First REW measurements - what am I doing wrong?

[Sagnet]

Ran my first 0-200 Hz sweeps today, and got this curve. What am I doing wrong...? It peaks out at 152 dB.

Mic used is the UMIK-1.

 

[Sagnet]

Here is the same curve, more compressed.

 

[Sagnet]

I think the problem must be that the mic is too sensitive. Is it possible that the mic is broken? It is five years old, and has for some time been stored at sub zero temperatures (celcius). Could that maybe be a cause?

 

[raindance]

What does the rest of the freq range look like? Do a full range measurement.

 

[Sagnet]

The problem seems to be that the UMIK-1 was not configured properly, so it was the onbord mic on my laptop doing the readings.

Have installed ASIO4ALL now, but the UMIK-1 does not show up there. Scratching my head at the moment, trying to figure out how to get the UMIK-1 connected to REW.

 

[raindance]

It is a setting in REW. Not sure ASIO4All has anything to do with it.

 

[RayDunzl]

Have installed ASIO4ALL now, but the UMIK-1 does not show up there. Scratching my head at the moment, trying to figure out how to get the UMIK-1 connected to REW.

Use JAVA in REW for the UMIK-1

Make sure to point REW to the calibration file. I renamed mine, so...

There may be other ways to set it up, but this is the most straightforward.

 

[daftcombo]

Did you look here already?
https://www.minidsp.com/applications/acoustic-measurements/umik-1-setup-with-rew

 

[Sagnet]

Figured out the mic. Was a problem with the cable not being properly connected. Ran a full sweep. Got this, looks better:

 

[Blumlein 88]

Usually in REW if you have the Umik attached it will be a drop down option for input. Open REW and look in Preferences. Look under Input Device.

Don't get too flustered, I remember messing around with REW and getting frustrated. Just kept at it until I figured it out. You'll get there and it is such useful software.

 

[Sagnet]

Not very impressive, but at least I got my first proper reading. Thanks guys.

 

[Erici]

Good job! Now it would be good to get separate readings for each speaker. Also, variable smoothing seems to provide reasonable detail without being too detailed to understand.

 

[Sagnet]

What should be the key takeaways from this graph? Is it normal that the response is so much lower from 100 Hz and onwards? Is it because the subs roll off from there?

The curve in post #11 is when running stereo with subs (2.1 channels). When running the same sweep on 5.1 channels ("multichannel stereo" on my AVR), the response from 100 Hz and onwards is almost 10 dB louder.

 

[RayDunzl]

When running the same sweep on 5.1 channels ("multichannel stereo" on my AVR), the response from 100 Hz and onwards is almost 10 dB louder.

Doubling the number of speakers, each playing at the same level, can add as much as 6dB to the sound level.

Left, Right, and Left + Right

Subtraction/cancellation can occur, too - phase problem at 48Hz above when both sides play simultaneously.

With your five speakers (not counting sub):

If each speaker plays the same level, say 85dB, and they hit the microphone with the same phase, the sum would be as much as

1 - 85dB
2 - 91.02dB
3 - 94.54dB
4 - 97.04
5 - 98.98dB

 

[Wayne A. Pflughaupt]

What should be the key takeaways from this graph? Is it normal that the response is so much lower from 100 Hz and onwards? Is it because the subs roll off from there?

The curve in post #11 is when running stereo with subs (2.1 channels). When running the same sweep on 5.1 channels ("multichannel stereo" on my AVR), the response from 100 Hz and onwards is almost 10 dB louder.

That's actually an excellent graph - most of us would kill for one that good. It is normal for the lows (subs) to be elevated compared to the mains.

Regards,
Wayne A. Pflughaupt

 

[Sagnet]

That's actually an excellent graph - most of us would kill for one that good. It is normal for the lows (subs) to be elevated compared to the mains.

Regards,
Wayne A. Pflughaupt

I see, that's good to hear. But is it normal that the difference is that large? Looking at the graph from post #11, it's almost 20 dB between the peaks at 40 and 58 Hz, and the dips at 800 and 1200 Hz.

 

[RayDunzl]

But is it normal that the difference is that large? Looking at the graph from post #11, it's almost 20 dB between the peaks at 40 and 58 Hz, and the dips at 800 and 1200 Hz.

In room, listening position measure, yes... Can be...

1/12th smoothing with my psycho system (no EQ) at the listening post (10 feet), no subs

 

[Wayne A. Pflughaupt]

I see, that's good to hear. But is it normal that the difference is that large? Looking at the graph from post #11, it's almost 20 dB between the peaks at 40 and 58 Hz, and the dips at 800 and 1200 Hz.

Apply 1/3-octave smoothing. That will get you a graph that looks more like what you're actually hearing.

Regards,
Wayne A. Pflughaupt

 

[Sagnet]

Apply 1/3-octave smoothing. That will get you a graph that looks more like what you're actually hearing.

Regards,
Wayne A. Pflughaupt

Thanks, will do. That graph is with 1/12 smoothing.

 

[Sagnet]

This is L+R+subs with 1/3 octave smoothing. Audyssey XT32 reference EQ applied. I appreciate that this is a pretty decent result.

If I want to improve this, what should I be looking at?