Class 1 Measurement Microphone 1/2", 1/4" vs class 2 calibrated

[Savi]

Thanks ScottG and 1audio for your help. I confess I was a bit lazy and didnot try to understand the different mic spec. Based on your answers, I tried and found this link and test: https://www.jochenschulz.me/en/blog...7150-vs-behringer-emc8000-vs-beyerdynamic-mm1.
For me, the most interesting was the foreword with mic characs explanations.

 

[ScottG]

You're welcome!

Note that even the very expensive DPA mic.s are registering 1% distortion at 136 and 139 db respectively. My guess is that the OM1 is very comparable to those mic.s at only .5% at 133 db.


Also, in context to the Umik-1: I should point-out that MLS Impulse Response measurements are reasonably phase/time accurate with USB mic.s (they don't need Dual-channel). The problem with them (USB mic.s) however is that the distortion levels of the mic./sound-card audio-interface/amp makes the MLS measurement less accurate (even with large sampling), particularly at lower freq.s.. (..MLS doesn't handle added distortion well.) This is comparative with sine-sweeps (including ESS/Logarithmic) and other periodic (pulsed) measurements (like synchronized swept-sine) that need *dual-channel to get accurate phase/time data.

*or a system with a reference clock like Clio or Clio Pocket. Though note that Clio Pocket is quite limited:

Practical Test & Measurement: Sound Cards for Data Acquisition in Audio Measurements (Part 7) - Pitting the CLIO Pocket Against a Sound Card

In this article series, Stuart Yaniger decided to do a comparison experiment between the Audiomatica CLIO Pocket - and its own sound-card based system.

audioxpress.com

This is in-ADDITION to the problems regarding nearfield non-linear testing.


Basically - you are still better off with an analog mic. instead of solely using a USB mic. like the Umik-1.

 

[Speaker_tweaker]

I compared a Dayton EMM-6, last calibrated 5 years prior, to an Earthworks M23 and there was only maybe 1/2dB difference between the two. See attached images. The second graph is the difference of the two responses. These are gated measurements so low frequency data is not meaningful. Both microphones had calibration files applied. The Earthworks file came from the factory, and the EMM6 file came from cross spectrum labs. It seems to me even the cheapest mics can be relatively stable and accurate, provided you have a good calibration file.

Regarding standards, even a Class 1 mic is really not accurate enough for careful audio work without calibration. If you are shopping for a microphone to design speakers with, it MUST have a good calibration file. To my knowledge, there are no microphones available for less than $500 that come with a factory calibration file that has any basis in reality. Twenty years ago, you could send a mic to Kim Girardin for calibration, but he seems to have disappeared. Today, as far as I'm aware, the only way to get a dirt cheap mic with a good calibration file is to buy an EMM6 or UMIK-1 or 2 from cross spectrum labs. The next step up are the Earthworks M23/M23R/M30. One nice thing about the M23R is the calibration file is not really needed, so you don't have to worry about having the file enabled correctly or having the compensation inverted.

Someone was concerned that Earthworks mics are somehow less professional or accurate than mics from larger companies like B&K, but I haven't seen any compelling evidence for that. I knew a loudspeaker driver engineer who worked for a couple of high profile loudspeaker companies and also did consulting for a tech giant, and his microphone was the M30.

 

[DonH56]

I think I mentioned before that my CSL-calibrated UMIK-1 essentially overlies the response of my Earthworks M30. I no longer have the B&K 4191; I thought I would replace it, but it was more than twice the price of the Earthworks, and the M30 was almost 10x the price of the UMIK-1. Couldn't justify it for amateur use since I am no longer doing it for fee. The UMIK-1 doesn't match the self-noise and dynamic range of the Earthworks (or B&K) but is close and more than good enough for typical audio measurements. The price is great albeit bittersweet to me, and allowed me to retire my preamp. Add REW (free, compared to my old pricey "pro" acoustic SW) and you've a very powerful measurement suite for very little money (assuming you already have the PC, and even so a decent laptop PC plus UMIK-1 is far less than my previous measurement system).

 

[jhaider]

To my knowledge, there are no microphones available for less than $500 that come with a factory calibration file that has any basis in reality.

Do you have reason to doubt iSemcon’s calibration files? Their mic is $300 and change. The other nice thing about them is, unlike Earthworks, they also provide a grazing incidence cal file.

 

[Trell]

Do you have reason to doubt iSemcon’s calibration files? Their mic is $300 and change. The other nice thing about them is, unlike Earthworks, they also provide a grazing incidence cal file.

