• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Time aligned speakers - do they make sense?

Digital_Thor

Senior Member
Joined
Mar 24, 2018
Messages
381
Likes
330
Location
Denmark
My understanding is that Earl's subs are bandpass subs, so their top end has a built-in third order lowpass function before any additional filtering.

My experience with distributed multisub systems has been that it's a good idea to aggressively roll off the top end of any subs that are positioned well away from the main speakers (like alongside or behind the listening area), so they don't betray their locations by passing audible upper bass/lower midrange energy. I use a 4th order lowpass filter, and my understanding is that that's what Earl uses as well (3rd order acoustic lowpass + 1st order electrical lowpass).
Now... that's interesting. I use a fourth order on my 4 subs too, and seem to get pretty nice results, when merely following his advice, on having the mains with no high-pass, or at least a first order. So I have a first order on my mains at 60Hz, then a different cross-over point for each sub, but with high enough filter, for them not to be localized. Now, when I measure in the seated position, there only seem to be variance around the area where my subs cross - when I move the mic half a meter around where I usually sit.
Another thing. I find it to be more balanced, with woofers that are minimum 8", when combining them with subwoofers that are 12" and 15". The overall impact and SPL seems more balanced this way. Tried with 6,5" woofers, since they only have to play down to 60Hz.... but I can't make it work like with bigger woofers.
 

Holmz

Major Contributor
Joined
Oct 3, 2021
Messages
2,018
Likes
1,241
Location
Australia

I know that there is great food everywhere - if you look. Same here in Denmark...

Actually had one of the best steaks in Amsterdam.
The Noma place in Copenhagen is supposedly worth getting the snout into the trough at.
 

Dougey_Jones

Addicted to Fun and Learning
Forum Donor
Joined
May 27, 2020
Messages
552
Likes
460
My mom moved to LA in the 60's and had my half-brother, which I love just like my "full" sister. And the stories are pretty straight up, fitting with what you see and hear in the media. I guess history is repeating and that most cultures have their things sticking pretty well, for quite some time, making up our identity, if you will. Look at us Danes.... we are still vikings, like Americans are cowboys. Guess tv and movies did their trick :p
Nothing is black and white, but funnily enough, my new colleague, who just moved in from New York, a few months ago.... brings up stories that fit so damn well... that it's almost scary. My other colleague, who worked in a restaurant in the US, confirm several stories too, while his dad gave Wall Street a go.... still trying... and that is a rough story all by itself.
Nothing is perfect.... we could easily turn the story upside down with someone moving from Denmark. But I believe it's like speakers.... there are compromises - but there are also clear trends. Some of my absolute favorite comedians, like George Carlin, Bill Burr, Christopher Titus and more, joked about all of this, for years.
If our tiny country can admit that we have our faults, then hopefully your huge country, can admit that there's maybe a bit of dust and mold in the corners ;)

But I'm sorry for letting myself getting carried away - I will try and stay on the subject :)
There’s plenty wrong with the good ol’ US of A, that we’re all able to be here together enjoying this hobby together is what matters. Not to get too political, but, I’ve been greatly heartened to see how quickly the world has circled the wagons around Ukraine and rejected naked aggression. Excited to welcome new Nordic friends to NATO and reject whatever this ideology is that Russia seems to be pushing.
 

dualazmak

Major Contributor
Forum Donor
Joined
Feb 29, 2020
Messages
2,823
Likes
2,951
Location
Ichihara City, Chiba Prefecture, Japan
Now... that's interesting. I use a fourth order on my 4 subs too, and seem to get pretty nice results, when merely following his advice, on having the mains with no high-pass, or at least a first order. So I have a first order on my mains at 60Hz, then a different cross-over point for each sub, but with high enough filter, for them not to be localized. Now, when I measure in the seated position, there only seem to be variance around the area where my subs cross - when I move the mic half a meter around where I usually sit.
Another thing. I find it to be more balanced, with woofers that are minimum 8", when combining them with subwoofers that are 12" and 15". The overall impact and SPL seems more balanced this way. Tried with 6,5" woofers, since they only have to play down to 60Hz.... but I can't make it work like with bigger woofers.

Interesting...

