Saturday, 24 January 2026

miniDSP UMIK-2 USB Microphone: analog gain & optimizing self-noise. Plus audiophile gear pictures from "No Other Choice".


Last week, I discussed the miniDSP UMIK-2 USB measurement microphone and as I mentioned then, it's not often that we see much discussion about these devices among audiophiles even though they are important in helping us chase down sonic problems and of course to calibrate the sound system. To me, that calibration process is an essential part of achieving a truly "accurate" sound system once we sort out basic room set-up; assuming high-fidelity accuracy is what you're after of course!

So, before heading off to other topics, I thought I'd explore a little more about this device.

The first thing I want to examine is the analogue gain adjustment using miniDSP's software "miniDSP-MIK".

As mentioned previously, you'll need to contact miniDSP directly if you want a copy of this program as the agreement is not to distribute as an end-user.

Here's what the program looks like when connected with the UMIK-2 at default setting: 


You start the program with the UMIK connected, then press the top right "Connect" button and the analog gain section will reflect the current setting when it says "Connected" with the checkmark.

From here, just click on the +18dB value and change it with up/down arrows between the 0-42dB range. As suggested before, unless you're planning to record very loud content >100dB SPL like in a concert or want to push your speakers to their upper limits to check for distortion, there's little need to change from the default.

For illustration, let me drop it to +12dB which is a reasonable setting and I believe in the past, some original UMIK-1 microphones might have had this as default:

The program will automatically save this gain setting to the microphone when you exit the program.

This change in the analogue hardware gain will need to be reflected in the calibration data so software like REW is able to compensate for the level. We do this by editing the downloaded calibration file from miniDSP; specifically the "Sens Factor" (sensitivity factor) to reflect the -6dB reduction from default, and we can change "AGain" (analog gain) to reflect this value as well:

Obviously, make sure to always keep a copy of the default file. I'll save the adjusted calibration file with an appropriate filename like "UMIK-2 *******_90deg [+12dB Gain]" (the *** portion is the serial number as per the usual miniDSP naming scheme).

Point REW to this calibration file in the "Cal files" preferences tab and we can now run another sweep with the microphone at the same location in my room to check that the adjusted file did the job.

Let's grab a few frequency responses with gain set to 0dB, +6dB, +12dB, and the +18dB default:

As you can see, the gain-adjusted calibration worked as expected with levels overlaid almost exactly. Unlike digital "Bits Are Bits" accuracy and basically identical results which can be seen when comparing bit-perfect devices, whenever we switch a thing or two in the analogue domain, it's worth checking for slight variations. At the ~80dB SPL frequency sweep, clearly the change in analog gain does not impact the result from the UMIK-2; this is excellent.

I'm tempted to keep the device at +12dB gain just to allow a little more overhead and a slightly lower analogue gain might improve low-level self-noise. If the software you're using doesn't read the sensitivity calibration, it will interpret the input level lower than default so just keep an eye on that. 

Speaking then of self-noise, let's explore this a little further. As you can see in the previous post, if I wrapped the microphone up with multiple layers of blankets on a very quiet evening, I can measure the limits of the microphone's self noise and plot out the peaks and valleys to show that some of this is likely a result of mains hum seeping through the USB port connected to the computer.

As such, we should be able to lower that component of self-noise with a USB isolator and a quiet lithium battery power source, right?

Here's what I did:


Since I didn't feel like making an anechoic chamber 😉, at the bottom-left of the image in yellow is the UMIK-2 cozily swaddled up with a couple of blankets in layers which should result in a very quiet environment. And on the carpet we see the Intona 7055-C USB 3.0 isolator and beside it with the LED lit, a typical USB power bank feeding the isolator.

So let's compare the noise level either with the UMIK-2 directly connected to the computer USB port, or through that USB isolator powered with the Li battery. Does doing this improve the UMIK-2's self-noise?


First, notice that I measured my room's ambient noise shown in dashed red on the graph. This was done in the daytime (around noon) resulting in the ambient noise around 26dB(A) - a little higher than evenings where it drops to below 25dB(A).

In blue I show the spectrum of the UMIK-2's self-noise as it's wrapped up in the blankets. Notice where the electrical noise from the USB port approaches the room ambient noise, specifically at 60Hz. Thankfully at that frequency, the A-weighting curve is around -25dB so it doesn't affect the calculated value too much.

In green then, by using the Intona USB isolator with a quiet lithium battery, we see that indeed, the UMIK-2 self-noise level dropped further. We can clearly see that with the Intona + battery power, the 60Hz hum has been attenuated by a good 6dB.

