A history of Time Aligned Loudspeakers

[gberchin]

... I do like to use higher-order xovers than MDS allows (or at least allows at my skill of MDS implementation).

In the latest MDS paper (version 20250714) I outline how to achieve basically any transition band slope that one might desire.

 

[norman bates]

You all are so lucky with the digital tools available nowadays !!!!!

 

[KL....]

You all are so lucky with the digital tools available nowadays !!!!!

Thank you, @norman bates, it can be also be suggested that Active Speakers (utiliseing DSP/FPGA PQGA/other) can (easyly) solve/resolve all of the issues mentioned in this thread (no (physical) XOs required) and more (GGNTKT M3), can't they, although/nonetheless, nice/enjoyable to discuss the History....

 

[Thomas_A]

Thank you, @norman bates, it can be also be suggested that Active Speakers (utiliseing DSP/PQGA/other) can (easyly) solve/resolve all of the issues mentioned in this thread (no (physical) XOs required) and more (GGNTKT M3), can't they, although/nonetheless, nice/enjoyable to discuss the History....

What they solve is quite much dependent what you define as "issues". Audible issues? Measurable issues? Distortion issues?

 

[KL....]

What they solve is quite much dependent what you define as "issues". Audible issues? Measurable issues? Distortion issues?

Thank you, @Thomas_A, Yes, but which would you suggest, at this time? It could be suggested as All (the thread posts indicate, possable), and even the Room, although perhaps not quite yet (at least at a reasonable price)....

 

[Thomas_A]

Thank you, @Thomas_A, Yes, but which would you suggest, at this time? It could be suggested as All (the thread posts indicate, possable), and even the Room, although perhaps not quite yet (at least at a reasonable price)....

Frequency response linearity and low distortion (audible stuff). For crossovers < 500 Hz one can have a second look at phase linearity.

 

[Waveform Fidelity]

Can someone comment on "blended' crossovers--initially 1st order then transitioning to higher orders?

The transfer function (order) of XO filters used in loudspeakers is fixed. There are a class of filters that adapt their transfer function (think noise cancelled headphones) based upon a feedback signal - they are called adaptive filters.

Perhaps you are thinking of filters that have a variable attenuation with frequency? Yes one can make different slopes by adding series/parallel R-C, R-L networks to standard filters, which some manufacturer's do to fine tune responses/suppress resonances etc. What ever you can do with passive components, you can do in DSP with IIR filters.

Lastly, FIR filters allow one to design filters with arbitrary gain and phase - so you can create any correction filter that you like to produce any acoustic response you like. Using that type of filter the constraints are the number of filter “taps” and the corresponding delay it produces. At low frequency you need many taps, so its common to combine IIR and FIR filters, or you can use Warped FIR filters which reduces the number of taps by using logarithm spacing in the design process.

Hope this helps

 

[Ultrasonic]

US manufacturer Meadowlark made numerous speakers that qualify for this thread I think. I used a pair of Shearwaters for many years (now sitting surplus to requirements upstairs). Here's a photo of their better-looking Nighthawk as an example:

Edit: I can't quickly find relevant measured data for these but here's some for the Sharwater model including the step-function:

Meadowlark HotRod Shearwater loudspeaker Measurements part 2 | Stereophile.com

In the vertical plane (fig.6), a presence-region suckout that develops immediately above the tweeter axis might balance the off-axis flare. On the other hand, sit lower than the tweeter and more presence-region energy becomes apparent, which will exaggerate the subjective consequences of the...

www.stereophile.com

 

[RetroStereo]

Wouldn't Jon Dahlquist's first speaker released under his own brand fit this category? The DQ10 was a remarkable speaker.

 

[nagster]

Earthworks Sigma 6.2

 

[norman bates]

interesting how the tweeter distances to edges are 3 different distances……. That will smooth out peaks / valleys. Still diffraction, but at least spread out to 3 different times and less due to round overs…

Except for the big diffraction source from the to of the woof cabinet………

 

[Trifonov Audio]

A few examples of full Time compensation - Not Quasi, Not Pseudo! With crossovers higher than first order.....

