Can you help interpret these?

[synapse-md]

Hello, everyone --

I'm new to this and have been trying to understand my room's acoustics. I was hoping you guys might be able to help me.

For measurement, I'm using REW V5.20 beta 47 and a UMIK-1. The computer is connected directly to the MIC and to my KEF LS50Ws via USB. The measurements are all done at 1M/96kHz, for a 20Hz-20kHz sweep around 10 seconds. The volume is set to provide a 75dB output of random pink noise.

An MDAT is attached with the following measurements (excerpted from an afternoon of testing!):
1. L+R Standard - The main speakers alone at "Standard" bass extension
2. 80Hz Standard HP - The main speakers with a 80Hz high-pass filter engaged
3. Sub 40deg no EQ - The main speakers plus a subwoofer, with the sub signal low-pass filtered at 80Hz. The main speaker HP filter remains at 80Hz and the bass extension has been reduced to "Less," which made a slight improvement in the integration.
4. Sub 40deq EQ - as #3, with two parametric filters engaged on the sub, computed by REW from the #3 measurement.

I have a few sets of concerns:
1. The nulls at 94Hz, 134Hz, and 229Hz, and the peak at 192Hz. All of which are quite invariant over the different setups tested.
2. The peaks at 38Hz and 70Hz, which are substantially smoothed by the subwoofer but not absent.
3. The general underlying choppiness of the response below 1.5kHz.

There does seem to be a dip in the group delay at 134Hz, but I'm not sure what this means or how to fix it.

The subwoofer is (for now) an SVS SB-3000. The room is 15'10" from front to back wall, 8'10" in height, and quite variable in width (11.5-16.5') due to a bay window on the right and an open hallway and stairwell at the front left.

The listening position is about 2.5 ft from the back wall.

Thanks in advance,
Neil.

 

[John Mulcahy]

The UMIK is a 48 kHz device, so no point in doing measurements at 96 kHz, the OS will be resampling the data. 48 kHz and 256k sweeps are fine for acoustic measurements. Other than that the measurements look fine. The level of variation is fairly typical. If you have any flexibility in your listening or sub position worth trying other positions for each. A quick way is to play the Pink periodic noise test signal and watch the response on the RTA as you move things around.

 

[synapse-md]

Thanks, John. I was using 96kHz thinking that, perhaps, the waveforms produced by the speakers might be more accurate. Is there any issue with aliasing between the output and input sampling? I’ll try 48kHz/256k and see what they look like.

The room does sound a bit lean and I wonder if this is because of all of those dips in the 90-250Hz range. The room is as yet untreated. I wonder whether bass traps (like the ASC tube traps or a Psi Audio AVAA) might correct this?