The moving microphone method (MMM) for Dummies using REW

[kolestonin]

What am I doing wrong?
Where am I supposed to see the decibels?

 

[Keith_W]

What am I doing wrong?
Where am I supposed to see the decibels?

Your mic isn't connected. There should be some metering on the input channel. Check that you haven't muted your mic or something like that.

BTW I am assuming you can hear pink noise since REW is indicating that it is sending out signal?

 

[kolestonin]

I was indeed muting the microphone but getting a pink noise.
Now I managed to unmute the microphone, but I don't get any pink noise.
The graph on the out bar seems to have output but I can't listen anything as I was listening before.

edit: when I set output device to 'default device' I can listen the pink noise.
Should I leave it like this?
The other settings look ok?

 

[Keith_W]

Forget everything on the left for the time being. Focus on the "Preferences" window. Press the "check levels" button. You should see metering on the output, metering on the input, and hear pink noise. Once everything is OK, then proceed to take the MMM as per the guide.

 

[kolestonin]

I managed to take one measurement!
First screenshot is after I copied the finished measurement to the main REW window by clicking on the "Current" button in the top center of the RTA window, and second screenshot is after I applied VAR smoothing.

Does it look ok? Do you see any flaws in my methodology or my settings?
Asking as I would like to repeat the measurement later cause my apartment face on avenue where there is some traffic and therefore some background noise now.

 

[Keith_W]

I can't tell if your MMM is flawed or not. That graph certainly looks realistic, but who knows. ALWAYS compare your MMM to a normal sine wave sweep as a reality check. Remember that MMM's can introduce artefacts of their own, e.g. "wind" artefacts from whooshing the mic around too fast, your body (if you are holding the mic in your hand), cables slapping around and cable microphonics, etc. There will be differences between a sine wave sweep and the MMM (that's the whole point of doing the MMM!) but if you see that something is way off, then there is a problem with one of the techniques.

When I do my MMM's, I convert my mic tripod into a pole and use velcro cable ties to secure the cable to the pole so they don't slap around when I move the tripod. Then I stand off to the side and slowly move the mic around. The only time I hold the mic in my hand is if I want to measure the effect of my body in the listening position.

 

[Waldemar-Johann]

Good day everyone, this is my first post here and I am referring to this thread as this is exactly what I have now taken from RandomEar and tried out.
@RandomEar thank you very much for your effort!

Since the topic is new to me and I'm trying to work my way in and understand it, I started with REW yesterday.
The first thing I did was to create a calibration file for my audio interface (ULN-2 3D).
For this I created 2 *.mdat files and 2 calibration files one with virtual-balanced-input-on and one virtual-balanced-input-off.
I have not noticed any difference here?

This morning I started with “The moving microphone method”.
The first thing I noticed was that it got very, very loud.
Thanks again to RandomEar, everyone who does this should take point 2 seriously.
Here I also recorded 2 *.mdat with my left speaker (Quested V2108).
Microphone used SONARWORKS Calibrated Measurement Mic.

Before I continue, I would like to know if the data makes sense or if something was not done correctly?

I'm thinking about doing something with MiniDSP or Arc Studio, but maybe the room is so bad that it's not worth spending the money.
I have almost no way of treating the room.
Also attached is a calculation for the room modes for the room (rectangular).
Also attached is a mdat.zip file.

Thank you very much for your help and reply.
Have a nice Sunday

Sorry for the long post.

 

[Keith_W]

Guten Tag, Herr Johann. Wilkommen bei Audiowissenschafts-Review! I had a look at your measurements.

Here are all 14 of them, aligned to the same SPL. As you can see, they are all identical. So we can have high confidence that you took a correct measurement and your technique is consistent. I do not see any artefacts which are typical of a badly performed MMM. Well done.

There are some room modes which are expected of every speaker in every room. As your calculation shows, there is a huge spike at 113Hz. It's nice when science works as expected, isn't it?

That frequency response shows that there very little bass below 100Hz. I looked for other measurements of your speaker and I found this (from gearspace):

It looks as if your speakers should roll off from 60Hz downwards. Meaning that there is a HUUUUGE null in your freq response between about 55Hz - 110Hz. If you supply a drawing of your listening room, indicating how far the speakers are spaced from the wall and where the listening position is, it may be possible to determine why you have that null.

Re: loudness. Do not measure so loud that it hurts your ears. You will push your speaker out of its linear operating range and it will distort. Prolonged loud pink noise will also heat up the voice coil and change its performance. Not to mention, if it heats up enough, it can damage the speaker. The resulting measurement will not be representative of your normal use. If you are measuring loudly because you are concerned about the signal to noise ratio, it is better to measure at a more normal volume for longer.

Apart from that, there isn't much to say. Enjoy your time on ASR!

 

[Waldemar-Johann]

Guten Tag, Herr Johann. Wilkommen bei Audiowissenschafts-Review! I had a look at your measurements.

View attachment 463018

Here are all 14 of them, aligned to the same SPL. As you can see, they are all identical. So we can have high confidence that you took a correct measurement and your technique is consistent. I do not see any artefacts which are typical of a badly performed MMM. Well done.

There are some room modes which are expected of every speaker in every room. As your calculation shows, there is a huge spike at 113Hz. It's nice when science works as expected, isn't it?

