Time aligned speakers - do they make sense?

[Plcamp]

Time alignment of drivers seems to me to simply be alignment in the z (back to front) axis, so I don’t see how that would be less important than aligning drivers in the y (vertical) or x (horiz, use a coax) axis?

 

[gnarly]

Very interesting, I learned a lot, thank you. I am still a little fuzzy on some details. After reading your post my understanding is that linear phase crossovers don't ring because the ringing is cancelled out by the two filters at the crossover point. So here a few questions. How about minimum phase crossover filters and post ringing? Seems like they would be more prone to ringing? It is popular with DIY digital crossovers to not use the exact same slope or even the exact crossover point in order to "blend" the drivers together from flatter FR but with what you are saying this would be a bad idea from the "ringing" standpoint. Is that right? Do you have any insight into how audible "ringing" is for either crossovers or independent high and low pass filters? Thanks

Thx, glad it made sense.

I don't know what to think about minimum phase post-ring.
I know high order IIR xovers, and high Q filters of any type, are generally shunned.
Are they shunned due to greater post-ring? Or does the ear simply not like abrupt phase shifts? I'd love to know the answer....

I think the prevailing DIY digital xover practice of often using asymmetrical xover slopes is correct....due to the way they go about it.
Which is generally to use a xover slope to sum with the natural driver rolloff to achieve a final symmetric acoustic xover slope/order.
The thing is, if the drivers' being summed together don't have the same raw response slopes/orders of natural roll-off, they are asymmetric from a raw perspective. And this requires asymmetry from an electrical perspective, to create the complementary acoustic symmetry.

It's my strong belief that acoustic xovers need to be fully complementary. That is the goal of all electrical xovers imho, to achieve acoustic complementary..
I think the typical DIY approach that uses either asymmetric or symmetric xover orders, and massages them against driver roll-offs, does that.
(although there are easier techniques like described below, imo)

I can use steep linear phase electrically complementary xovers, because I do minimum phase pre-adjustments to the drivers' natural roll-offs.
Basically, i flatten mag and phase as far out-of-band as I want the summation to occur with neighboring driver.
This flattening is the inverse of the driver's natural roll-off, typically an order or two, maybe 4th order at most.
This flattening boost of course raises drive levels beyond the point of safety, particularly on the low end of the driver.
Then i slap a steep 16th order or brickwall xover, on top of the few orders of boost used for the out-of-band flattening, and the net electrical xover order remains very high...always at least 12th order or greater, ime.

The idea of flattening out-of-band, prior to adding xovers, can be used with any order xover high or low. It just requires that the summation of the flattening and the xover order, provide enough electrical attenuation for safe operation.

This flatten first, add xovers method, ends up providing the exact same solution as the typical DIY process...provided the same acoustic order is reached. Again, for any final acoustic order, low or high.

Steep just makes it really easy, and really safe.
Price of steep is it must be linear phase, otherwise we are back to our post-ring quandary when trying to use steep IIR.

 

[Plcamp]

The idea of flattening out-of-band, prior to adding xovers

I like this, and sort of tried it…but with Minidsp 2x4 hd there’s not enough fir taps to correct out of band phase and gain (or at least I couldn’t figure it out adequately). I was trying to equate the two drivers an octave out of band prior to L-R crossover.

How do you achieve this?

 

[gnarly]

I like this, and sort of tried it…but with Minidsp 2x4 hd there’s not enough fir taps to correct out of band phase and gain (or at least I couldn’t figure it out adequately). I was trying to equate the two drivers an octave out of band prior to L-R crossover.

How do you achieve this?

Aah cool!

For the out-of-band flattening, use the 2x4HD's ten bands of parametric EQ. Series of shelving filters are particularly effective.
The out-of-band flattening should be IIR minimum phase anyway, where mag and phase are fix one, you fix the other too.
Oh, fix in band IIR too.

Then just use the FIR taps for linear phase xovers you made in rePhase or wherever. Good way to start.

Ultimately, if you get the FIR bug, automated FIR generators that make FIR files to match target curves are the way to go.
Essentially, matching response to a target curve does the same thing as flatten first, and add xover....only without so much work !

 

[Plcamp]

Heh, I can say I did precisely as you suggest…except I failed in any attempt to use FIR…2x4 hd has limited available taps, and I was trying to align under 500 hz between twin 15” woofers and a fullrange. as I recall there are only 1200 taps available per channel…and I remember convincing myself I can’t get there from here.

 

[Digital_Thor]

Ultimately, if you get the FIR bug, automated FIR generators that make FIR files to match target curves are the way to go.
Essentially, matching response to a target curve does the same thing as flatten first, and add xover....only without so much work !

well... I tried that. And you really have to be careful, because any automated process, can't give you any better result, than the data that you put in. So you might end up with quite a few weird sounding results, before you feel that it sounds somewhat good enough, to worth the effort, compared to IIR - IMO.