Just for our reference, using 8-wave, 3-wave and single-wave tone burst signals, we can rather easily objectively measure as well as objectively visualize (using Adobe Audition) and objectively adjust time alignment between subs and woofers, cleanliness and tightness of the sound given by subs, woofers and most importantly subs+woofers at the overlapped Fq zone. (I have done these recently. See my post #86 on this thread.)

You wrote "when I measure in the seated position...". Did you use tone burst signals and visual analysis tools like Adobe Audition?

This type of objective measurements and adjustments are also very much useful for determination of the XO frequency and selection of XO filter slopes.

Of course, any of objective/subjective tuning on our audio system should be finally evaluated by careful subjective listening sessions at our listening position in our listening environments, though.
 
Last edited:

gnarly

Addicted to Fun and Learning
Joined
Jun 15, 2021
Messages
992
Likes
1,390
Interesting...

Just for our reference, using 8-wave, 3-wave and single-wave tone burst signals, we can rather easily objectively measure as well as objectively visualize (using Adobe Audition) and objectively adjust time alignment between subs and woofers, cleanliness and tightness of the sound given by subs, woofers and most importantly subs+woofers at the overlapped Fq zone. (I have done these recently.)

Hi dualazmak,

First let me say I've enjoyed reading your project links you posted in #86 of this thread. That's the most extensive use of wavelet bursts for timing, that I've seen. Nice effort you put in sharing that.

I've been using ARTA to study the use of wavelets for timing. It has a built-in storage o'scope that makes captures and comparisons pretty easy.
Also use REW's scope which has really nice graphical capability, but alas no storage facility.

Here's a recent example of tuning a speaker and then checking it with wavelet bursts that might be of interest (to those who think time-alignment is valuable ;))

Speaker is the same 4-way main i posted about on first page of this thread.
In place were a 120Hz high pass, and xovers at 300Hz, 750Hz, and 6.3kHz. All those were linear phase 96dB/oct LR.
The speaker was tuned/timed to about 1m using the conventional technique of matching frequency response to a target curve. Target being Flat mag and phase....
here's the acoustic mag and phase measurement.
syn9t spot tune for burst.JPG



Ok, then to see how well wavelets looked for a "near perfectly tuned" speaker, I ran 1.5 cycle bursts at various frequencies to check how well their time alignments held up.

One immediate pleasant surprise was how extraordinarily clean the microphone captured, acoustic wavelet captures looked.
So I tuned off the FIR processing for each driver and compared, and wavelet captures clearly degraded, ranging from a little to a lot, depending on where in the driver's passband i was testing.
Anyway, that's not the point of this post....I'm just trying to explain how the next set of captures can look so good...like pure electrical, not acoustic.
It's simply the power of FIR to a spot, which will obviously degrade some off the spot (but not nearly as much as folks seem to think, ime.)

back to topic....how do the wavelet bursts time align...
Here are comparisons of 150Hz to various frequencies, 300Hz, 700Hz, 2000Hz, and 6.3kHz....in order.
Note the time cursor is locked at 178ms on all.
arta spot burst 150 Hz & 300Hz.JPG

arta spot burst 150 Hz & 700Hz.JPG

arta spot burst 150 Hz & 2000 Hz.JPG

arta spot burst 150 Hz & 6300Hz.JPG


It's pretty clear to me that flat mag and phase produce excellent wavelets with spot-on timing.
They are equivalencies imo, just like perfect mag and phase, equals perfect impulse, step, and square waves...etc


Of course, any of objective/subjective tuning on our audio system should be finally evaluated by careful subjective listening sessions at our listening position in our listening environments, though.
Absolutely agree. After all, in the end it's all about pleasing the ears, huh? :)

I think if more folks could hear speakers with precise time alignment (which again for me is flat mag and phase), they might change their minds about its audibility.
 

Digital_Thor

Senior Member
Joined
Mar 24, 2018
Messages
381
Likes
330
Location
Denmark
Hi dualazmak,

First let me say I've enjoyed reading your project links you posted in #86 of this thread. That's the most extensive use of wavelet bursts for timing, that I've seen. Nice effort you put in sharing that.

I've been using ARTA to study the use of wavelets for timing. It has a built-in storage o'scope that makes captures and comparisons pretty easy.
Also use REW's scope which has really nice graphical capability, but alas no storage facility.

