Measurement Microphones - what is the consensus?

[Blumlein 88]

How good do they have to be?

You are too good at asking just the right question sometimes. At a minimum they have to be better than your ears/brain at measuring frequency response. And this is a wide gulf between good enough to be useful and theoretical perfection. Maybe a case of don't let the perfect be the enemy of the good.

 

[No. 5]

To illustrate that, here is the deviation from linear of 250 UMIK-1's, as posted by Curvature earlier in this thread:

And here is the deviation from linear of an un-EQ'd room response with the same vertical scale:

And here is an approximation of the difference between the true in-room response and what the worst of the uncalibrated mic's from the above graphic would measure at low frequency:

Certainly not perfect, but in my experience, still very usable.

 

[FeddyLost]

Certainly not perfect, but in my experience, still very usable.

For personal home use - when we don't really need proper FR at LP and anyway will voluntarily EQ according to tastes - it's enough.
For professional duties of course it's a bad compromise.

 

[FeddyLost]

By the way need some advice regarding calibrated mic. It's cheap if it's not from Earthworks/B&K, so i decided to try it.

I've investigated some options available and found that I already have
1) ECM8000
2) KRK ERGO mic (without any calibration file, originally used by an interface only up to 500 Hz)
3) NOS usb ARC microphone with some Anthem calibration file .cal, that is supported only by ARC Genesis and I don't know how to use it standalone.

So, most of other options available at decent price now are either UMIK-1 (maybe with some minidsp gear) either Sonarworks mic with or without SW.
What is your experience with these? Is UMIK reliable as a standard measurement mic?

PS Or maybe converting .cal file for Anthem if that possible somehow. Did anyone tried?

 

[Blumlein 88]

By the way need some advice regarding calibrated mic. It's cheap if it's not from Earthworks/B&K, so i decided to try it.

I've investigated some options available and found that I already have
1) ECM8000
2) KRK ERGO mic (without any calibration file, originally used by an interface only up to 500 Hz)
3) NOS usb ARC microphone with some Anthem calibration file .cal, that is supported only by ARC Genesis and I don't know how to use it standalone.

So, most of other options available at decent price now are either UMIK-1 (maybe with some minidsp gear) either Sonarworks mic with or without SW.
What is your experience with these? Is UMIK reliable as a standard measurement mic?

PS Or maybe converting .cal file for Anthem if that possible somehow. Did anyone tried?

The Umik 1 is pretty good for most purposes and comes with a cal file. You can get one from Cross Spectrum for a few extra bucks and they do a more thorough calibration.

Cross·Spectrum - Calibrated MiniDSP UMIK-1 Microphones for Sale

Purchase calibrated MiniDSP UMIK-1 measurement microphones, low-cost option for performing precise and accurate sound measurements

cross-spectrum.com

Limitations are they have some noise at lower levels, but that is very rarely a problem the way they are used. The big convenience is they are pretty well accurate with calibration and so easy because you plug them into the USB on a laptop, use them with REW and they automatically read out SPL levels accurately without you having to bother with cross matching levels with something else. And you don't need a recording interface since they are USB microphones. If you were trying to investigate very low level sounds you'd be better off with a different mic that feeds an interface.

Now an ECM 8000 is pretty accurate and if you already have a recording interface to feed it into I'd just use that. Plenty good enough to start doing things.

 

[FeddyLost]

Plenty good enough to start doing things.

But not good enough to make these things correct. That's why I thought again about calibrated mic.

Maybe it will be the cheapest way of making sound more correct with D-EQ. Even if I don't like it, I'll be able to save preset.

 

[No. 5]

PS Or maybe converting .cal file for Anthem if that possible somehow.

If you open the .cal file in a text editor, what do the contents look like? My suspicion is that it's just a text file, in which case, it can be used with other software. And if you plug that ARC microphone into your computer, does it recognize it as a USB microphone?

But not good enough to make these things correct.

Question: how correct do you want things, and what things do you want corrected?

 

[FeddyLost]

My suspicion is that it's just a text file,

Depending on encoding maybe ...

29 188 bytes
I'm not a reverse engineer, so it's not easy for me.

And if you plug that ARC microphone into your computer, does it recognize it as a USB microphone?

After installation of ARC Genesis with all required drivers. It's a PBK-1 Paradigm USB audio. 1 channel up to 16 bit 48KHz.
USB\VID_1B35&PID_1003&REV_0100&MI_00
USB\VID_1B35&PID_1003&MI_00

how correct do you want things, and what things do you want corrected?

