UMIK-1 vs UMIK-2

[Sonnie]

I know there are posts about this in some other threads that are not specifically dedicated to this very subject, so I am posting it here to keep up with it. I have sent the info to miniDSP support.

I am getting two totally different measurements with the UMIK-2 vs the UMIK-1. I noticed after using the UMIK-2 with REW that the response curves were looking suspicious, and thought maybe the firmware had not been updated. I downloaded it and updated it to be sure. As you can see below, the bottom and top end is significantly different. I certainly don’t trust what I am seeing with the UMIK-2 and it give

Below are comparisons of what Dirac measures with the UMIK-1 vs the UMIK-2 … and what REW measures after Dirac calibration using each mic with their respective calibration files.


UMIK-1 in Dirac

Here is UMIK-2

Here is the response I am measuring for the Dirac calibrated response with the UMIK-2 … look at 5kHz and up… not correct by any means… the cal file is show above the measurement.

Here is what it looks like with the UMIK-1… which I trust, as it looks correct.

 

[Mike-48]

I read somewhere that versions below Dirac 3 do not support the UMIK-2 correctly. Don't know if that applies to you, Sonnie.

 

[Sonnie]

I read somewhere that versions below Dirac 3 do not support the UMIK-2 correctly. Don't know if that applies to you, Sonnie.

It shouldn't, I'm using the latest version 3.0.14.

I'm gonna give it another go in the next few days to make sure I didn't have something set wonky.

 

[JStewart]

It’s interesting the SPL differences almost appear to be the inverse of the umik 2 cal file plot.

 

[Sonnie]

I noticed one time I got different results and the tilt was not quite as bad. It may be something to do with the resolution settings.

I'm going to do some more testing for miniDSP to try to figure it out. Apparently they are still seeing issues.

 

[Focus Fidelity]

I noticed one time I got different results and the tilt was not quite as bad. It may be something to do with the resolution settings.

I'm going to do some more testing for miniDSP to try to figure it out. Apparently they are still seeing issues.

Hi Sonnie, are they still seeing issues beyond those described here? https://www.minidsp.com/forum/umik-questions/18099-umik-2-measuring-different-from-umik-1#52094

 

[Sonnie]

Hi Sonnie, are they still seeing issues beyond those described here? https://www.minidsp.com/forum/umik-questions/18099-umik-2-measuring-different-from-umik-1#52094

Apparently I had some setting off somewhere. I stepped thru the UMIK-1 this afternoon, then swapped over to the UMIK-2 and it worked fine. I fiddled with it for several hours the other night and never could get it working like it should. I'm just not sure what I missed, UNLESS, I may have had double calibration files loaded in REW. This time I cleared everything out, made sure I had all the settings correct... ran a bunch of sweeps, both mic with their cal files matched up perfectly, so it was all me.

 

[AustinJerry]

Glad you figured it out. When i received my UMIK-2, I ran extensive comparisons with my UMIK-1, both using CSL custom calibration files. I saw no significant differences. As advertised, the UMIK-2 seems to be slightly more sensitive.

 

[NBPK402]

Was it worth upgrading to?

 

[AustinJerry]

Was it worth upgrading to?

To satisfy my curiosity, yes. To achieve measurably better results, no.

 

[Sonnie]

Yeah... I'm not sure I'd ever be able to note any differences between using either mic.

 

[thothsong]

I got a UMIK-2 today, and of course tried comparing it with my UMIK-1, and I'm getting obvious differences from about 2 kHz up, just doing simple frequency sweeps. So like the guy with two watches, I don't know which one to trust. Here's an example of sequential measurements, with only one mic physically connected at a time.

 

[AustinJerry]

I got a UMIK-2 today, and of course tried comparing it with my UMIK-1, and I'm getting obvious differences from about 2 kHz up, just doing simple frequency sweeps. So like the guy with two watches, I don't know which one to trust. Here's an example of sequential measurements, with only one mic physically connected at a time.

