Room Measurement

[ernestcarl]

I guess the question I have is that I have 10 or so quasi-anechoic measurements per speaker (one for every 10 degrees), and only one MMM per speaker. So what should then correspond with what?

Make your own quasi-anechoic listening window (LW) curve and use that overlayed on top of your MMM curve. This helps prevent you from over-equalizing above the transition zone or Schroeder freq. of the room. My own LW curve is made up of the rms avg of a select combination of horizontal (0-30 degrees left and right) and vertical (0-22.5 degrees up and down) group/family of curves .

From trace #1, my moving microphone measurement/method (MMM) indicate that I should apply more EQ cuts below 1 kHz. BUT, my very nearfield quasi-anechoic curves say otherwise -- speakers are point-source coaxes making it somewhat easy to make very nearfield quasi-anechoic measurements in-room. The necessary placement of the speakers inside the room (very close to numerous boundaries causing unwanted interferences/reflections) makes equalizing in between the transition zone very tricky. Relying on multiple references -- again, use 'Overlays' view window/tab in REW -- helps one to avoid applying excessive and unnecessary EQ corrections.

 

[jlo]

I guess the question I have is that I have 10 or so quasi-anechoic measurements per speaker (one for every 10 degrees), and only one MMM per speaker. So what should then correspond with what?

Oh now I understand your concerns comparing QA and MMM. I should be clearer :
- or you compare MMM in real room and MMM in anechoic room
- or you compare MMM in real room and Listening Window response in anechoic room. See LW definition in ANSI/CTA2034 for home loudspeakers , it is a mean frequency response over 9 angles near axis. It is generally near to On-Axis response but with a small slope.
Above 600/800Hz, MMM in room is also very near from anechoic OA and LW but with a bigger slope. This slope can be deduced from room dimensions, RT60 and loudspeakers directivity.
Sorry, I just saw that ernestcarl already answered !

 

[dc655321]

MMM is a spatial averaging. Depending on the scanned volume, it smoothes the high frequencies and not the whole spetrum. That's why many reflections are smoothed out and the measured MMM response in a room is nearly the same as MMM done in anechoic room above a certain frequency (see my measurements linked above).

I think we’re saying almost the same thing.

It’s just the characterization of MMM as quasi-anechoic that I disagree with.

 

[jlo]

It’s just the characterization of MMM as quasi-anechoic that I disagree with.

For sure, I agree that you can disagree.

 

[dc655321]

For sure, I agree that you can disagree.

lol. A more French response was never uttered.

 

[Rednaxela]

I blame the quasi prefix.

 

[Rednaxela]

After looking at the REW results, I will almost surely be in for the Thomann t.racks DSP plus a bundle of cables to get 2 ch. in and 4 out to the living room player. Add $100 + fgt. and about $40 for the wiring.

Depending on your gear (and your plans) you may be able to do a POC with what you already have. Have you considered this?

 

[kthulhutu]

Because MMM is a measurement method that smoothes out most reflections above 600Hz : for a comparison of same loudspeaker in anechoic room and other rooms, see https://www.ohl.to/archives/492
Speakers with nice and flat anechoic responses (Neumann, PSI, Genelec) are still smooth above 600/800Hz when measured in-room with MMM : see some measurements on loudspeakers.audio

May I ask what the purpose of MMM is when you can average single point measurements and slap on var or ERB smoothing to get basically the same result (speaker curve looks the same as anechoic performance above 1khz or so). Unless you god forbid want to make high q corrections above the transition frequency.

 

[jlo]

May I ask what the purpose of MMM is when you can average single point measurements and slap on var or ERB smoothing to get basically the same result

It is because you need about 20 single point measurements to get same precision as a one 20 seconds MMM scan : MMM is much quicker.

 

[Jim Shaw]

Why has there been silence from me on the adventures into room measurement?

I got the Behringer microphone. I got the USB interface and phantom supply. I got a mic & boom stand. I downloaded REW.
So, what's the holdup?

Well...
As I was starting to connect everything up, I dumped a half cup of latte onto my open laptop. Carumba! On the upside, I still had a half cup of the delicious liquid left to drink. On the downside, the laptop took the day off -- and a few subsequent ones. Meanwhile, I've been living like a Millennial: life through the 6-inch porthole of my phone.

I got Mr. Lenovo back from the repair shop this afternoon, and ($200 later) it is its old, usual self. So, hold onto your bonnets; I'm back in the 'get set' position.

Our intrepid Gump goes where so many have gone before. Film at eleven... curses are expected long before that.

 

[juliangst]

I have this Behringer measurement mic too and I also have a Behringer audio interface (UM2).
At what level should I set the interface's gain knob? Setting it to max volume/gain is probably too noisy (?).
How would my setup compare to a UMIK 1 setup?

 

[CauliflowerEars]

Why has there been silence from me on the adventures into room measurement?

I got the Behringer microphone. I got the USB interface and phantom supply. I got a mic & boom stand. I downloaded REW.
So, what's the holdup?

Well...
As I was starting to connect everything up, I dumped a half cup of latte onto my open laptop. Carumba! On the upside, I still had a half cup of the delicious liquid left to drink. On the downside, the laptop took the day off -- and a few subsequent ones. Meanwhile, I've been living like a Millennial: life through the 6-inch porthole of my phone.

I got Mr. Lenovo back from the repair shop this afternoon, and ($200 later) it is its old, usual self. So, hold onto your bonnets; I'm back in the 'get set' position.

Our intrepid Gump goes where so many have gone before. Film at eleven... curses are expected long before that.

So, did the adventure continue at all?
Just learning REW and hoping to find an inexpensive microphone that'd be enough for the job ...

 

[OCA]

I have quite different views on some of the comments above regarding measurements and wanted to share.

MMM, at least the way I do it, does not smooth out the HF response. It just dims it down due to averaging of different phase angles of the same frequencies. It may be useful to calibrate for a more uniform bass response across the room but nothing that cannot be achieved with multiple mic position measurements aligned with cross correlation.

Near field woofer response measurements are only useful for port/crossover phase correction. You want to correct for the room reflections below the transition frequency.

On the contrary, HF correction should leave out any reflections which can be done efficiently by decreasing the right window size of the measurement to a few milliseconds (just enough to eliminate first reflection - usually from the floor) at MLP.