Moving mic vs. averaged vs. single point measurements - which is better?

[Blumlein 88]

This is probably a ridiculous question, but is there any way to start and stop the RTA measurement with a delay? I can't quite reach the mouse from the MLP. A longer cable is obviously another solution, but perhaps I'm missing some setting.

I use a wireless keyboard with a TouchPad. You could use a cheap wireless mouse.

 

[dominikz]

After a lot of testing of various methods I now use MMM almost exclusively when I do room EQ. The reasons:
- It is quick and simple to do
- Measurement results are very repeatable and comparable between takes
- It does spatial smoothing/averaging that hides high-frequency irregularities that we don't hear anyway, but keeps the low frequency ones (which are very audible)
- It is easy to find the preferable target curve slope based on it
- MMM result is usually directly comparable to anechoic PIR above ~1kHz
- Filters calculated based on an MMM measurement with variable smoothing and a target curve that follows the MMM approximate slope always sound good to me (note that I only EQ below 300Hz, and only use subtractive IIR PEQ - no boosts!)

 

[ozzy9832001]

Honestly, I think it really depends on how much you move and how varied your position can be. When I'm really listening to music I'm leaning back in my chair, I don't move all that much, so I single measurement is more than enough. If I was really interested I'd take 2 other measurements the width of my head and avg them with the center position, just to make sure there aren't any crazy things going on. Have to move the MLP if that happens.

When I'm working on my computer, I'm closer to the wall, so the bass is louder as I sit forward in my seat. I have a separate EQ setting for that, which was done quick and dirty with MMM.

The results are close enough to a single point.

 

[jhaider]

The spatially averaged responses do not deviate significantly from the single point measurement at the listening position for small spatial averaging displacements (± 0.1 m).

I've emphasized the important bit. I'm not convinced that an 200cm/8" sphere is terribly useful.

Note also this paper is a convention paper. For a peer reviewed study on spatial averaging, see Geddes & Blind, 1986.

E-library page - AES

AES E-Library ← Back to search

www.aes.org

FWIW, I generally do spatial averaging because I don't like to be in the room with sweeps. (I tend to run REW on one computer and control it from a different room.) Now it's relatively inexpensive to do fixed multi-mike. REW Pro is a $100 add on to unlock multi-mike input and then you can use whatever mics you want or have. If you're good with being in the same room then MMM seems like a fine method too. Cheaper as well, given that you only need 1 mike.

 

[droid2000]

Just use Dirac.

 

[Blumlein 88]

This is the best paper I've seen on MMM. Note among other details that most people move the microphone much too fast. The paper mentions some call the proper MMM measurement as the Tai Chi method when you do it at the proper slow speed. It also notes MMM gives much closer repeatable results than other methods. This paper references Geddes, Blind, some work from Harman and other JAES papers.

https://www.ohl.to/audio/downloads/MMM-moving-mic-measurement.pdf

Yes this is the paper linked the OP. I feel like lots of people commenting have not read it. It is only 18 pages and half of that are graphs.

 

[droid2000]

What's the point of spending so much time quibbling on measurement strategies when the avg joe just creates REW filters? Adjusting for one dimension is inadequate in a 3d world.

 

[thewas]

In my personal experience up to room transition frequency where it makes sense to correct based on LP measurements sweeps and MMM give very similar results https://www.audiosciencereview.com/...ement-using-the-mmm-method.24131/#post-815429 , but still prefer using MMM as it doesn't take much time and can be smoother at the high frequencies which I only use as a reference for the level "stitching" of my low frequencies correction as I correct the higher frequencies based on anechoic or windowed listening window measurements.

 

[JeremyFife]

Very basic question on MMM with a UMIK-1 for stereo listening: point the mike at the speakers, or at the ceiling (standard calibration file or 90deg calibration)?
I've seen both recommended. Seems as if the 90deg option is for Home Theatre, but I can't seem to get a definitive answer.

Thanks

 

[staticV3]

Very basic question on MMM with a UMIK-1 for stereo listening: point the mike at the speakers, or at the ceiling (standard calibration file or 90deg calibration)?

Please read the paper in the OP.
As Blumlein wrote three posts above yours, it is only 18 pages long with half of them being graphs.
On page 11 is the answer to your question.