Here's a recent example of tuning a speaker and then checking it with wavelet bursts that might be of interest (to those who think time-alignment is valuable ;))

Speaker is the same 4-way main i posted about on first page of this thread.
In place were a 120Hz high pass, and xovers at 300Hz, 750Hz, and 6.3kHz. All those were linear phase 96dB/oct LR.
The speaker was tuned/timed to about 1m using the conventional technique of matching frequency response to a target curve. Target being Flat mag and phase....
here's the acoustic mag and phase measurement.
View attachment 208202


Ok, then to see how well wavelets looked for a "near perfectly tuned" speaker, I ran 1.5 cycle bursts at various frequencies to check how well their time alignments held up.

One immediate pleasant surprise was how extraordinarily clean the microphone captured, acoustic wavelet captures looked.
So I tuned off the FIR processing for each driver and compared, and wavelet captures clearly degraded, ranging from a little to a lot, depending on where in the driver's passband i was testing.
Anyway, that's not the point of this post....I'm just trying to explain how the next set of captures can look so good...like pure electrical, not acoustic.
It's simply the power of FIR to a spot, which will obviously degrade some off the spot (but not nearly as much as folks seem to think, ime.)

back to topic....how do the wavelet bursts time align...
Here are comparisons of 150Hz to various frequencies, 300Hz, 700Hz, 2000Hz, and 6.3kHz....in order.
Note the time cursor is locked at 178ms on all.
View attachment 208205
View attachment 208206
View attachment 208207
View attachment 208210

It's pretty clear to me that flat mag and phase produce excellent wavelets with spot-on timing.
They are equivalencies imo, just like perfect mag and phase, equals perfect impulse, step, and square waves...etc



Absolutely agree. After all, in the end it's all about pleasing the ears, huh? :)

I think if more folks could hear speakers with precise time alignment (which again for me is flat mag and phase), they might change their minds about its audibility.
I'm curious... how do your speakers measure off-axis - relative to on-axis?
 

gnarly

Addicted to Fun and Learning
Joined
Jun 15, 2021
Messages
992
Likes
1,390
I'm curious... how do your speakers measure off-axis - relative to on-axis?
Sure. here's same speaker with some horiz measurements made indoors 0-10-20 degrees.
Which gives a fairly generous 40 degree wide listening area.
Tuning was for optimizing 0-20 degs, not trying to be perfect to a spot like for the wavelet tests in my prior posts.

syn10 indoor polars 0-20 H 10deg V.JPG


I don't really like or trust polars made indoors, as i think best measurements are far-field and without gating.... which for me requires outdoors....and at >3m.

Hopefully, I'll be outside soon, putting together a more extensive set of off-axis measurements.....
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
My understanding is that Earl's subs are bandpass subs, so their top end has a built-in third order lowpass function before any additional filtering.

My experience with distributed multisub systems has been that it's a good idea to aggressively roll off the top end of any subs that are positioned well away from the main speakers (like alongside or behind the listening area), so they don't betray their locations by passing audible upper bass/lower midrange energy. I use a 4th order lowpass filter, and my understanding is that that's what Earl uses as well (3rd order acoustic lowpass + 1st order electrical lowpass).

Yes, I think you're right... though, he did not specifically mention the "bandpass" design of his subs in the video I watched him discuss it.
 

dualazmak

Major Contributor
Forum Donor
Joined
Feb 29, 2020
Messages
2,823
Likes
2,951
Location
Ichihara City, Chiba Prefecture, Japan
I'm curious... how do your speakers measure off-axis - relative to on-axis?

My strategy/policy on possible (or impossible?) off-axis measurement in my mulitchannel multi-driver multi-way multi-amplifier stereo setup is almost identical to @gnarly's above approach; for satisfactory off-axis measurements, I believe that I need to bring my entire system into anechoic chamber or move to wide open outside field those would be unrealistic at all.

Consequently, I always dare to limit myself to quasi-on-axis measurements at the listening position in my actual room environments.

