic centers and optimize phase around the crossover region, that's standard engineering. But claiming that any listening without such 'technology' is by definition flawed sounds a bit like audiophile dogmatism.

At most, it's known that larger offsets cause interference and change the directional characteristics. But those effects are often much less audible than internet discussions would suggest. Especially in a normal listening room, with reflections and the listener's head movement.

So in my opinion, the original poster did a perfectly sensible practical experiment, and his conclusions are not 'wrong' at all. They just didn't confirm a theory that someone really wants to believe in.

Everyone is free to experiment endlessly even without a proper Assignment and with the self right to subsequent subjective comments.
For some, the Stage and if the sound no longer comes out of the speakers - may be unimportant, but for others it is crucially important!
We are talking about a mathematically proven method (Technology), not about amateur experiments through which the importance of Phase-Time Coherence is denied!!
Achieving a very good Time Domain is not possible only with audiophile "optimistically-desirable" Trial-Error, Trial-Error, but with very serious engineering knowledge!
Of course, not everyone can appreciate and notice the benefit of better Time, but without criticism and self-criticism - there is no progress.....
Until the finalization of the special Technology - there were many arrogant and nonfrendly "remarks".....

.....

 

[gnarly]

I believe time alignment is more audible on "sharp" sounds, such as woodblock or Snare drum, jingling key, rain, etc.

I do too.

To be bold....

I think time alignment, in terms of absolute constant fixed time.... is more important the higher the frequencies involved.
The impulse peaks of multiple drivers separated in time is plain frigging audible as separate events !!!!! Pure timing.
Phase up high? Less important. Of minor timing importance. Matters more in terms of smooth frequency mag response.
Down low? Phase totally rules imo/ime. Must be aligned before addressing time.



It really just comes down to this imo,.....in every measurement audio we make, time is in the fundamental denominator.
The even includes basic frequency response that ignores phase.

Time and/or phase alignment doesn't matter? Good luck with that.....
Better to ask, why is it with the current reproduction technology, that time and phase audibility are hard to detect..

 

[SteveC]

I've had a pair of these for many many years. They are the daily drivers for my video system (I'm old school and prefer stereo to surround for home movie watching). They've never let me down! https://www.hifi-classic.net/review/thiel-cs2-124.html

 

[gberchin]

Time and/or phase alignment doesn't matter? Good luck with that.....
Better to ask, why is it with the current reproduction technology, that time and phase audibility are hard to detect.

Audiophile claims that cable differences are audible are met with skepticism on ASR because they are not measurable.

Engineering measurements of errors in time-alignment are met with skepticism on ASR because they are not audible.

The irony is delicious.

 

[Holmz]

Audiophile claims that cable differences are audible are met with skepticism on ASR because they are not measurable.

Engineering measurements of errors in time-alignment are met with skepticism on ASR because they are not audible.

The irony is delicious.

But the time alignment is measurable.
Assuming fidelity is how well the reproduced signal matches the input signal, then there may be some hifi definition there.

 

[NTK]

Audiophile claims that cable differences are audible are met with skepticism on ASR because they are not measurable.

Engineering measurements of errors in time-alignment are met with skepticism on ASR because they are not audible.

The irony is delicious.

Total BS. Cable differences are easily measurable, e.g. at RF frequencies. Can any of these audiophools claims that these differences are audible but unmeasurable withstand properly conducted double blind tests?