The moving microphone method (MMM) for Dummies using REW

[Keith_W]

The y split I'm talking about is a double y split (soldered) it is TRS connected to main cable and another TRS also connected to main cable (and the previous trs) then the main cable on the other end has 2 rcas.

This makes no sense. Please draw what you are talking about.

 

[JD1]

This makes no sense. Please draw what you are talking about.

Minute 2

 

[Keith_W]

That loopback cable you constructed is a "loopback + monitoring" cable. It is used for recording audio from a PC and listening to the recording at the same time. It is not suitable for measurements.

How that cable works is something like this:

Windows audio source (e.g. Youtube, computer games, etc) --> OUTPUT 2 --> LOOPBACK --> INPUT 2 --> Recording software

and simultaneously:

Windows audio source (e.g. Youtube, computer games, etc) --> OUTPUT 2 --> CABLE --> amps, speakers.

Like I mentioned, the purpose of loopback audio measurements is to compare the measurement circuit to the loopback circuit. This is a very important concept to grasp. You can not have your loopback circuit and measurement circuit on the same channel. Well you can, but it would only be suitable for measuring very short impulses. If the length of the impulse is longer than the latency of the measurement circuit, then the loopback impulse will bleed into the measurement impulse since they are both on the same impulse response.

Please go back and read this post again.

 

[JD1]

That loopback cable you constructed is a "loopback + monitoring" cable. It is used for recording audio from a PC and listening to the recording at the same time. It is not suitable for measurements.

How that cable works is something like this:

Windows audio source (e.g. Youtube, computer games, etc) --> OUTPUT 2 --> LOOPBACK --> INPUT 2 --> Recording software

and simultaneously:

Windows audio source (e.g. Youtube, computer games, etc) --> OUTPUT 2 --> CABLE --> amps, speakers.

Like I mentioned, the purpose of loopback audio measurements is to compare the measurement circuit to the loopback circuit. This is a very important concept to grasp. You can not have your loopback circuit and measurement circuit on the same channel. Well you can, but it would only be suitable for measuring very short impulses. If the length of the impulse is longer than the latency of the measurement circuit, then the loopback impulse will bleed into the measurement impulse since they are both on the same impulse response.

Please go back and read this post again.

But what is capturing sound is input 1 beyerdynamic mm1 not input 2



Input 2 is simply for timing reference





This works for sweeps.







My question wasn't about input 2 or if the loopback works or not.



It was exclusively about output 2 y splice and how that works since I can't select mono in rew, and if I choose uncorrelated is it actually in fact correlated since it's Y split to left and right wiim channels, and if the summing of a equal right signal to left and right is summed properly for subwoofer.



Basically how this works

"(2) use your Y splitter from the Focusrite to drive both L/R inputs on the Wiim simultaneously".

 

[Keith_W]

I am having trouble picturing how you have cabled your setup. I understand your loopback cable. Please draw a diagram.

 

[JD1]

Focusrite output 2 TRS into input 2 TRS and both to main cable that then Y splits into 2 rca that go into wiim left and right line in input.
Xlr mic beyerdynamic mm1 into focusrite input 1.

In rew preferences I have:
microphone: input 1
output: output 2
Timing reference: output 2
Loopback: input 2

 

[Keith_W]

As mentioned, there is no way you can use your loopback cable as you constructed for timing measurements. There is no REW setting that will work, the way you have cabled it. I am sorry, but there is nothing more for me to say.

 

[RickS]

Focusrite output 2 TRS into input 2 TRS and both to main cable that then Y splits into 2 rca that go into wiim left and right line in input.
Xlr mic beyerdynamic mm1 into focusrite input 1.

In rew preferences I have:
microphone: input 1
output: output 2
Timing reference: output 2
Loopback: input 2

While REW's creator is a member here, the support forum is hosted elsewhere:

Official REW (Room EQ Wizard) Support Forum

Guides / Links / Notes / Setup / Connection / Testing Official Support Forum for Room EQ Wizard (REW) by John Mulcahy

www.avnirvana.com

Suggest you try there as perhaps they may be able to add some additional perspective.