I am always interested in some improvement(s) of the total sound quality at my listening position in my listening environment, as @Purité Audio (keith) simply and kindly wrote here; "You must hear equipment in your own room in your own system, compare unsighted if there isn’t an immediately apparent difference/improvement. To go further if there isn’t a significant improvement then don’t change anything, the largest gains are speakers and room."
 
Last edited:

Frgirard

Major Contributor
Joined
Apr 2, 2021
Messages
1,737
Likes
1,040
My strategy/policy on possible (or impossible?) off-axis measurement in my mulitchannel multi-driver multi-way multi-amplifier stereo setup is almost identical to @gnarly's above approach; for satisfactory off-axis measurements, I believe that I need to bring my entire system into anechoic chamber or move to wide open outside field those would be unrealistic at all.

Consequently, I always dare to limit myself to quasi-on-axis measurements at the listening position in my actual room environments.

I am always interested in some improvement(s) of the total sound quality at my listening position in my listening environment, as @Purité Audio (keith) simply and kindly wrote here; "You must hear equipment in your own room in your own system, compare unsighted if there isn’t an immediately apparent difference/improvement. To go further if there isn’t a significant improvement then don’t change anything, the largest gains are speakers and room."
The ground plane measurement on a parking
 

dualazmak

Major Contributor
Forum Donor
Joined
Feb 29, 2020
Messages
2,823
Likes
2,951
Location
Ichihara City, Chiba Prefecture, Japan
Of course, any of objective/subjective tuning on our audio system should be finally evaluated by careful subjective listening sessions at our listening position in our listening environments, though.

I recently started the still-ongoing "post series", therefore, here and thereafter entitled "Excellent Recording Quality Music Albums/Tracks for Subjective (and Possibly Objective) Test/Check/Tuning of Multichannel Multi-Driver Multi-Way Multi-Amplifier Time-Aligned Active Stereo Audio System and Room Acoustics; at least a Portion and/or One Track being Analyzed by Color Spectrum of Adobe Audition in Common Parameters" (sorry for rather long tile) .

Your visits and participations will be highly appreciated.
 
Last edited:

witwald

Senior Member
Forum Donor
Joined
Jul 23, 2019
Messages
478
Likes
502
Over the many previous decades, approximate solutions to the driver time-alignment problem have been attempted by a number of loudspeaker manufacturers, sometimes with surprisingly good results. However, loudspeakers with passive crossovers seem to need 1st-order crossovers to achieve any semblance of a genuinely time-aligned behavior, owing to the excess phase shift introduced by higher-order crossovers (which are of course necessary for many other practical reasons). Some examples of square wave reproduction from past loudspeakers that had been designed to be time-aligned follow.

The B&W DM6 from the 1970s:
1677300015937.png

The Duntech PCL25:
1677300260087.png


Duntech Crown Prince:
1677367398227.png

Dunlavy SC-I Monitor:
1677368486194.png

Which modern loudspeakers can get close to that level of theoretically-ideal performance?
 
Last edited:

sigbergaudio

Major Contributor
Audio Company
Forum Donor
Joined
Aug 21, 2020
Messages
2,639
Likes
5,397
Location
Norway
I missed this thread the first time around. Just wanted to share our approach on this in case it's interesting / relevant to the discussion.

With our active speakers we use assymetric / parabolic filters that start out as 1.order, and then a steeper filter is added later to protect the drivers, but late enough that linear phase is maintained.

With regards to using hard DSP to force flat response and / or phase, we now use only very low Q filters to avoid any audible ringing. This means the response isn't perfectly flat, but it does sound more natural. The difference here is very subtle, but I think it's there.

Related (and completely anecdotal): We're testing the waters towards studio audience, and a recent mixing/mastering engineer who visited to audit the Manta speakers had an interesting comment: He did not enjoy the sound of modern, very flat studio monitors (I will not mention brands), as they came across as "processed" and fatiguing (his words). In my understanding it wasn't about a bright sound per se. That this impression may be a result of the filters is speculation on my part, but it was an interesting comment.
 

ppataki

Major Contributor
Joined
Aug 7, 2019
Messages
1,216
Likes
1,358
Location
Budapest
I missed this thread the first time around. Just wanted to share our approach on this in case it's interesting / relevant to the discussion.

With our active speakers we use assymetric / parabolic filters that start out as 1.order, and then a steeper filter is added later to protect the drivers, but late enough that linear phase is maintained.