I got a grazing incidence cal file for my M23R when I asked, and it's likely auto-generated. That said, the M23R frequency response is -+ 0.5 dB before any calibration file is applied, but Earthworks will also sell you more M23R that is exactly matching any other as long as you've the serial number. I guess the grazing incidence calibration file is quite accurate.

 

[DonH56]

I got a grazing incidence cal file for my M23R when I asked, and it's likely auto-generated. That said, the M23R frequency response is -+ 0.5 dB before any calibration file is applied, but Earthworks will also sell you more M23R that is exactly matching any other as long as you've the serial number. I guess the grazing incidence calibration file is quite accurate.

How much difference is there between the direct and 90 degree files for your M23R? Since it's a small-diaphragm omni mic I always sort of wondered how large the difference... I do not see much with my M30 IIRC but it has been years since I looked closely.

Curious, thanks - Don

 

[Trell]

How much difference is there between the direct and 90 degree files for your M23R? Since it's a small-diaphragm omni mic I always sort of wondered how large the difference... I do not see much with my M30 IIRC but it has been years since I looked closely.

Curious, thanks - Don

Sure, attached two calibration files with the on axis the measured one Earthworks do for all M23R, and the 90deg calculated one. Possibly based upon another M23R measured at 90 deg incidence.

They do have a calibration device that cost $75. Since that would mean back and forth from Europe to USA with transporting and duties, I skipped that. For my usage it’s just fine.

 

[Speaker_tweaker]

Do you have reason to doubt iSemcon’s calibration files? Their mic is $300 and change. The other nice thing about them is, unlike Earthworks, they also provide a grazing incidence cal file.

No, there is no reason to doubt it! I forgot about iSemcon, that looks like a great choice. The claimed typical frequency response is very impressive. It's nice they also also provide diffuse field calibration.

I haven't looked at this in a few years. Since then, it looks like Audix has started selling the TM1 plus, which includes calibration data. I also see Beyerdynamic now provides calibration files on request for the MM1, but that's only on axis and down to 50hz.

I do believe the factory provided calibration data on the EMM6 and UMIK are not accurate. I have verified this myself on the EMM6.

Don, yes, I saw your previous comment on how the FR of both mics overlay. I also saw PKaudio's overlay of the same mics, but they diverge quite a bit on the high end.

 

[DonH56]

Sure, attached two calibration files with the on axis the measured one Earthworks do for all M23R, and the 90deg calculated one. Possibly based upon another M23R measured at 90 deg incidence.

They do have a calibration device that cost $75. Since that would mean back and forth from Europe to USA with transporting and duties, I skipped that. For my usage it’s just fine.

I started to plot these but they seem a little strange... The on-axis data starts at 10 Hz then jumps to 765 Hz, ending at about 24.5 kHz. The 90-degree file starts at 739 Hz and runs to ~52.5 kHz, way beyond the M23's range. And I am surprised there are no points between 10 Hz and 700+ Hz. I don't think I have a cal file for my old Earthworks, but the plot and table they provided shows a nice line of data points IIRC.

It does show lower level and faster roll-off for the 90-degree plot...

 

[jhaider]

I got a grazing incidence cal file for my M23R when I asked, and it's likely auto-generated. That said, the M23R frequency response is -+ 0.5 dB before any calibration file is applied,

You are correct - I omitted the key word "individualized" in my post above. When I asked for a grazing incidence file for my old matched pair of M30s, Earthworks supplied what they identified as a "generic" file. I agree that given the precision of the microphones it's good enough for the relevant tasks (sampling the in-room response for automated/manual room correction or subwoofer calibration) but it is not an individualized calibration.

Earthworks will also sell you more M23R that is exactly matching any other as long as you've the serial number. I guess the grazing incidence calibration file is quite accurate.

M23R has active circuitry to effect the match, I believe. The also sell pre-packaged matched pairs, in nice wooden boxes.

 

[Blumlein 88]

I do believe the factory provided calibration data on the EMM6 and UMIK are not accurate. I have verified this myself on the EMM6.

I don't know if I agree. I have compared a Umik-1 with factory calibration and the little sub-$20 IMM 6 meant for smartphones (also with factory calibration) and they match nearly to perfection. I didn't know about the Cross Spectrum labs when I purchased the Umik-1.

I also have taken published specs from recording condenser microphones and created my own cal files from those. With 1/12 octave smoothing, they match perfectly to around 2 khz and have some slight differences above that. Even those are usually no more than a decibel or so. So they might not be accurate enough for lab use, but they are in a pretty small ball park. Maybe I got lucky twice.

It would be interesting if someone had a few Umik1 with and without the CSL cal files to see how much they differ.

