Acoustic phase measurements of loudspeakers

[daftcombo]

Yeh, have done something like this with a pair of 3-way speakers with LR4 crossovers.

I measured the acoustical phase of the speakers and then corrected it with a filter generated in RePhase (I only partially corrected the bottom end roll-off, though), then measured to confirm the filter had worked (which it had).

In that case, I thought I could hear a difference when I first implemented the filter, but then was not able to reliably discern a difference when informally ABX testing myself. I didn't try too hard though - perhaps with more work I could have got to the point of reliably detecting a difference. But if there was anything audible there, I decided it was so subtle as to be unimportant.



If you convolve the RePhase filter with a recording, you can ABX it against the original using Foobar.

Yes, that's what I did. To convert a few songs, I used convolver as an option of the Foobar2000 converting tool. I got two series of tracks: the original ones and the convolved ones. Then I tried to ABX with the Foobar ABX plugin.
I meant that I failed the tests!

 

[andreasmaaan]

Yes, that's what I did. To convert a few songs, I used convolver as an option of the Foobar2000 converting tool. I got two series of tracks: the original ones and the convolved ones. Then I tried to ABX with the Foobar ABX plugin.
I meant that I failed the tests!

Aha! Had misunderstood you. So yes, we had exactly the same experience

I was not so surprised though, as I'd previously failed a more controlled form of the same test, in which I created a filter that emulated the phase distortion produced by that same pair of speakers and then ABX tested myself with that using headphones, the use of which is known to increase sensitivy to phase distortion.

 

[QMuse]

Lots of overlap, but amusing variations.

You'd have much less variations and phase graphs would look "normal" if you apply FDW of say 6 cycles to all these measurements. Alternatively you can measure from closer distance, say 1m. When you measure from LP reflections are causing 360deg dropouts so phase graph doesn't make much sense.

 

[RayDunzl]

You'd have much less variations and phase graphs would look "normal" if you apply FDW of say 6 cycles to all these measurements.

I see, but don't really understand.

 

[QMuse]

I see, but don't really understand.

What is that you don't understand?

 

[ReaderZ]

You have to make a few educated guesses or inferences based on what you know about the speaker to interpret it. It's definitely not as straightforward as a phase response graph.

There's a good basic explanation here, and a better, more detailed explanation here (about 1/5 of the way down the page). It was also discussed in some depth on ASR here.

Hope that helps a bit?

Hmm, how do you read a phase response graph? I really don't understand all the 180, 270 360 phase shift shows on speaker reviews or linear phase = good idea(why? can we hear it?)

 

[Linearphass]

I wonder why the MLs produce such a flat response. Reflections at the microphone position should cause wild deviations. Something to do with them being a line source? Another factor that I'm overlooking?

It is the close mic position that makes reflects small and hence phase smoother.

 

[Linearphass]

I am sceptical it will tell you anything about perceived audio quality. I have yet to find any studies that show any significant impact of phase response (or any other time domain behavior, for that matter) on loudspeaker preference in realistic conditions.

However, I am interested in phase response for a different reason: it can help selecting speakers that can be combined together in one system. For example, let's say I want to buy a pair of front speakers and a (different) centre speaker. Ideally, I'd want to use a centre speaker that has a similar phase response to the other speakers, otherwise, they will be out of phase in some parts of the spectrum, and will therefore interfere with each other. That, for me, is a good reason to publish phase data.

I believe that phase changes are subtly audible BUT are often obscured by room reflections. The reflections will produce waveform distortions similar to phase shifts which obscure the underlying information. It is difficult to set up a test to validate this. Headphones are probably the easiest way. if you have the ability to create an unity gain all pass phase shifter you can switch it in and out and hear for yourself. The changes are most audible on transients. They are also audible on massed strings.