The result with USB isolator + battery is a measured microphone noise floor dropping from 23.7dB(A) down to 22.2dB(A) (averaged over 8 samples). While I did not test with +6dB and 0dB gain settings, I suspect it's possible to achieve even lower readings down to maybe 20dB(A) based on some of the miniDSP specs I've seen. Obviously, I'm not going to be doing this in day-to-day measurements with REW or with Dirac Live, but it's interesting to show the effect of the USB isolator and low-noise battery just in case anyone needs to aim for the lowest self-noise possible from this USB microphone!
[I think the UMIK-2 self-noise level is pretty good and probably competitive with some more expensive measurement microphones + preamps one might consider. Here's a thread showing noise floor measurements of Earthworks mics with a comparison to UMIK-1 and UMIK-2 down the page - you might be surprised and possibly impressed by the UMIK-2 result!]
As a hobbyist who doesn't want to fiddle around with external microphone pre-amps, the UMIK-2 and its predecessor UMIK-1 all-in-one USB measurements microphones really are very convenient. Now with the UMIK-2's lower noise and higher samplerate abilities, this will expand the range of potential acoustic measurements we might want to try.

With that, I think I've satisfied my curiosity around the UMIK-2 microphone to an adequate degree and will start using it more ahead as my go-to tool. đŸ˜Š

--------------------

Despite a busy work week, I was able to catch a showing of the recent Korean movie No Other Choice (어쩔수가없다) (2025) directed by Park Chan-wook the other night. It was nominated in the Golden Globes last week but surprisingly not nominated for the Oscars. In the movie, there's an "all analogue" audiophile character with a few scenes showing his vinyl system.

It's always fun to see our hobby represented in movies (like this one, my favourite). Here are some screenshots:



Look at the ceiling room treatments! This guy's the real deal.

And side wall diffusers too.

I suspect the vintage hardware folks will be able to identify the turntable and some of the gear beyond the obvious McIntosh 275. It's an interesting movie to check out if you're into dark humor and satire related to the extremes one might consider during desperate times, channeling societal worries around mechanization and human labor as AI capabilities increase (along with the growth of personal debt worldwide).

To end, here's a fun song used in the movie during the humorous extended scene taking place in the sound room - Cho Young-pil's "Redpepper Dragonfly", circa 1981. A notable bit of BeeGees-esque falsetto to the singing:


As usual, beyond the hardware and all this tech-geekery, I hope you're all enjoying the music, dear audiophiles!

5 comments:

  1. Sorry but I have nothing to say about those measurements. Me and measurements are poles apart anyway.

    But I have something important to say… Are you sure those things on the ceiling are real diffusers? They look more like fancy ceiling decoration to me. With such low ceiling, even real quadratic diffusers cannot work properly—need at least 8 feet or more height to scatter sound nicely. Interesting how Hollywood and Korean movies always put valve amps, turntables, and ‘audiophile’ stuff for cool feeling. Just for laughs!

    ReplyDelete
    Replies
    1. Oh man ST,
      Measurements are cool. Play with it if you have time. 😉

      Dunno man about whether the ceiling was meant to be diffusers or decorations. But the fact that they're up there looking like diffusers is pretty cool! Of course whether they work is a matter of measurements perhaps!

      Yeah, the "audiophile" caricature these days is often represented as the older dude with the valve amps, turntables, and such. It's silly and funny. Would be interesting to see if this shifts and morphs over time.

      Delete
  2. Hej Arch,
    Interesting. I have not measured my room using a dedicated microphone and software such as presented here. I am intrigued and this will be next on my shopping list for 2026.
    This week I was made aware of MQAs attempt at a new comeback as reviewed by Golden Sound on the headphone show. He also references your blog and review of MQA. “Is MQA back from the dead?”
    https://youtu.be/-AgyddT1ytU?si=MuF0GS2MS86rft2B
    My conclusion is that this is the same nonsense but with a different name.
    I came across another video which might be of interest. “Why don’t movies look like movies anymore?” https://youtu.be/EwTUM9cFeSo?si=m5bCtEod3ZeKcR8L
    Cheers
    Mike

    ReplyDelete
    Replies
    1. Thanks for the note Mike,
      Yeah, have a look at your room measurements this year. Enjoy!

      Yeesh, MQA. Like herpes, the audiophile gift that nobody really needed (or wanted) that keeps on giving. 🤣 Until Lenbrook actually releases their streaming service (wasting corporate dollars on something that almost definitely will fail), there's absolutely nothing new and I don't see how any audiophile can find what they're doing in any way exciting.

      This whole "new" MQAd2a is yet another way to repackage the poor quality minimal phase filter that Bob Stuart wants us all to believe makes a meaningful difference. Such nonsense. Anyhow, I put in an addendum in the Lenbrook MQA article last year reflecting what I thought was the most important finding in the new video.

      Interesting video on why movies look different these days. Hmmm, looks like he prefers a more contrasty, maybe more saturated look to his movies? I agree that when not done well, VFX and green screen can look bad (I'm reminded of the sterile look of Star Wars Episodes 1-3 in the early 2000's). But pumping up contrast can be done with good HDR films if the moviemakers want that look.

      Delete
    2. hi Mike5959,

      I watched the video titled why films don't look like films anymore, and it's all simply a matter of opinions. He clearly has an absolute opinion on what he likes. I don't understand what his point is. Was it just to simply say, I like how old analogue cameras look?

      Also, he has used a fairly random set of examples, seemingly cherry-picked from his personal favourites, then compared them against mainstream film examples designed to make money, not designed for creativity and art, especially using the children’s film Wicked as an supposed example of a bad-looking film.

      It's a strange video.

      Delete