 

[Digital_Thor]

Heh, I can say I did precisely as you suggest…except I failed in any attempt to use FIR…2x4 hd has limited available taps, and I was trying to align under 500 hz between twin 15” woofers and a fullrange. as I recall there are only 1200 taps available per channel…and I remember convincing myself I can’t get there from here.

I would say that the benefit of FIR, is the precision, compared to IIR. But at lower frequencies, it really ain't worth the hassle - IMO.
The wavelength is so long, that you mostly listen to a complex mix of reflections anyway - not direct sound, which in return means, that phase only serves as a tool to align individual bass drivers to create a total sum of sound = correct amplitude at the listening position. You have to remember, that you "only" change the phase of the signal driving the amplifier, driving the speaker unit. As soon as the sound bounces around the room.... the room fully takes over, and it's now an acoustical problem.
Maybe the sound will be more focused/clear, if you use FIR on the more direct sound - above the transition point, of around 500Hz - in most rooms.
But FIR filtering vs IIR, is still up for debate. And there is still no real consensus, whether one is absolutely better than the other, in all situations.

Some claim to have learned what to listen for - and now can't live without linear phase - for one or several reasons.

Some need to learn what to listen for, and then a special track or sound, to hear the difference - and some even need headphones on top of this, to really sense when actual linear phase even makes an audible difference. So imagine how it pans out in practical everyday listening situations, where you just hear a few tracks on the stereo, while half sleeping on the sofa before bed - in a normal reflective living room.

Some people sleep better, when they finally obtain that close to perfect step response. But just imagine how difficult it is to remain unbiased, when you've just spend hours and hours, to tune in that filter.

Some day I might hear a system that actually plays better with this technology, rather than some other. And I applaud anyone trying to improve sound reproduction overall. But until now - it seems like - no one solution, is always the right one.

 

[gnarly]

Heh, I can say I did precisely as you suggest…except I failed in any attempt to use FIR…2x4 hd has limited available taps, and I was trying to align under 500 hz between twin 15” woofers and a fullrange. as I recall there are only 1200 taps available per channel…and I remember convincing myself I can’t get there from here.

Hi, just making sure i understand...
Are you saying you need a xover at 500Hz? or rather two of them for stereo?
If so, here's the slippage for a 1200 tap @96kHz lin phase hpf....it's not perfect but it should work, no? (lpf slippage is less)

There's nothing else going in the FIR file than the xover, right? All the flattening is via IIR, yes?

 

[gnarly]

well... I tried that. And you really have to be careful, because any automated process, can't give you any better result, than the data that you put in. So you might end up with quite a few weird sounding results, before you feel that it sounds somewhat good enough, to worth the effort, compared to IIR - IMO.

Haha...ain't that the truth!!!!

Automated FIR is so dang powerful...and sooo dang easy to screw up.... badly.
Tis why starting out with full manual work in rePhase is such a great training ground.

Hopefully, by the time corrections make sense manually, we've actually learned enough and gained enough measurement experience, to know what measurements are valid for correction.
And hopefully by then, we realize there is no shortcut vs driver by driver correction. IOW, No global FIR corrections (i'm talking at the speaker building level)

If all this comes together, the results are so very worth it ime....especially down low,..... in contrast to the bad luck you seem to have had??

 

[mhardy6647]

Time alignment? Who needs time alignment?

View attachment 207020

so-called "Time alignment" actually turned out to be critical in two way loudspeakers in the early days of the "talkies".
I'll bet you knew that already.
I think this was already mentioned before in this thread.

source: https://worldradiohistory.com/Archive-All-Audio/Archive-Audio/70s/Audio-1977-08.pdf
Hilliard reference cited above (Ref. 1) J. Hilliard, "Historical Review of Horns Used for Audience Type Sound Reproduction" J.A.S.A. 59 (1), 1 (Jan. 1976).

Industry Pioneers #18: Dr. John K. Hilliard, The Standard Of Motion-Picture Sound

Hilliard set Hollywood's audio standards, designed some of the earliest cinema speakers, and made innovations in microphone technology.

www.soundandcommunications.com

 

[MakeMineVinyl]

so-called "Time alignment" actually turned out to be critical in two way loudspeakers in the early days of the "talkies".
I'll bet you knew that already.
I think this was already mentioned before in this thread.

View attachment 207050
source: https://worldradiohistory.com/Archive-All-Audio/Archive-Audio/70s/Audio-1977-08.pdf
Hilliard reference cited above (Ref. 1) J. Hilliard, "Historical Review of Horns Used for Audience Type Sound Reproduction" J.A.S.A. 59 (1), 1 (Jan. 1976).

Industry Pioneers #18: Dr. John K. Hilliard, The Standard Of Motion-Picture Sound

Hilliard set Hollywood's audio standards, designed some of the earliest cinema speakers, and made innovations in microphone technology.

www.soundandcommunications.com

I was being facetious. Those Western Electric horns were one of the reasons the Shearer horn research began. Their time delay between drivers must have been measured in minutes.

 

[mhardy6647]

I was being facetious. Those Western Electric horns were one of the reasons the Shearer horn research began. Their time delay between drivers must have been measured in minutes.

tap-tap!

whew! Sorry I missed the facetiousness... facetiousnosity... umm... whatever
I was worried about you for a moment, there!