One speaker system where this is audible is a set of Linkwitz LX minis with Veledyne subs. This system has been converted to an linear phase device using a DEQX processor and a 4th order Linear phase crossover The drivers were measured with the system elevated so they were centered vertically in a room with a 20' ceiling. The nearest wall was also about 10' away. The calibrated mic was placed 1M on the tweeter axis. From this we generated the speaker correction files using FIR filters. Basically this makes the speaker very close to perfectly flat phase and frequency response at one point in space.

Then the speakers were careful positioned in a large listening room that had been carefully treated using 3 spaced quilts on each sidewalk, a cloud over head and bass traps with diffusion plates in each corner. There is also a very large riser maybe 8x14' that was built that has two low Q limp mass bass traps tuned to the null frequencies of the room. The traps were placed at the location of pressure maxima at those null frequencies. Each device was careful placed and adjusted for maximum results this took over 100 hours of effort. The RT60 is very uniform at around 250mS across the entire spectrum down to 100Hz using an OMNI source.

Room EQ was adjusted using minimum phase parametric EQ at each of the remaining peaks. The resultant response is flat from 50Hz to 15Khz +/-1.5dB at the MLP.

The listening result was a distinctly better when compared with the same speakers in the same room and position using LR24 at the same frequencies. Levels were closely matched as was the frequency response.

The improvements in clarity and articulation are the most notable.

 

[RayDunzl]

It is the close mic position that makes reflects small and hence phase smoother.

If 10 feet is close then I measured with a close mic position.

Both the ML and JBL were measured at the listening position.

The JBL phase "flattens out" if close mic'd.

The ML - stand with your nose to the panel, and they still "sound" 2~3 feet distant. The sound is very low level at any square inch of the 720 sq inch area.

 

[Linearphass]

My mistake. For some reason I thought you were measuring very close. After re-rereading I agree the ML's are much more directional than the JBL's. That dramatically reduces the number and intensity of reflections and make for a smoother phase curve. Moving the mic closer to the speaker does the same for most speakers. I like to use linear vertical scale when showing ETC. I think it is labelled %, I also use normalize to 100% and normalize at the IR peak.
Makes for a much more readable display. We typically go after any reflections that exceed 10% in that mode. That should be equivalent to 20dB down.
I have little experience measuring large panel speaker but would assume you have to get very far away to truly be in the diffuse field.

 

[RayDunzl]

I like to use linear vertical scale when showing ETC. I think it is labelled %, I also use normalize to 100% and normalize at the IR peak.
Makes for a much more readable display. We typically go after any reflections that exceed 10% in that mode. That should be equivalent to 20dB down.
I have little experience measuring large panel speaker but would assume you have to get very far away to truly be in the diffuse field.

Here's ETC, linear %, normalized, and zoomed in, JBL red, ML black, at 10 feet, both (L/R) speakers active:

 

[Linearphass]

Well clearly the ML's interact with the room much less that the JBL's. This is as good as any I have seen. Do you have diffusion or absorption on the front wall? I note that your room is moderately dead and fairly flat on the R60 curves.
how do the image in that room?

 

[RayDunzl]

Well clearly the ML's interact with the room much less that the JBL's. This is as good as any I have seen. Do you have diffusion or absorption on the front wall?

https://www.audiosciencereview.com/forum/index.php?threads/you-can-call-me-ray.1814/

The corners behind the speakers are rockwooled - 7x22x48 softslabs standing tall on top of each other, just put them there and that was the end of that.

The TV in the picture was 50", replaced with 75". It's about flush with the front of the speakers, 44" from the wall behind.

Two more slabs stand on end behind the couch to block the PC fans and some of the wall reflection when you are seated. Ears are about 5 feet from the wall behind.

Left rear corner is open to kitchen/breakfast area.

I note that your room is moderately dead and fairly flat on the R60 curves.
how do the image in that room?

The ML image as well as they have in two other locations. Sharp focus for individual imstrumemts and up to 90 degrees left-right on those freaky Q-sound recordings.

The JBL aren't as definined. At one point we could barely tell whether they were in or out of phase with music playing. They're OK (daily casual drivers), but get put away when a real listening session occurs.