Based on that measurement, you have bigger things to worry about than the minor differences between the two mics. On the positive side, in the area that you should be working on below 500Hz, the mics measure essentially the same.

 

[thothsong]

Sure, but beside the point.

 

[AustinJerry]

Sure, but beside the point.

I have both mics as well, and each has a custom calibration from Cross Spectrum Labs. As recently as earlier this week, I compared measurements using both mics. Here is my result:

I am not seeing the difference in the high end that you are seeing, so something may be amiss with your measurements. If you are sure that the calibration files you are using are correct, I wouldn't worry about the differences that you see. The point of my previous post is that the top end doesn't need work, regardless of the differences in the measurements that you are observing. The low end, on the other hand, looks pretty bad. You can work on improving the low end without worrying about the differences between the mics.

Does that make sense to you? I assume you see the difference between your low end and my low end.

Edit: With the UMIK-2, are you measuring at 96KHz and using a full 1MB sweep length? If not, maybe that is something to try.

 

[thothsong]

Sure, I understood you the first time. I'm sure the cal files are as downloaded from the miniDSP site. I've opened a support ticket with miniDSP to see what they have to say.

 

[thothsong]

I measured both at 48 kHz, 1M sweep.

 

[AustinJerry]

I measured both at 48 kHz, 1M sweep.

Can't hurt to try the UMIK-2 @96KHz to see if the measurement is any different.

 

[thothsong]

Thanks for the suggestion, but it measures the same at 96 kHz.

 

[Mike-48]

@thothsong - Are you using the same type of calibration file for each mic? Don't the straight and 90-degree files differ mostly in the high frequencies? Just a guess of what might cause such a thing.

 

[thothsong]

Yes, I'm selecting the 90 degree cal file in both cases. Mic stand stays fixed in placed, just replacing which mic is put there.

 

[Focus Fidelity]

Hi @AustinJerry How do yours compare if you use the MiniDSP calibration files instead of the cross-spectrum ones?

 

[AustinJerry]

Hi @AustinJerry How do yours compare if you use the MiniDSP calibration files instead of the cross-spectrum ones?

I don't have immediate access to an actual measurement showing a comparison, but here is a comparison of the actual calibration files. The legend at the bottom identifies which curve is which. IMO, the calibration files from the MiniDSP web site are quite close to the CSL files. OTOH, the two mics have significantly difference calibrations, which is probably to be expected. This comparison does not give me a strong feeling that either mic is more accurate than the other, and actual in-room measurements show that they measure quite close to each other. If I need to use a mic, a flip of the coin would determine which mic I would use.

 

[Mike-48]

IMO, the calibration files from the MiniDSP web site are quite close to the CSL files.

In looking at your informative graphs, that wouldn't be my opinion. Yes, from 100 Hz to 3 kHz, the two calibrations are quite close, but the differences above and below are pretty large. In my experience, 4 dB at 8 kHz or 30 Hz is enough to change the character of the sound quite significantly. I've been using DSP in my audio systems for almost 20 years, and I've always have suspected a serious weak point is the uncertain accuracy of the mic calibrations. Your graphs are a good example of that. Thank you for posting them!

 

[AustinJerry]

In looking at your informative graphs, that wouldn't be my opinion. Yes, from 100 Hz to 3 kHz, the two calibrations are quite close, but the differences above and below are pretty large. In my experience, 4 dB at 8 kHz or 30 Hz is enough to change the character of the sound quite significantly. I've been using DSP in my audio systems for almost 20 years, and I've always have suspected a serious weak point is the uncertain accuracy of the mic calibrations. Your graphs are a good example of that. Thank you for posting them!

Question is, which is more accurate? I would be inclined to say the CSL calibration files. Does it really make a difference? Here is a measurement comparing the two mics. Not much difference above 20Hz, and below 20Hz I don't care.