With regards to using hard DSP to force flat response and / or phase, we now use only very low Q filters to avoid any audible ringing. This means the response isn't perfectly flat, but it does sound more natural. The difference here is very subtle, but I think it's there.

Related (and completely anecdotal): We're testing the waters towards studio audience, and a recent mixing/mastering engineer who visited to audit the Manta speakers had an interesting comment: He did not enjoy the sound of modern, very flat studio monitors (I will not mention brands), as they came across as "processed" and fatiguing (his words). In my understanding it wasn't about a bright sound per se. That this impression may be a result of the filters is speculation on my part, but it was an interesting comment.
Very interesting approach indeed!
What do you mean by 'very low Q'? Is it like below 5 or even below 2?
Thank you
 

Bjorn

Major Contributor
Audio Company
Forum Donor
Joined
Feb 22, 2017
Messages
1,286
Likes
2,562
Location
Norway
I see some speaker brands claiming to be physically time-aligned, usually by titling the speaker back so the tweeter is farther from the listener. Is this truly effective? It seems if there’s a positive effect, it would be for only one narrow listening position (which may not be where you’re sitting ). Are there any demonstrated, measured benefits to this design, if your head isn’t in a clamp? Or are there other benefits to this design?
If you compare a speaker with no time alignment between drivers and adding time alignment with for example LR 4th order crossover (which isn't linear phase), I would say the difference is very audible. The time aligned speaker will sound more precise, has better imaging and less fatiguing. The difference will depend but is quite obvious in an AB test.

If this is done physically, this almost always have trade offs. For instance will tilting the speaker lead to more ceiling reflections and which can be quite detrimental. And with horn speakers, diffraction become a major issue. So time alignment with delay and DSP is much better.

When a crossover is added after time alignment with delay, like LR 4th order, there isn't 100% time alignment anymore and phase isn't linear. This can be approved with for example a FIR filter. Does this give a clear audible improvement over crossover with delay and LR 4th? IMO, that's at best icing of the cake and will depend on the frequency.

So we need to distinguish between almost 100% time alignment using delay, and 100% time alignment with FIR, all pass filter or 1th order with linear phase. The first is very important IMO but the latter is more a minor improvement in most cases.

Hope that makes sense.
 

sigbergaudio

Major Contributor
Audio Company
Forum Donor
Joined
Aug 21, 2020
Messages
2,639
Likes
5,397
Location
Norway
Very interesting approach indeed!
What do you mean by 'very low Q'? Is it like below 5 or even below 2?
Thank you

It's below 0,7.
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,755
Likes
16,208
If this is done physically, this almost always have trade offs. For instance will tilting the speaker lead to more ceiling reflections and which can be quite detrimental. And with horn speakers, diffraction become a major issue. So time alignment with delay and DSP is much better.
I would add having all three is even better, for example using a concentric design as otherwise you have the alignment only for one vertical axis and not for the ceiling and floor reflections.
 

Bjorn

Major Contributor
Audio Company
Forum Donor
Joined
Feb 22, 2017
Messages
1,286
Likes
2,562
Location
Norway
I would add having all three is even better, for example using a concentric design as otherwise you have the alignment only for one vertical axis and not for the ceiling and floor reflections.
Everything is a compromise. I have tested a good number of concentric drivers. They all have their trade offs IMO. They aren't 100% time aligned either, which effects the frequency response and needs to be corrected with delay. Synergy/unity horns will also have their own issues as well. It's matter of picking the best compromise.
Personally I prefer a driver that covers a wide frequency area.
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,755
Likes
16,208
They aren't 100% time aligned either, which effects the frequency response and needs to be corrected with delay.
That's why I wrote above to combine them with delay. I agree and also always say that everything is a compromise and each has to find the one that fits his needs best.
 

Bjorn

Major Contributor
Audio Company
Forum Donor
Joined
Feb 22, 2017
Messages
1,286
Likes
2,562
Location
Norway
That's why I wrote above to combine them with delay. I agree and also always say that everything is a compromise and each has to find the one that fits his needs best.
Out of curiosity and off-topic, which is your favourite driver/design here?
 
Top Bottom