 

[ripvw]

I first experienced the potential benefits of time-alignment by purchasing a pair of B&O Beovox M70 floor standing speakers after hours of comparison testing against my existing Acoustic Research AR11's.

Bang & Olufsen M70 Receivers (1978) — Specs & History | VTA

Bang & Olufsen M70 (1978) — 4-ohm floor-standing speaker with 38Hz–20kHz response, 10-inch woofer, and time-aligned baffle. Priced at $700–$790 in period reviews.

vintagetechnologyarchive.com

I was working for Circuit City when they picked up the DCM line of speakers. I compared the Time Window 3 against the best JBL and Infinity speakers available to me and ended up buying a pair. I had a chance to pick the brain of designer Steve Eberbach over the next couple of years and came away very impressed. He was truly an expert in filter theory and passive crossover design. I also admired but never owned his top of the line Time Window 7's.

https://www.akdatabase.com/AKview/albums/userpics/10007/DCM%20TW3%20Manual%20and%20Sell%20Sheet.pdf

My last time aligned speaker purchase was the venerable Vandersteen Model 2CE. My wife and I compared them to a number of conventional speakers at our local high end dealership after I had moved on from consumer electronics retail and was working as an IT analyst for ExxonMobil. There have been a number of different Model 2's over the years and it was the speaker that put Vandersteen on the map so to speak. They are apparently still being made:

Vandersteen Model 2Ce Signature III | Vandersteen Audio

Vandersteen Model 2Ce Signature III

www.vandersteen.com

John Atkinson and others tested the Quad ESL-63 and found that it could reproduce a visually perfect square wave as well as deliver the classic triangular shaped step response. Despite its capabilities the designer Peter Walker was not sold on the importance of perfect phase response. This brilliant speaker designer seemed to have made the ESL-63 time coherent because he could, not because he was a true believer in the benefits. I loved listening to but never owned a pair

Quad ESL-63 loudspeaker Measurements | Stereophile.com

Sidebar 2: Measurements (from June 1989, Vol.12 No.6) As I had dragged my motley collection of test equipment (footnote 1) over to Larry Archibald's listening room to carry out a set of measurements on the Mirage M1 loudspeaker he has reviewed elsewhere in this issue, I thought it might be a...

www.stereophile.com

Of course it has been mentioned that modern active loudspeakers using DSP can deliver excellent frequency and phase response and nowhere it that more clearly indicated then ASR's own test of the Neumann KH 150:

Neumann KH 150 Monitor Review

This is a review, listening tests and measurements of the Neumann KH150 DSP 2-way studio monitor (active speaker). It is on loan from the company and costs US $1,750 (each). The design language is not changed of course. The main woofer is 6.5 inches now powered by 120 watt dedicated...

www.audiosciencereview.com

I am very happy with my Genelec 8320A's and twin 7350 subs as a final system but if I was buying new today I think I would listen first to the KH 150's from Neumann...

 

[Trifonov Audio]

There are many misrepresentations, misuses and even frauds with this concept: Time Alignment or Phase Coherent.
If there is no publication of a real measured Step Response - it is very likely that the advertisement was created for gullible payers with low technical and sound culture!....

 

[Chris A]

A few examples of full Time compensation - Not Quasi, Not Pseudo!

But only if the microphone is on-axis--not off-axis.

For full time alignment, you need electronic delay of higher frequency driver signals for drivers arranged on a flat baffle or in multiple horn apertures...or drivers arranged in less than quarter wavelength axial separation distances at crossover frequency (and dividing network filter delays) inside a multiple entry horn.

Chris

 

[Trifonov Audio]

But only if the microphone is on-axis--not off-axis.

For full time alignment, you need electronic delay of higher frequency driver signals for drivers arranged on a flat baffle or in multiple horn apertures...or drivers arranged in less than quarter wavelength axial separation distances at crossover frequency (and dividing network filter delays) inside a multiple entry horn.

Chris

Of course - only on the axis and in the "selfish" Sweet Spot! Who cares what is heard on the balcony or in the kitchen.....?
Indeed, complete alignment - is not an easy task! Because even ONLY the exact arrangement of the Acoustic Centers does not give an flat Frequency Response, even a good Step Response!