I was most surprised by the IMM 6 being so inexpensive. Alas it was meant for use with an Iphone which no longer has the jack for it. I don't know if it works with a USB C dongle. Something maybe worth trying out when I get around to it. It had issues with most Android phones (not the IMM6's fault). I had an old Lenovo laptop with a 4 channel headset input and it worked great with it. More handy than the Umik. Alas newer laptops at least that I have don't have that jack either.

 

[DonH56]

Here is a new plot thanks to @jhaider for the grazing file, scaled and normalized so the direct and grazing files are 0 dB at 10 Hz (below). More what I would expect, with fall-off at higher frequencies for the grazing angle. Actually a little more than I expected but it has been far too long since I measured microphone responses.

 

[mtg90]

I don't know if I agree. I have compared a Umik-1 with factory calibration and the little sub-$20 IMM 6 meant for smartphones (also with factory calibration) and they match nearly to perfection. I didn't know about the Cross Spectrum labs when I purchased the Umik-1.

I also have taken published specs from recording condenser microphones and created my own cal files from those. With 1/12 octave smoothing, they match perfectly to around 2 khz and have some slight differences above that. Even those are usually no more than a decibel or so. So they might not be accurate enough for lab use, but they are in a pretty small ball park. Maybe I got lucky twice.

It would be interesting if someone had a few Umik1 with and without the CSL cal files to see how much they differ.

I tested my pair of UMIK-1's with the miniDSP calibration files loaded and with the CSL calibration files loaded plotted against my calibrated M23. For further info the two UMIK-1's were also purchased from CSL 7 years apart in 2014 & 2021.

A link to the other tests I did was posted a few pages back as well:

MTG Designs - Measurement Mic Tests

When I was restarting the great waveguide shootout in early 2021 I noticed odd behavior in the distortion measurements of the compression drivers. Specifically I wasn't seeing the amount variation in the 2nd order harmonic distortion from the different compression drivers or horns that I was

www.mtg-designs.com

 

[DonH56]

This post shows the difference in 0, 45, and 90 degree CSL files for my UMIK-1: https://www.audiosciencereview.com/...ent-tutorial-for-dummies-part-2.5/#post-37383 Note these are the calibration files, not measurement files, so show the modifications needed to the response. After applying these, the response is essentially flat.

I have not found the plots where I compared the factory files to CSL. When I got mine, UMIK-1 cal files were not individual but just a standard file for all mics. I think that has since changed and miniDSP provides individualized cal files but am not sure.

 

[gnarly]

I'm in the process of building an outdoor test rig, where 4 or more mics will be mounting vertically on a 22ft high painters pole.
The pole is held plumb with guy wires, like old time antennas. I'm making it where i can set it up, take it down pretty quickly, to get mics on and off.

Anyway, could easily encompass a few learning mistakes, and don't want to risk a good mic going into "timber" mode if the pole comes crashing down haha.
So I'm gonna use ECM8000's.
Already have two of them , about to order several more, based on a quick comparison this morning to my Isemcon 7150 (which has been my reference mic)

Here's 7150 in green; the two ECM8000's are orange and purple on top of each other.
All three had the 7150's free-field calibration file in place.
(Measurement was indoors at about a meter, of a 4-way box high passed at 120Hz)

I have high hopes for the new test rig. Idea is to capture 4-5 vertical angles at once, for each horizontal spinorama angle.
I think the ECMs will work just fine.
And I must admit, i really don't sweat small dB variations in VHF....i think they are inevitable, unless maybe truly in a lab setting.

 

[Speaker_tweaker]

I don't know if I agree. I have compared a Umik-1 with factory calibration and the little sub-$20 IMM 6 meant for smartphones (also with factory calibration) and they match nearly to perfection. I didn't know about the Cross Spectrum labs when I purchased the Umik-1.

Again, I'm not sure about the Umik-1 or IMM-6, but the Dayton EMM6 I ordered from PE in 2018 came with a calibration file that was garbage. If I remember correctly, the mic needed a ton of compensation, but there wasn't much compensation in the file. The file also had a bunch of resonances that I doubt were in the mic itself. I tested it against my other cross spectrum calibrated EMM6 and Earthworks M23, which match each other when compensated. Again, that was 2018, it's possible they have changed their methods since then.

Maybe I got lucky twice.

I'm starting to wonder that about my microphones. I just found an old 2009 post by Herb Singleton of CSL in which he speculates the error of his calibrations could be as high as +/- 2-3 db at higher frequencies. Again, that was 2009, so he may have improved his methods since then. Here is the full text of that post:

"My cals (probably) are not +/- 0.5 dB accurate, certainly not across the full bandwidth of the measurement. If we start with the reference mic (which I get regularly calibrated by Scantek , an NVLAP/NIST-traceable lab), their cal methods have an uncertainly of 0.2 dB from 200-8000 Hz, 0.6 dB from 8-10 kHz and 0.9 dB from 10-20 kHz using an electrostatic actuator.

