Confused about microphone cal and graph interpretation

[Gregory Livingstone]

Is the resulting frequency graph supposed to be adjusted by the microphone calibration data. I thought it would be, but now I'm confused. I tried my set up with a different microphone and calibration file. It shows a deeper frequency response.

I have been using REW with Dayton microphone & sound card. Today I also did the same measurements using Behringer microphone from SMS-1 kit. The measurements are both of a 2.1 system using a Dirac target filter lowered to put the lowest frequencies at 0DB. The only change is this microphone and the mic calibration file.

Looking at the calibration files the Dayton file's first entry is at 20 hz at 8.9 as the microphone calibration trace shows. The Behringer file starts at 10hz at -3.61051. If the frequency response trace does include the mic cal then one of the microphones is wrong or has an incorrect calibration file. For instance if the Dayton is flatter at the lower frequencies than the cal file indicates, it would falsely show roll off starting at 26 hz (not sure what happens below 20 hz).

Any thoughts as to how to confirm that one mic is more accurate than the other? I want to think that the Behringer is the more accurate, There are dual SVS SB3000 subwoofers on the .1 channel which should be better than -10dB at 20hz.

I am willing to get the recommended usb microphone, but think I still need a sound card as HDMI isn't a viable option. I presume using the laptop stereo out is a bad idea.If I go that route is there a sound card recommendation?

 

[John Mulcahy]

The mics are only as accurate as their individual cal files, which are in turn only as accurate as the process and references used to generate them. That Dayton file looks pretty odd at the low end though.

There's nothing wrong with using the laptop output for acoustic measurements.

 

[Gregory Livingstone]

I bought a Umik-1 and provide 2 comparisons below, The Unik traces are red and the REW SPL was between 74.7 & 75.2 dB when setting level volume using full b/w pink noise. The earlier measures were also intended to be at 75dB, but I am less sure of accuracy. The First is w/o filters & 2nd with Dirac Full Live target filters.

Going on assumption that Umik is most accurate, the Dayton bass is definitely wrong. I redid getting the cal file and it had same high.offsets at low frequencies. I also looked at a cal for mic with next serial in line and that was flatter (less offset) for low frequencies. For the Behringer or Velodyne SMS mic I hav a cal file, but am clueless as to how I got it.

The Umik trace shows the high end is flatter. The speakers are B&W 704 and I thought their tweeters were suppose to go high. At my age not too concerned about that and doubt I hear it.

What I find puzzling is the Umik differences between 60-80 Hz and 200-300 Hz. While there are differences at low and high ends for the mics. The Behringer and Dayton follow similar paths as does the Umik, except at the noted ranges. There the Umik is lower than the other two by 10dB and appear to be opposing dips and peaks. Could that be from the mic positioning being different by a little (2 inches is my guess at variance though I think I am spot on)

Given that my set up to date was based on the Dayton which I think is wrong in the bass, I think I need to redo my sub B phase change made to align with sub A thinking it might be incorrect. The Dirac Live measure for the sub only also shows the peak at 20Hz and then a slope down to 55 Hz which I didn't see in my sub only measurements with REW/Dayton.

 

[roger2]

Which Dayton mic are you testing?

 

[Gregory Livingstone]

It was an EMM6