I'd like to know that FR that measured with mic in reality is close to its form.
Voluntary EQing system with possibility of such mic responce variance as posted before is ... less than pointless. Almost like applying soft parmetric EQ with random settings.
It's always nice to have some physical reference even if you don't like it "flat". Otherwise you'll never know what exactly you are listening to.

 

[Blumlein 88]

But not good enough to make these things correct. That's why I thought again about calibrated mic.

Maybe it will be the cheapest way of making sound more correct with D-EQ. Even if I don't like it, I'll be able to save preset.

If you lack confidence in what you have spend a few bucks on the Cross spectrum umik.

 

[No. 5]

Depending on encoding maybe ...
View attachment 216207
29 188 bytes
I'm not a reverse engineer, so it's not easy for me.

Yeah, that's above my paygrade. but it was worth checking just in case it ended up being usable. The microphone might work with something like REW, but without the calibration file, it won't do what you want to do.

I'd like to know that FR that measured with mic in reality is close to its form.

I can certainly get behind that desire, and I think it's worth pursuing. That's why I have a calibrated mic for designing speakers and measuring room acoustics, and why I have an off and on debate with myself about getting a diffuse field microphone. If you want some amount of assurance, as Blumlein 88 recommended, picking up a mic from Cross-Spectrum Labs is a way to get what you want, and knowledgeable professionals that I trust have remarked that a calibrated UMIK-1 gives about the same results as their high dollar gear. Or you can send them one of the mics you have in for them to calibrate it.

Voluntary EQing system with possibility of such mic responce variance as posted before is ... less than pointless. Almost like applying soft parmetric EQ with random settings.
It's always nice to have some physical reference even if you don't like it "flat". Otherwise you'll never know what exactly you are listening to.

I certainly would not say that it's less than pointless. At least not for low frequency.

To effectively apply low frequency equalization, you need to be able to identify the center frequency(s), bandwidth, and level of an offending resonance. An uncalibrated microphone will have zero effect on the first two of those three things. In other words, an uncalibrated microphone will not make a real 50Hz high Q resonance appear as a low Q dip at 90Hz. On the other hand, the level of the resonance will indeed be shown incorrectly, but look at the in-room measurement I posted earlier, there's some pretty offensive boom centered at 47Hz that's underreported by 2dB or so in the uncalibrated version. But that 47Hz boom is still clearly visible and can therefore be addressed. And if it can be addressed, the system will sound better, if the system sounds better, it was not less than pointless.

And note that the response variations of the UMIK-1 are less than+/-1dB from mid-bass up to 2.5kHz. To put some perspective on that, CTA-2034-A states that a +/-1.5dB variation between two separate laboratory measurements of loudspeaker's is considered good.

Again, not perfect, but still very usable.

 

[dougi]

I have an UMIK-1 calibrated by CSL, and recently compared it to a regular ECM8000.
Excluding potential diferences with positioning, i was impressed on how accurate the ECM8000 was:

My "Sonarworks" calibrated ECM8000 0 degree cal file is basically flat up until the high treble, and then only varies by a dB or so at most. I think most ECM8000s are pretty accurate, but I guess you don't know until you calibrate one, given the earlier picture of the spread of their performance.

 

[Speedskater]

the "Home Theater Shack" has a sub-forum on cal mic & SPL meter calibration.

https://www.hometheatershack.com/forums/spl-meters-mics-calibration-sound-cards.167/

but in seems to be less active inn 2022 than in past years.

 

[DonH56]

I tried a ECM8000 and it worked OK but had a fairly high noise floor. Not a problem for home use by any means. My main measurement mic for many years was an Earthworks M30, a well-regarded measurement mic. Before that I had a B&K or Sennheiser mic (decades ago, do not recall exactly, sorry). Some years back I purchased a CSL-calibrated UMIK-1, and an AVP I purchased came with its own UMIK-1. Comparing them, the calibrated UMIK-1 essentially overlaid the Earthwork's frequency response, though it did not have the noise floor or dynamic range of the M30. The uncalibrated UMIK-1 was almost identical though the midrange, deviating a little at the deepest bass (<50 Hz) and very high end (>10 kHz). A small electret I had sitting around was about as flat. The "recording" mics I have, just a few, were nowhere near adequate (things like AKG C1000/3000/414, Shure 57/58, etc.) for measurements.