 

[JeremyFife]

Please read the paper in the OP.
As Blumlein wrote three posts above yours, it is only 18 pages long with half of them being graphs.
On page 11 is the answer to your question.

View attachment 314527

Thanks ... I was opening the paper as I posted the question ... shockingly lazy!

"... Mic shall not be pointed directly to loudspeaker : due to semi-diffuse field, microphone can be pointed randomly during scanning or kept at 90° from source ..."

 

[dominikz]

Very basic question on MMM with a UMIK-1 for stereo listening: point the mike at the speakers, or at the ceiling (standard calibration file or 90deg calibration)?
I've seen both recommended. Seems as if the 90deg option is for Home Theatre, but I can't seem to get a definitive answer.

Thanks

I'd suggest to point the microphone to the ceiling with the 90° calibration file applied. Perhaps you will find the explanation in this post useful.
Though microphone orientation is not really that important if you only do correction in the low frequencies.
Actually, I'll go even further and say that in my experience you can sometimes even use non-measurement microphones without calibration with good results (as long as they have relatively flat LF response) - see this example.
But on the other hand without a calibrated reference it is difficult to know whether your results are valid or not - so there definitely is a lot of value in a calibrated measurement mic!

 

[Bjorn]

If the question is related to do a full frequency correction in the listening position, both methods are basically flawed but just in different ways. There's no way neither methods can be applied in a way to correct only what's minimum phase except for perhaps sub frequencies.

 

[Keith_W]

OK I have a very strange result with my MMM measurement. Specifically, the tweeter.

I use the Acapella TW-1S tweeter. This is a plasma tweeter with integrated electronics with a volume control. The stupid thing about this tweeter is that the volume control is stepless, and has no markings. The ONLY way to set the volume accurately is with measurements. When I initially set up the speaker, I levelled the volume between the two tweeters with a test tone played by REW and an SPL meter.

TLDR:
- MMM shows diverging frequency response at the tweeter crossover point, with the LEFT speaker dB louder than the right. We will call the louder left tweeter the culprit tweeter. This is not seen on the SPS which says both are normal.
- Swapping the culprit left tweeter now makes the RIGHT tweeter louder. This confirms that the volume of the culprit tweeter has been set too loud.
- I performed SPS at LLP, MLP, and RLP. At MLP, the volume of tweeters is equal. At LLP the right (culprit) tweeter is slightly louder. At RLP the right (culprit) tweeter is much louder. This explains why the MMM shows a diverging tweeter response.
- I pulled out my SPL meter and measured the SPL at both tweeters at the mouth of the horn. Both tweeters measure exactly the same.
- I swapped the tweeters back, and reduced the volume of the left (culprit) tweeter using MMM. Now the MMM looks perfect, but the SPS shows that the culprit left tweeter is much lower in volume than the right.
- I wondered if it was a room issue, so I took spectros of both tweeters at the MLP with REW. The spectros look exactly the same.

The Evidence
I performed this measurement at the MLP:

The SPS was performed using Acourate's mic alignment method - Acourate plays a series of alternating high frequency clicks between left and right tweeter, and you move the mic until the amplitudes align exactly. It does not depend on the volume of the amplitude, only the timing. As you can see, with the mic dead center, the sweep shows that the tweeters are equal in volume. However, when I performed the MMM, it shows that the LEFT tweeter is louder by 3dB, implying that I did not set tweeter volume correctly.

So I performed this experiment - what happens if I swap the left tweeter for the right? This was the result:

Two MMM's performed at MLP. You can see that the louder tweeter, which was on the left, is now on the right. Conclusion: this is very strong evidence that the tweeter volume has been set incorrectly.

But this leaves the question: why does the SPS show that the volume of the tweeter is the same? So I performed another experiment. Using Acourate's mic alignment tool, I took a SPS dead center. Then I moved the mic 30 samples to the left and right, and took more SPS. This was the result:

I windowed the measurements very tightly to ensure that room influence plays minimal part in the result. As you can see, the SPS shows that the right channel is louder when the mic is anywhere but at dead center.

I swapped the tweeters back. Now what happens if we were to use the MMM to adjust the volume of the tweeter?