That frequency response shows that there very little bass below 100Hz. I looked for other measurements of your speaker and I found this (from gearspace):

View attachment 463019

It looks as if your speakers should roll off from 60Hz downwards. Meaning that there is a HUUUUGE null in your freq response between about 55Hz - 110Hz. If you supply a drawing of your listening room, indicating how far the speakers are spaced from the wall and where the listening position is, it may be possible to determine why you have that null.

Re: loudness. Do not measure so loud that it hurts your ears. You will push your speaker out of its linear operating range and it will distort. Prolonged loud pink noise will also heat up the voice coil and change its performance. Not to mention, if it heats up enough, it can damage the speaker. The resulting measurement will not be representative of your normal use. If you are measuring loudly because you are concerned about the signal to noise ratio, it is better to measure at a more normal volume for longer.

Apart from that, there isn't much to say. Enjoy your time on ASR!

Hello Keith, thank you very much for your quick reply and information.
My name is actually Waldemar and my middle name is Johann, Waldemar was already taken, so Waldemar-Johann, but no offense ;-)

I am very happy that the measurement went well.
I read a lot beforehand, the problem is always the same, everyone says something different. It's best to just do it!

Yes, science works, but people don't always!

As for the zero, if it's that simple, I still have a subwoofer, but I don't think it will solve the problem.

I'll just send you some pictures now, that might be the easiest way.
You already know the size of the room.

Just so you know, I will “close” the open entrance with a heavy process.
That being said, this is my workspace and it's a priority!
I've moved around a few times but this is the way I feel most comfortable.

Best regards from small to big island.
Waldemar

 

[Soniclife]

It looks as if your speakers should roll off from 60Hz downwards. Meaning that there is a HUUUUGE null in your freq response between about 55Hz - 110Hz.

Or the speakers bass was turned down?
@Waldemar-Johann are they set flat?

 

[Keith_W]

Pictures won't help me calculate why you have that null I am afraid. I have a suspicion it is SBIR or LBIR since your room dimensions don't correlate with a huge dip centred around 80Hz. Don't forget that the side walls also affect SBIR, and if both speakers are symmetrically set up with 1m distance between each speaker and the corresponding side wall, it will add up and give you a huge dip at 80Hz. Anyway, it looks as if it may be difficult for you to move your speakers so perhaps a subwoofer might be the best solution.

 

[Waldemar-Johann]

Or the speakers bass was turned down?
@Waldemar-Johann are they set flat?

Hi Soniclife, thank you for getting in touch.
It looks like they are flat, see attachment.
I have professional tools but am an amateur myself.
Any idea how I should set up the speakers for the test?

Thank you!

 

[Soniclife]

Hi Soniclife, thank you for getting in touch.
It looks like they are flat, see attachment.
I have professional tools but am an amateur myself.
Any idea how I should set up the speakers for the test?

Thank you!

Flat would be the default before measuring, but it can be worth experimenting to see the impact they have, and in case the manufacturers idea of flat is off.

I'd also remeasure at a lower volume, unless that is how loud you use them, just in case some compression or limiter was kicking in. It's probably not that, but worth ruling out before going to far.

 

[Waldemar-Johann]

Pictures won't help me calculate why you have that null I am afraid. I have a suspicion it is SBIR or LBIR since your room dimensions don't correlate with a huge dip centred around 80Hz. Don't forget that the side walls also affect SBIR, and if both speakers are symmetrically set up with 1m distance between each speaker and the corresponding side wall, it will add up and give you a huge dip at 80Hz. Anyway, it looks as if it may be difficult for you to move your speakers so perhaps a subwoofer might be the best solution.

Hell Keith,

thank you for coming back.

I will do a Plan with GIK.
Now more pictures that are meaningful for a start.
Hope this help.

Thank you!

 

[Salt]

The dip is 60 up to 100 Hz (center 80 Hz) so not speaker issue but room mode. Moving speakers and remeasuring might confirm (or not) this.

 

[Waldemar-Johann]

Flat would be the default before measuring, but it can be worth experimenting to see the impact they have, and in case the manufacturers idea of flat is off.

I'd also remeasure at a lower volume, unless that is how loud you use them, just in case some compression or limiter was kicking in. It's probably not that, but worth ruling out before going to far.

Hello Soniclife,

thank you for the fast response and answer.

What exactly do you mean by “lower volume”?
Do you mean changing the volume in the “generator” to e.g. -24dBFS, see attachment?
I stuck to the RandomEar setting the first time.

No compression, limiter etc.

Now I'm unsure and don't want to get anything wrong.

Thank you very much for your understanding and help.

All the best,
Waldemar

 

[Waldemar-Johann]

The dip is 60 up to 100 Hz (center 80 Hz) so not speaker issue but room mode. Moving speakers and remeasuring might confirm (or not) this.

Good evening Salt,
thank you for your interest and help.

I am willing to move the speakers for testing, but where?

 

[Soniclife]

What exactly do you mean by “lower volume”?
Do you mean changing the volume in the “generator” to e.g. -24dBFS, see attachment?

How do you normaly control volume when listening, use that control. Aim for the loudest you would ever listen critically.

 

[Waldemar-Johann]

Sorry for the low response. Like every Sunday, the Internet is dead. I habe now only 3G with my phone.
That is the disadvantage of living on an island ;-)

 

[Waldemar-Johann]

How do you normaly control volume when listening, use that control. Aim for the loudest you would ever listen critically.

I don't think I understand your question now.
I entered this value, then pressed the “Play” button and then adjusted the volume on the audio interface until I reached 80 dB(Z) S in the SPL meter.
I then started the measurement.