I also found that REW (free) was more than equal to my previous "professional" SW package, and I did not need a preamp with the UMIK-1, so have not actually used the Earthworks for several years now (except as an omni mic for some recordings).

FWIWFM - Don

 

[Speedskater]

The ECM8000 is a model designation, it's been around for decades. The internal circuit and construction have changed from time to time. So be cautious about making generalities about the mic's performance.

 

[DonH56]

The ECM8000 is a model designation, it's been around for decades. The internal circuit and construction have changed from time to time. So be cautious about making generalities about the mic's performance.

Good to know. I no longer have mine, but it was purchased around 2010, so perhaps current models are better. My M30 was purchased long before that and rebuilt once. The ECM8000 was to be part of a DEQ2496 system but I went a different direction. From what I recall, bearing in mind it's an old memory, the response was very flat up until around 5~6 kHz then exhibited a series of frequency-response ripples around 1~2 dB or so. I was told why at one point but do not remember the answer. I have a vague memory of someone (Arny K?) commenting on the ECM8000 at one point that it had lower noise than another high-end measurement mic and was comparable to my Earthworks M30, which was one reason I got it -- hoping for a similar measurement capability for 1/10 the price. I also don't remember the actual noise floor; it was not as quiet as the M30, but I don't recall it being an issue for taking measurements. A friend was getting a DEQ2496 so I gave him the mic.

Based on all of that, I should have just skipped the comment on the ECM8000, foggy old memory...

 

[Speedskater]

I have a vague memory of someone (Arny K?) commenting on the ECM8000

Yes, I also think that it was Arny K (RIP), but I can't find any links or notes.

 

[DonH56]

Yes, I also think that it was Arny K (RIP), but I can't find any links or notes.

I'd have to dig, sorry. The only reason I thought it was Arny is because it stood out in my mind and drove me to look at that particular mic. I picked up a UMIK-1 and REW around then and switched to that. Bittersweet that a $100 mic (ECM8000 or UMIK-1) and free SW (REW) beats my prior $10k investment in mic, preamp, and SW...

I do wonder if I was thinking of a different mic, maybe one of the little Dayton's, instead of the ECM8000. I looked at several mics back then and may be mixing them up in my little pea brain.

Onwards - Don

 

[Blumlein 88]

Dayton Audio - iMM-6 iDevice Calibrated Measurement Microphone

iMM-6 iDevice Calibrated Measurement Microphone

www.daytonaudio.com

This goofy super cheap Datyon with cal file actually matches my Umik well. Not perfectly, but quite close. Worked fine on iPhones when they had a plug for this. I've had problems on Android phones getting it to work right. Worked perfectly with an old Windows laptop with this trrs plug. Unfortunately I don't have that laptop anymore. I need to figure out an adapter to some other input I can use. It is $40 now, but when I purchased it I think it was only $19.

 

[Blumlein 88]

Here is a chart where I compared a Umik 1, and two recording microphones. I used published response graphs of the two recording microphones to create a calibration file. I was using a speaker as the source which is why it is up and down, but I was interested in the relative response. I also think though I was careful the differences above 2 khz are partly due to non-exact placement of the microphone diagram. These would all be useful to a bit beyond 10 khz. Plus if I wanted, I could create a cal file using the Umik as a reference and get nearly exact results. The other two microphones were a pencil condenser omni and an LDC set to an omni pattern. These would get a little directional vs the Umik above 10 khz. The Avantone Ck-1 is a $150 microphone with three capsules including an omni. The CAD M179 is a multi-pattern microphone for less than $200. You can often grab them for $75-100 used. Plus you could record with them, I'd share my cal file if anyone is interested.


Here is an overlay of a Umik 1(red), Avantone CK-1 omni SDC (green), and CAD M179 omni LDC(blue). 1/12 octave smoothing.

 

[MRC01]

Wow, that's precise. As you may remember from this thread, I never got that kind of precision. However, what I did get was good enough to dramatically improve my in-room sound quality.

Another thing I learned during that experience: human hearing isn't omni at midrange & treble frequencies, and is not entirely omni in low frequencies, but is at least partly directional, so take the mic results with a grain of salt. Omni or cardiod, we don't hear exactly what the mic hears. Though it's close enough to be useful in most cases.