I adjusted the tweeters until they were the same with MMM. Then I did a SPS:

And as expected ... the Left tweeter is now softer.

So I was now wondering whether some kind of room acoustic issue is causing the volume mismatch. I cracked out REW and adjusted the tweeters back to the same volume:

And then looked at the Spectrogram:

Left

... and Right.

 

[3ll3d00d]

it's in a separate enclosure like this right? if so, take them outside (or into the centre of a suitably sized room) and measure them quasi anechoically

 

[OCA]

OK I have a very strange result with my MMM measurement. Specifically, the tweeter.

I use the Acapella TW-1S tweeter. This is a plasma tweeter with integrated electronics with a volume control. The stupid thing about this tweeter is that the volume control is stepless, and has no markings. The ONLY way to set the volume accurately is with measurements. When I initially set up the speaker, I levelled the volume between the two tweeters with a test tone played by REW and an SPL meter.

TLDR:
- MMM shows diverging frequency response at the tweeter crossover point, with the LEFT speaker dB louder than the right. We will call the louder left tweeter the culprit tweeter. This is not seen on the SPS which says both are normal.
- Swapping the culprit left tweeter now makes the RIGHT tweeter louder. This confirms that the volume of the culprit tweeter has been set too loud.
- I performed SPS at LLP, MLP, and RLP. At MLP, the volume of tweeters is equal. At LLP the right (culprit) tweeter is slightly louder. At RLP the right (culprit) tweeter is much louder. This explains why the MMM shows a diverging tweeter response.
- I pulled out my SPL meter and measured the SPL at both tweeters at the mouth of the horn. Both tweeters measure exactly the same.
- I swapped the tweeters back, and reduced the volume of the left (culprit) tweeter using MMM. Now the MMM looks perfect, but the SPS shows that the culprit left tweeter is much lower in volume than the right.
- I wondered if it was a room issue, so I took spectros of both tweeters at the MLP with REW. The spectros look exactly the same.

The Evidence
I performed this measurement at the MLP:

View attachment 338942
The SPS was performed using Acourate's mic alignment method - Acourate plays a series of alternating high frequency clicks between left and right tweeter, and you move the mic until the amplitudes align exactly. It does not depend on the volume of the amplitude, only the timing. As you can see, with the mic dead center, the sweep shows that the tweeters are equal in volume. However, when I performed the MMM, it shows that the LEFT tweeter is louder by 3dB, implying that I did not set tweeter volume correctly.

So I performed this experiment - what happens if I swap the left tweeter for the right? This was the result:

View attachment 338943
Two MMM's performed at MLP. You can see that the louder tweeter, which was on the left, is now on the right. Conclusion: this is very strong evidence that the tweeter volume has been set incorrectly.

But this leaves the question: why does the SPS show that the volume of the tweeter is the same? So I performed another experiment. Using Acourate's mic alignment tool, I took a SPS dead center. Then I moved the mic 30 samples to the left and right, and took more SPS. This was the result:

View attachment 338944
I windowed the measurements very tightly to ensure that room influence plays minimal part in the result. As you can see, the SPS shows that the right channel is louder when the mic is anywhere but at dead center.

I swapped the tweeters back. Now what happens if we were to use the MMM to adjust the volume of the tweeter?

View attachment 338963

I adjusted the tweeters until they were the same with MMM. Then I did a SPS:

View attachment 338965

And as expected ... the Left tweeter is now softer.

So I was now wondering whether some kind of room acoustic issue is causing the volume mismatch. I cracked out REW and adjusted the tweeters back to the same volume:

View attachment 338966

And then looked at the Spectrogram:

View attachment 338967

Left

View attachment 338968

... and Right.

Did you check toe-in angles?

 

[Keith_W]

it's in a separate enclosure like this right? if so, take them outside (or into the centre of a suitably sized room) and measure them quasi anechoically

Yeah those are the ones. Brilliant idea, I will take them outside tomorrow, weather permitting.

Did you check toe-in angles?

Not recently, but the speakers have not been moved. The toe in was set with long rulers and a protractor a few years ago. Those speakers are big and heavy and they won't move, so I am sure they would be the same.