Strange REW measurements

[polmuaddib]

Hi, i have bought Sonarworks xref20 microphone and downloaded a calibration file for my mic from their website. I use Stainberg ur22 mkII interface on Windows 10 and REW software.
For a while I started measuring my front speakers. The reason for measurements is to find the best possible location for them and myself in the room. Also, I have two subwoofers that i use for stereo and HT. They are connected by high level inputs with front speakers, so they are always on when i play music.
My room is 5.1m by 4.7m (but one side wall is 4.5m and the other is 4.9m, so not a perfect square, but very close, unfortunately). Ceiling is 3.15m (originally 3.3m, but i lowered it with drywall and placed 4 Atmos speakers there.
So, since i got my mic, i have been getting the same consistent results, whatever i do. I have changed speaker and ML positions (+ or - 30 cm, which is a lot for my room) in many configurations and i have even bought a bigger carpet, I have changed crossover settings for subs and i have changed the measuring mic position from on axis to 30 degrees of axis. And the results are almost always the same. I have an enormous dip from 2 to 5 khz. And it is the same for both speakers. And this is more then 20 db abbyss.
Can microphone be wrong? It is like my midrange drivers are off. I would understand one, but both??? I bought speakers used. And I hear them just fine.
I tried archiving mdat file to zip, but ASR will not accept the file, for some reason.... it starts uploading, but fails... ( I only have RAR and 7zip, but i choose ZIP for format..)
Thanks for suggestions.

 

[pozz]

What kind of speakers do you have? Are you measuring using sweeps? It might just be an artifact of where you place the microphone.

Try the MMM method:

• http://www.ohl.to/audio/downloads/MMM-moving-mic-measurement.pdf
• https://www.audiosciencereview.com/...ving-microphone-method-help.12641/post-374973

Or it could be an incorrect calibration file.

 

[AnalogSteph]

What's the MIX control on the Steinberg set to? If I understand the monitoring correctly, this should be turned all the way to DAW in loopback testing.

I'll second turning off the calibration for a test. In the frequency range the deviation would be quite small anyway.

 

[polmuaddib]

Thanks. They are a three way floorstanders, Avalon opus ceramique. I measured in REW doing sweeps from 20hz to 20khz. Those are measurements from different mic positions. That is why i am puzzled...

 

[polmuaddib]

Steinberg is set to input. But I am not using Steinberg for output. Only input. I use my DAC for output. I will try Steinberg only. Thanks

 

[polmuaddib]

I used Steinberg for input and output. But i set the mix to middle between input and DAW. I will try later all the way to DAW. Much better now.

 

[tuga]

I used Steinberg for input and output. But i set the mix to middle between input and DAW. I will try later all the way to DAW. Much better now.

Not really much better...
It would help if you kept a 50dB-wide vertical scale consistent.

Are you measuring both speakers playing at the same time?

Are you using any kind of weird immersive 3D processing?

 

[JohnPM]

I used Steinberg for input and output. But i set the mix to middle between input and DAW. I will try later all the way to DAW. Much better now.

The mix knob has to be all the way to DAW. The input must not be routed back to the output.

 

[daftcombo]

Do you measure each one at a time or both playing together? If the latter, it would be normal as they most probably cancel each other at those frequencies.

 

[Asinus]

long shot but did you by any chance disconnected the speakers when you moved them? and have you checked the polarity on the cables? I was once getting similar measurements from the rear channels (a dip in the mid-highs) and it turned out the bananas were reversed.

 

[polmuaddib]

In the last measurement i posted, i measured only one speaker from 1m distance on axis, tweeter level approximately. And mostly I measured that way. One speaker at a time, seldom both, just to try.
Only one channel playing, no 3d effects. After that, i turned mix to DAW all the way per advice. Didn't save it, but i will repeat it tomorrow. When I turned all the way to DAW, i just got louder measurments, but the same line...
And, no, Asinus, polarity is fine.
Now, Tuga, I know it is not perfect, but i meant that when i used Steinberg for input and output, I got at least a logical result. What do you think it is more? Speakers or room. My room is untreated except for furniture and i moved speakers closer to the wall to get more low end. Is this zoom in better?
Thank you all so much.

 

[RayDunzl]

I tried archiving mdat file to zip, but ASR will not accept the file, for some reason....

I think there's a 2 megabyte limit, if that applies to your problem.

REW can open multiple measurement files at a time to reconstitute the measures you want to share.

 

[polmuaddib]

You are right. I didn't realise those mdat files are big. Thanks