I haven't done a complete statistical analysis of the accuracy and precision of my method, but I've compared results I get with my setup with the results Scantek has generated for my other ANSI-rated mics to get an idea of the accuracy.

I use a free-field substitution method for the data about 100 Hz. The results about about 8 kHz is extremely sensitive to mic placement and room effects which cause most of the error. Below 100 Hz I use the substitution method in a pressure chamber and the error comes from imperfect seals around the mic capsule in the coupler (this is an issue with the EMM-6 and ECM8000 since they are slightly smaller in diameter than my 1/2 ACO 7052, so I have to pad out the diameter with electrical tape to get a good fit in the coupler).

Given the above, the data I provide from 100 Hz to around 3-5 kHz is accurate to around +/- 1 to 1.5 dB. Below 100 Hz and above 8 kHz or so the error increases, probably in the neighborhood of +/- 2-3 dB (perhaps more in the highest or lowest octave). The substitution method in principle is relatively simple but in practice there is a lot that can cause error. I think this is one of the reasons why NIST-traceable labs don't bother with this method (they all use electrostatic actuators now).

(in another thread either here, on hometheatershack.com htguide.com, or parts-express.com I have some accuracy numbers that were based on some quick calcs I did on data, and this post is an attempt to remember those numbers. If there is any conflict between this post and the earlier post, go with the numbers in the other post).

I do know that my method is repeatable. Every once in a while, I have an overseas delivery go awry and I get a microphone returned to me after it's been sitting in a foreign post office for a month or so. When I test it again, the result is always within +/- 0.5 dB of what I got previously. Thank heaven for small favors."

 

[Speaker_tweaker]

a quick comparison this morning to my Isemcon 7150

I'm in the process of building an outdoor test rig, where 4 or more mics will be mounting vertically on a 22ft high painters pole.
The pole is held plumb with guy wires, like old time antennas. I'm making it where i can set it up, take it down pretty quickly, to get mics on and off.

Anyway, could easily encompass a few learning mistakes, and don't want to risk a good mic going into "timber" mode if the pole comes crashing down haha.
So I'm gonna use ECM8000's.
Already have two of them , about to order several more, based on a quick comparison this morning to my Isemcon 7150 (which has been my reference mic)

Here's 7150 in green; the two ECM8000's are orange and purple on top of each other.
All three had the 7150's free-field calibration file in place.
(Measurement was indoors at about a meter, of a 4-way box high passed at 120Hz)
View attachment 235593

I have high hopes for the new test rig. Idea is to capture 4-5 vertical angles at once, for each horizontal spinorama angle.
I think the ECMs will work just fine.
And I must admit, i really don't sweat small dB variations in VHF....i think they are inevitable, unless maybe truly in a lab setting.

That is unbelievable consistency for a pair of unmatched, uncalibrated $30 microphones. As you are probably aware, when cross spectrum was calibrating ECM8000s they published this graph of the response of 85 samples, see attached. They later switched to the EMM-6 because of quality control issues.

I've been doing quasi spinorama measurements on a pole for the last 8 years or so. I have a speaker "rotisserie" with a gear motor, and I just manually rotate the speaker with a switch. It's a pain. Having multiple microphones would be good, but that means many more cables and the weight aloft is already an issue. Having multiple cheap microphones also presents a calibration problem if they don't magically happen to match. Perhaps there is some automation in REW or Vituix that helps with this? I've been using ARTA.

My pole is about 23feet (?) tall, and collapses to 13ft or so. I think the ground bounce at that height results in 1db(?) of ripple. I have an anchor on a concrete pad with a hinge for the bottom of the pole, and four guy wires. What I do is attach the speaker to the collapsed pole, tilt the whole assembly up and fix the last guy wire. Then I use a pulley to crank the speaker up into the air. The whole thing is a little sketchy and I can barely tilt it up and then extend the pole. If the pole was fully extended before tilting up I would not be able to lift it.

 

[Blumlein 88]

Again, I'm not sure about the Umik-1 or IMM-6, but the Dayton EMM6 I ordered from PE in 2018 came with a calibration file that was garbage. If I remember correctly, the mic needed a ton of compensation, but there wasn't much compensation in the file. The file also had a bunch of resonances that I doubt were in the mic itself. I tested it against my other cross spectrum calibrated EMM6 and Earthworks M23, which match each other when compensated. Again, that was 2018, it's possible they have changed their methods since then.