 

[Holmz]

Using 16th order / brickwall filters to improve driver alignment and phase brings up the question of which is more audible phase issues or filter ringing neither of which are well studied as far as I can tell.

From a dispersion perspective, that will yield a massive change as one moves in frequency through that region.

I would say that the benefit of FIR, is the precision, compared to IIR. But at lower frequencies, it really ain't worth the hassle - IMO.
...

Group delay EQ could be a good case for FIR

 

[pseudoid]

RichardVandersteen's floor standing speakers attempt both at time-alignment and "phase-correct designs" of the drivers.
A tough act!

 

[gnarly]

RichardVandersteen's floor standing speakers attempt both at time-alignment and "phase-correct designs" of the drivers.
A tough act!

Yeppers, back when passive xovers ruled, stuff by Vandersteen, Theil, and Dunlavy among others, helped demonstrate the value of phase-correct, time-aligned designs.

Nowadays, phase-correct and time-aligned is almost a trivial task, via multi-way active with DSP.

But folks apparently like to cling to passive type thinking, and still see/hear value, that simply doesn't make "sound sense" anymore given the difficulty of pulling off correctly..

 

[gnarly]

From a dispersion perspective, that will yield a massive change as one moves in frequency through that region.

Hi, no clue what you mean...
massive change in what? thru what region?

I just look at polar responses. On -axis vs off-axis measurements.
Ime, High order lin phase xovers have helped definitely those converge, with less variance

Whatcha mean? thx

 

[ernestcarl]

When one is using multiple subs, say, 3 or more for multiple seating positions or rows in a home theater, does perfect (or near-perfect) linear phase down to the bass even make much practical sense at all? Has anyone actually achieved a perfect looking step response in all seats in a home theater? I remember Geddes asking how one's supposed to even go about getting it "perfect" (linear phase and ideal-looking step response below Schroeder) in his dedicated home theater/listening room.

 

[Holmz]

Hi, no clue what you mean...
massive change in what? thru what region?

I just look at polar responses. On -axis vs off-axis measurements.
Ime, High order lin phase xovers have helped definitely those converge, with less variance

Whatcha mean? thx

I mean a 12” woofer will have a certain horizontal radiation at 490 Hz.
If a 3” MR is 100% at 510 Hz, then the dispersion angle will do a 4:1 jump over a small frequency range.

This is true when phase tracks, and if the phase is being wild, then it can be worse again.


The opposite example would be a wideband driver with no crossover… and the horizontal and vertical dipersion will slowly narrow as the frequency increases.

 

[gnarly]

When one is using multiple subs, say, 3 or more for multiple seating positions or rows in a home theater, does perfect (or near-perfect) linear phase down to the bass even make much practical sense at all?

Good question, and one i don't know the answer to.
I'm not a fan of the multiple sub approach.
I use stereo subs closely co-located with stereo mains, such that they are within 1/4 wavelength of each other through the xover summation range.

Has anyone actually achieved a perfect looking step response in all seats in a home theater? I remember Geddes asking how one's supposed to even go about getting it "perfect" (linear phase and ideal-looking step response below Schroeder) in his dedicated home theater/listening room.

I expect not. perfect impulse/step will occur only to a specific spot ime.
I think the goal is a reasonably defined impulse/step over as wide an area as possible.

HT is not my gig anymore. I spent about 15 years, 1990-2005, working on it and multi-channel audio. Even built large dedicated room, acoustically designed/treated, using high quality prosound boxes. Great sounding room and setup...far better than my longstanding time with electrostats and planars.
Then i made the mistake of taking the prosound speakers outdoors....broke my heart how much better outdoors sounds....no turning back for ultimate SQ.
Now, I tune DIY speakers outdoors for best quasi-anechoic I can.....and critical listening is outdoors. Truly glorious....worth the effort and wait (most of the time LOL)

I know from exchanges with Geddes, he prefers the sound of speakers and multiple subs indoors. Likes the room envelopment.
For me, it's the exact opposite. Indoors sounds like mud in comparison to outdoors...no matter how good indoors gets..
The mud vs clear distinction, only seems to grow as measurements get better and better. (maybe I have some confirmation bias going on...dunno.)

It probably has a lot to do with our underlying preference for direct vs reflected sound, and what type music we listen too...again, more dunno...

 

[gnarly]

I mean a 12” woofer will have a certain horizontal radiation at 490 Hz.
If a 3” MR is 100% at 510 Hz, then the dispersion angle will do a 4:1 jump over a small frequency range.

This is true when phase tracks, and if the phase is being wild, then it can be worse again.


The opposite example would be a wideband driver with no crossover… and the horizontal and vertical dipersion will slowly narrow as the frequency increases.

Gotcha. Thx.

yeah, a 12" woofer mated to a 3" mid range is bad acoustic design for sure....directivity mismatch train wreck Lol
Not sure how phase fits into a discussion about basic directivity. ...???