TOPIC:

I've had a UMIK-1 for a few months, and just recently bought a UMIK-2.  Using the miniDSP-supplied cal files in REW, I get obviously different measurements between the two mics. As a result, I don't know if I can trust the accuracy of either mic.  I filed a support ticket on 9 January, but I've yet to receive any response from miniDSP.  Here's an example measurement comparison, both using 48 kHz with a 1M sweep.  You can see clear differences above 2 kHz.  The UMIK-2 measures the same at 48 kHz and 96 kHz.
 

Did you definitely have the correct calibration file for the orientation of the microphone in each case? Getting this wrong for one of the microphones could cause this sort of difference I think.

How did you position your microphone for each measurement? Just asking because the highest frequencies will be most sensitive to small differences in where the microphone is located. 

Based on another thread it might also be that your signal level is a bit low for the UMIK-2. Are your results repeatable?

If you upload the mdat file I'll have a quick look in case anything jumps out at me. 

(I'm a bit surprised that you haven't had a response from Support yet. There have been a couple of people to post with issues where support requests apparently haven't been getting through. Hopefully someone from @devteam will reply to this but if you don't hear back soon you may want to try messaging them again - possibly with a different internet browser.)
 

Yes, I re-downloaded the cal files from miniDSP, and verified the files I have are the correct ones, and yes, I'm using the 90deg file in both cases.  The mic stand remains fixed in place, I simply switch which mic is placed in the holder, with the mic pointed at the ceiling.  Yes, results are repeatable.  I've just taken a new pair of measurements at ~5 dB louder, and the differences are just as obvious  I can't upload the mdat file as it exceeds the 2MB file size limit.  As to the support ticket, right after submitting it I received email from miniDSP acknowledging receipt and stating that a ticket had been created, so lack of response is not because it wasn't received.

If you wanted to share your existing data file you could do so via a file sharing website (e.g. dropbox). You'd also have a smaller data file to upload if you used a shorter measurement sweep.

Is your result for a single loudspeaker or two running at the same time? Where was the measurement made in relation to the speaker?

(Your response looks a little unusual to me in the 100-300 Hz range so just trying to get my head around what may be going on.)

FWIW I typically run sweeps at ~75 dB using a UMIK-1 but dropping 5 or 10 dB shouldn't really change much other than the SNR (also affected by your local background noise levels). I've not used a UMIK-2 myself to comment specifically there.

Yes, there are acoustic issues to look at below 600 Hz, but they are unrelated to the measurement differences, especially since the two mics agree in that range.  The measurements are from a 3-driver channel of a soundbar, with ~120 Hz crossover to a sub, in an untreated room.  Mic is about 7 ft away, at MLP.  65 vs 70 vs 75 dB makes no real difference (nor should it, really).  To make upload work, I've redone a pair of ~75 dB measurements with a shortened sweep and (just to focus on the issue) shorted the range to 1.5 kHz to 20 kHz, and put them in separate files.

Nothing obvious jumps out at me there.

I doubt it but just to rule it out, you haven't got the foam wind screen on one or both of the microphones have you?

Hopefully @devteam will reply to this thread soon if they don't contact you directly. If you're confident the ends of the microphone are in the same position and you aren't using the wind screens then this looks like something you'll need their input to resolve.

It's generally easier to look for anomalies in 'normal' measurements by the way, rather than unusually band-limited ones. I mentioned the 100-300 Hz features as when I looked quickly earlier I was vaguely wondering if they were partly reflecting a measurement artefact rather than a real acoustic effect, although on reflection I can definitely believe they're real. I wasn't criticising anything in case you thought I was  .

 

No, I'm not using the wind screens, and yes, the mic position is the same for both measurements.  Thanks for your responses.

TechSupport has finally reached out to me about my ticket, so I'll report back at a later date with any conclusion.

So it's been a week now since I sent the requested data to TechSupport, and I've received no further response yet. 

thothsong wrote:

So it's been a week now since I sent the requested data to TechSupport, and I've received no further response yet. 

Bear in mind most (all?) staff will have had time off for the Chinese New Year period.

Chinese New Year is no longer an excuse.

I had somehow the same issue but with the UMIK-2 microphone.
I pre-ordered the UMIK-2 and it was delivered right after getting public. I also made measurements to check any differences to the UMIK-1.
No significant differences between the UMIK-1 and UMIK-2. After I got the information, that miniDSP had an calibration issue, I downloaded new calibration files for my UMIK-2 and the measurements showed almost the same difference (about 5dB more @ 5kHz).

www.minidsp.com/media/kunena/attachments/43519/calfiles.jpg

That's very curious.