 

[RosalieTheDog]

This is such a useful guide. I always refer to it to measure and then do some quick PEQ on the low end. Thanks!

I am about to measure again (new OS) to apply some PEQ. I have two questions that are perhaps tangentially related to the measurement method, so apologies in advance if this is not the thread.

1) Sample rate settings. I usually set everything to the default 48hz / 16 bit, but since I am at my PC I guess I can choose higher settings too. Any point in this? I would apply use REW to generate EQ filters for Easy Effects on Linux.

2) Does it make sense or is it necessary to measure L and R outputs separately and EQ them separately? Or is it best to just measure L+R and EQ the result?

 

[Timcognito]

A nice tutorial here IIRC those topics are covered

The moving microphone method (MMM) for Dummies using REW

After getting a couple of weird results on my first MMM measurements, I went searching around the forums & YouTube for what went wrong. To save others from gathering various bits and pieces of info from a dozen sources, I decided to publish this quick write-up here. You can use it as a rough...

www.audiosciencereview.com

 

[Mort]

Is an MMM needed if you are only measuring Schroeder and below? The energy is omnidirectional right?

 

[Keith_W]

Is an MMM needed if you are only measuring Schroeder and below? The energy is omnidirectional right?

You probably haven't reached the part of the book where the MMM is mentioned The short answer is yes, you still need a spatially averaged measurement below Schroder. But I prefer doing multiple sweeps and averaging them when looking at low frequencies.

 

[janbth]

2) Does it make sense or is it necessary to measure L and R outputs separately and EQ them separately? Or is it best to just measure L+R and EQ the result?

If you are going to apply EQ in the bass region, you should measure L+R. This will take into account the interference you get between L and R. Go for the peaks, and be careful with boosting the dips.

 

[5omeone]

Great tutorial, many thanks!

1. Open the RTA window and configure the graph controls (gear symbol in the upper right) as follows:
   View attachment 341680
   • Mode "RTA 1/48 Octave"
   • Smoothing "1/48 smoothing"
   • Max Overlap "50" (default) to "75" (slightly quicker)

Just out of curiosity: Why are you deviating from the standard settings here for the ones marked in red? What exactly does “1/48 octave” mean?
Many thanks and best regards

 

[RandomEar]

Great tutorial, many thanks!


Just out of curiosity: Why are you deviating from the standard settings here for the ones marked in red? What exactly does “1/48 octave” mean?
Many thanks and best regards

1/48 is the RTA resolution in octave fractions, see REW help. In contrast to spectrum mode, this reduces the resolution for higher frequencies: You do not need to know where a peak is down to 0.X Hz at 15 kHz, but at 150 Hz it might be useful for determining room modes accurately. At the same time, using 1/48 smoothing gets rid of some noise. You may be fine with 1/24 for both or other combinations, I just eyeballed what looks good

I think the max overlap is 75% by default, but 50% was working totally fine for me. As explained by the creator of REW here, this setting seems to be relevant for clock drift correction.

 

[5omeone]

Great, many thanks for the clarification! Appreciate!

One final question: Why is a pink noise used instead of a sweep? If I remember correctly Sonos is using for their room correction a sine sweep with MMM approach.

 

[NTK]

Pink noise is a statistically stationary signal. When you are moving the mic, the source signal is statistically the same regardless of where the mic is at any moment.

When you use a sweep, the frequency of the source signal varies with time, and there is only one active frequency in the signal at any instant. If you are constantly moving the mic, the response the mic is measuring at position 1 is only of one single frequency, which is most likely unknown as you most likely doesn't keep track. And when the mic is moved to position 2, the response will be of another single unknown frequency. When you aggregate the measurements together, your frequency response will be: magnitude at frequency A is measured at (unknown) position 1, magnitude at frequency B is measured at (unknown) position 2, magnitude at frequency C is measured at (unknown) position 3, ... etc. Not very useful.

 

[5omeone]

All right, thanks so much to both of you for the detailed explanation! Now I have a much better idea of what I'm actually doing and measuring. Time to give it a try.