In room REW measurement for L+R channel help

[killitmore]

Hi guys,

I measured my room for the first time I used REW and measure both L+R channel. THe result was that my untreated room is too good to be true (only +-2db from 60Hz to 20khz).
I wonder if i did some thing wrong or my tools are broken. I followed the instruction on REW website. Loaded my sound card calibration file, no mic calibration file (the instruction stated that i can skip that file if the file is not available), calibrated input level to 80db
Tools i used: Icon upod pro soundcard, Mic isk at 100 (these tools are popular in my country)
Belows are some screenshots of my result.

 

[RayDunzl]

It does look a bit unusual to me.

REW creates a .mdat file.

For a single measurement, it should be small enough to post here.

Compress it and you can attach it as a .zip file here.

 

[killitmore]

@RayDunzl thank you very much. Please view the attached mdat file

 

[dasdoing]

a cardioid microphone isn't really usefull for meassuring. it will atenuate the reflections. is this a nearfield measurement?

 

[Jimi Floyd]

Something is wrong. I never saw any room with a frequency response remotely as flat as your result. It is physically impossible. I browsed your mdat file and there is no resonance and no deep at any frequency. It can't be, you probably selected the line out and not the microphone as input or something like that. Sorry.

 

[staticV3]

for an untreated room there's an unbelievable lack or room modes in your measurement:

not sure what might be the cause though.

I'd try repeating the measurement, but this time using the Moving Microphone Method.

Here is the PDF describing MMM: https://www.ohl.to/audio/downloads/MMM-moving-mic-measurement.pdf
Here are some videos showing MMM in action: https://youtu.be/6RiuwqzjqlQ https://youtu.be/SXWfJrl8aBw
Here are the relevant settings for REW: https://cdn.discordapp.com/attachments/860982405074649089/1041416797070557264/image.png

 

[Jimi Floyd]

for an untreated room there's an unbelievable lack or room modes in your measurement:
View attachment 294019
not sure what might be the cause though.

I'd try repeating the measurement, but this time using the Moving Microphone Method.

Here is the PDF describing MMM: https://www.ohl.to/audio/downloads/MMM-moving-mic-measurement.pdf
Here are some videos showing MMM in action: https://youtu.be/6RiuwqzjqlQ https://youtu.be/SXWfJrl8aBw
Here are the relevant settings for REW: https://cdn.discordapp.com/attachments/860982405074649089/1041416797070557264/image.png

It's not the microphone placement method, there is some more fundamental mistake going on

 

[staticV3]

It's not the microphone placement method, there is some more fundamental mistake going on

MMM has both different mic placement and a fundamentally different data acquisition method in REW. Two birds with one stone was my thought.
Better than trying to troubleshoot REW's default sweep IMO (especially since MMM is more accurate anyway).

 

[RayDunzl]

It would appear the measurement starts out being an electrical measure, then transitions to acoustic, which is very odd.

The impulse response appears to be electrical:

And so does the Step Response - but then it transitions to what "could" be a microphone/speaker response:

So, whatever is going on, it is not providing a usable acoustic measurement.

Try again.

 

[staticV3]

It would appear the measurement starts out being an electrical measure, then transitions to acoustic, which is very odd.

Maybe heavy crosstalk from analog out to mic in?

 

[dasdoing]

I think I know what happend. If you look at the soundcard correction it seams to overcompensate.
what probably happend is that he calibrated his soundcard, and didn't remove the loopback when meassuring the room. so this is soundcard and room combined.
something like that

 

[killitmore]

Thank you all for your help. Basically my measurement is not reliable, I will definitely re-do it by this weekend and update the results.

 

[sam_adams]

You over-smoothed the measurement and compressed the vertical scale to 3 dB. This is what the measurement really looks like:

It's largely nothing but noise. Some setting in your external soundcard is wrong—probably some monitoring feedback.

 

[killitmore]

Thank you all so much. I measured again today. This looks much more reliable but still need your help to review it once again.
If this is still wrong, i probably have to use another soundcard for the measurement this soundcard i borrowed from my friend, he uses this for karaoke at his hone
====
Edit: Just going through the whole manual for setting on the sound card again. This measurement is likely wrong again. Will perform another measurement by weekend

 

[Rednaxela]

You didn’t by chance record your laptop mic?

 

[killitmore]

You didn’t by chance record your laptop mic?

No. 100% sure not use my laptop mic

 

[killitmore]

Just going through the whole manual for setting on the sound card again. This measurement is likely wrong again. Will perform another measurement by weekend

 

[Jimi Floyd]

This new one looks reasonable. Are your speakers filtered with a high-pass filter?

 

[RayDunzl]

This new one looks reasonable. Are your speakers filtered with a high-pass filter?

I don't think so.

I measured again today. This looks much more reliable but still need your help to review it once again.

The data recorded makes no sense. Maybe just noise.

Attached is a valid measurement for you to compare.

Drag the unzipped file onto your REW. or use File/Open Measurement

Use the Overlays window to compare data.

 

[RayDunzl]

You can monitor what REW is seeing real time with the RTA window.

Here, background noise and a soft sine tone from the Generator window.

The red button top right starts and stops the monitor.