 

[Duke]

A few examples of full Time compensation - Not Quasi, Not Pseudo! With crossovers higher than first order.....
View attachment 549889
View attachment 549891
View attachment 549892

Thanks @Trifonov Audio! I remember when that "pregnant kangaroo" configuration was used to achieve time alignment, taking into account the effects of the crossover on inter-driver phase response. The inter-driver phase differential is compensated for by offsetting the highpass driver to the rear by a distance corresponding to the time delay arising from the relative phase lag of the lowpass driver at the crossover frequency.

A dirty little secret of the well-beloved 4th order crossover is this: Yes the highpass and lowpass drivers are "in phase" in the crossover region, but without time-domain compensation (either via one wavelength of physical offset or similar digital delay) the lowpass driver's output arrives 360 degrees (one wavelength) BEHIND the highpass driver's output! There is data indicating the ear is relatively insensitive to this one-wavelength inter-driver delay, but I don't think there's any data showing that correcting it is detrimental.

A somewhat under-the-radar technique exists for achieving physical time alignment in what looks, from the outside, like a simplistic two-way hybrid horn speaker:

Not obvious at first glance is this:

At the crossover frequency, the compression driver's diaphragm is physically offset one-half wavelength behind the woofer's diaphragm. And the crossover slopes are acoustically second order through the crossover region. The second-order slopes result in the woofer phase-lagging the compression driver by 180 degrees at the crossover frequency. The net result is, the one-half wavelength of distance offset compensates for that 180 degrees of phase lag!

So we get the full-on time-domain benefits of the "pregnant kangaroo" configuration without the appearance (and diffraction) thereof. Some tradeoffs have to be juggled to get it all to come together, and a rather robust compression driver is called for.

Credit to Earl Geddes for teaching me this.

 

[Trifonov Audio]

Thanks @Trifonov Audio! I remember when that "pregnant kangaroo" configuration was used to achieve time alignment, taking into account the effects of the crossover on inter-driver phase response. The inter-driver phase differential is compensated for by offsetting the highpass driver to the rear by a distance corresponding to the relative phase lag of the lowpass driver.

A dirty little secret of the well-beloved 4th order crossover is this: Yes the highpass and lowpass drivers are "in phase" in the crossover region, but without time-domain compensation (either via one wavelength of physical offset or similar digital delay) the lowpass driver's output arrives 360 degrees (one wavelength) BEHIND the highpass driver's output! There is data indicating the ear is relatively insensitive to this one-wavelength inter-driver delay, but I don't think there's any data showing that correcting it is detrimental.

A somewhat under-the-radar technique exists for achieving physical time alignment in what looks, from the outside, like a simplistic two-way hybrid horn speaker:

View attachment 549954

Not obvious at first glance is this:

At the crossover frequency, the compression driver's diaphragm is physically offset one-half wavelength behind the woofer's diaphragm. And the crossover slopes are acoustically second order through the crossover region. The second-order slopes result in the woofer phase-lagging the compression driver by 180 degrees at the crossover frequency. The net result is, the one-half wavelength of distance offset compensates for that 180 degrees of phase lag!

So we get the full-on time-domain benefits of the "pregnant kangaroo" configuration without the appearance (and diffraction) thereof. Some tradeoffs have to be juggled to get it all to come together, and a rather robust compression driver is called for.

Credit to Earl Geddes for teaching me this.

Correct and well said! Although if we include phase rotation in degrees instead of just (pure) Time Delay - some misdirection may occur....
One-half wavelength offset is good start for Time optimisation!

 

[gberchin]

The second-order slopes result in the woofer phase-lagging the compression driver by 180 degrees at the crossover frequency. The net result is, the one-half wavelength of distance offset compensates for that 180 degrees of phase lag!

I want to be certain that I understand what you're saying here. It looks like you have implemented a 2nd-order Linkwitz-Riley crossover, but instead of inverting the polarity of one of the drivers, you're offsetting it by one-half wavelength. Correct?

If so, then the attached graphic shows what happens to the frequency response.

 

[Space is the place]

I had a pair of DM16s, I had to machine face plates for the new tweeters, but they weren't bad for £80, friend still has them in his garage.