I'm starting to wonder that about my microphones. I just found an old 2009 post by Herb Singleton of CSL in which he speculates the error of his calibrations could be as high as +/- 2-3 db at higher frequencies. Again, that was 2009, so he may have improved his methods since then. Here is the full text of that post:

"My cals (probably) are not +/- 0.5 dB accurate, certainly not across the full bandwidth of the measurement. If we start with the reference mic (which I get regularly calibrated by Scantek , an NVLAP/NIST-traceable lab), their cal methods have an uncertainly of 0.2 dB from 200-8000 Hz, 0.6 dB from 8-10 kHz and 0.9 dB from 10-20 kHz using an electrostatic actuator.

I haven't done a complete statistical analysis of the accuracy and precision of my method, but I've compared results I get with my setup with the results Scantek has generated for my other ANSI-rated mics to get an idea of the accuracy.

I use a free-field substitution method for the data about 100 Hz. The results about about 8 kHz is extremely sensitive to mic placement and room effects which cause most of the error. Below 100 Hz I use the substitution method in a pressure chamber and the error comes from imperfect seals around the mic capsule in the coupler (this is an issue with the EMM-6 and ECM8000 since they are slightly smaller in diameter than my 1/2 ACO 7052, so I have to pad out the diameter with electrical tape to get a good fit in the coupler).

Given the above, the data I provide from 100 Hz to around 3-5 kHz is accurate to around +/- 1 to 1.5 dB. Below 100 Hz and above 8 kHz or so the error increases, probably in the neighborhood of +/- 2-3 dB (perhaps more in the highest or lowest octave). The substitution method in principle is relatively simple but in practice there is a lot that can cause error. I think this is one of the reasons why NIST-traceable labs don't bother with this method (they all use electrostatic actuators now).

(in another thread either here, on hometheatershack.com htguide.com, or parts-express.com I have some accuracy numbers that were based on some quick calcs I did on data, and this post is an attempt to remember those numbers. If there is any conflict between this post and the earlier post, go with the numbers in the other post).

I do know that my method is repeatable. Every once in a while, I have an overseas delivery go awry and I get a microphone returned to me after it's been sitting in a foreign post office for a month or so. When I test it again, the result is always within +/- 0.5 dB of what I got previously. Thank heaven for small favors."

I don't disagree with any of that. I do know my Umik has an individual cal file as I have looked at the file for others and they differ. Someone said at one time miniDSP used a single file for all mikes. I don't know when that changed, but I don't think it true now.

Measuring indoors you have to be picky, picky, picky about exact placement to compare two microphones. Tiny differences alter those upper frequencies.

Electrostatic actuators would be the way to go, but I'm guessing they are rather expensive.

 

[gnarly]

That is unbelievable consistency for a pair of unmatched, uncalibrated $30 microphones. As you are probably aware, when cross spectrum was calibrating ECM8000s they published this graph of the response of 85 samples, see attached. They later switched to the EMM-6 because of quality control issues.

I've been doing quasi spinorama measurements on a pole for the last 8 years or so. I have a speaker "rotisserie" with a gear motor, and I just manually rotate the speaker with a switch. It's a pain. Having multiple microphones would be good, but that means many more cables and the weight aloft is already an issue. Having multiple cheap microphones also presents a calibration problem if they don't magically happen to match. Perhaps there is some automation in REW or Vituix that helps with this? I've been using ARTA.

My pole is about 23feet (?) tall, and collapses to 13ft or so. I think the ground bounce at that height results in 1db of ripple at that height? I have an anchor on a concrete pad with a hinge for the bottom of the pole, and four guy wires. What I do is attach the speaker to the collapsed pole, tilt the whole assembly up and fix the last guy wire. Then I use a pulley to crank the speaker up into the air. The whole thing is a little sketchy and I can barely tilt it up and then extend the pole. If the pole was fully extended before tilting up I would not be able to lift it.

Awesome. Thanks for describing your rig.
Sounds like we are doing things very similar.

Although i don't hoist the speaker up. I have a deck about 12 ft off the ground, and either set the speaker & spinorama on a 3ft high acoustically inert barrel, or on top of 4ft high subs.
And yep, same thing as far as clipping two sets of guy wires in place, then standing the mic pole up, and clip the third.

It's hard to believe how close the two Beh mics i have, measured to each other. I ordered three more today and don't expect the same luck. But I can easily build calibration files to get them to match up (if even I care to do so.)

here's a pict of the painters pole taken from standing on the deck. Tomorrow I'll put the mic mounts on, and run/secure the XLR cables.
I'm thinking I'll have almost 30ms refection free.