I wish Earthworks offered axial and 90 deg calibrations, as CSL does as well as some newer contenders such as iSEMcon. But they don't.
Earthworks does have a generic 90 deg file available on request. For Dirac and the like it's good enough. That said, lately I've been using a Neumann MA-1 microphone for Dirac cals, along with a Focusrite interface. I figure the polar response is better just because it's a smaller capsule. Alas, Neumann offers 90 deg cal but NOT on axis cal, so you can't use the MA-1 as a calibrated mike to measure speakers.
If your measurement mike is a cylinder, without fancy supersonic nose cones, the difference between on axis & 90 is fixed. That's probably what the Earthworks file is. So just apply this '90 correction' to the on axis calibration.
This 'difference' is size dependent. If your 'measurement' mike is a different cylinder diameter, move it up or down in frequency as required.
The 'correction' is also available in B&K & GRAS datasheets and various Acoustics textbooks like Beranek & Olson.
Member
Joined 2003
Yes you can - but you need a very good measurement setup/situation.
And yes, a generic correction between 0° and 90° should be good enough up to 20kHz.
And yes, a generic correction between 0° and 90° should be good enough up to 20kHz.
A new mic is in town:
https://www.behringer.com/product.html?modelCode=0506-AAB
Use at 0 or 90, not much difference:
https://www.behringer.com/product.html?modelCode=0506-AAB
Use at 0 or 90, not much difference:
Alas it doesn't come with individual correction files. Hardly possible for that price. The Sonarworks Sound ID mike still seems the cheapest option with that.
This curve illustrates some of the caveats you should be looking out for in a 'calibration'.
The wriggles above 1kHz are almost certainly due to poor mounting so this 'calibration' is mostly nonsense.
I have been using a Cross Spectrum Labs calibrated UMIK-1 now for 10 years for all my designs. Seems to produce an acceptable result with the speakers I have come up with. I wouldn’t be too concerned about behavior above 16kHz. The voicing is all in the 400Hz to 10kHz range. Just make sure you use a calibrated SPL meter and a decent Vrms meter to get a handle on signal levels.
as mentioned above there is absolutely no issue with using a calibrated cheaper mic for response measurements. When you want to start reliably and accurately measure distortions that when problem starts. You need pay much more to cover that ground, just look up how much hificompas’ measuring setup cost 🙂
And when you start measuring with different microphones… be prepared have your mind blown that your microphone, your amp, your method eg. MLS vs Farina, your environment (room boundaries and acoustics, noise floor) measurement distance, mic gain all affect your result. Unless you take considered steps, what you’re measuring is your recording chain…
Last edited:
Hi All,
Plenty of images in this posts for fun.
Some tests on four microphones, two of them are calibrated and the others are not but it is an old laboratory and the other is cheap microphone for sound engineers.
Four microphones, Left to right on the photo below :
1. Superlux ECM-999 . No calibration file
2. SonarWorks XREF20 with a calibration file
3. EarthWorks M23R A calibration is supply but not need
4. A B&K 4155 capsule (1980) mount on an input stage B&K ZC 0020 coming from a B&K2231sound level meter. My first measurement microphone. Not have calibration. I made an enhancement to a XLR P48. You can find in the manual of the B&K2231 theses curves :
The photo
The DUT is the SEAS 27TAC/GB, on axis measurement 1m distance with a gate of 4.2ms.
The M23R reference measurement :
Now the comparaison between measurement starts :
Superlux ECM999. Here +2.5dB max but honestly for the price not sot bad
The XREF20, good calibration from Sonarworks 🙂 same response.
The B&K 4155. -1.5dB max of deviation but we are the tolerance of the curve above. Now I can make a correction curve.Note this microphone can go as high as the M23R. I think the signal noise ratio is better than the M23R.
My conclusion Is : you can do speaker measurements with cheap microphones in the audio range. Just need a good calibration file.
I should test distortion measurement to see the difference.
An other test I made with the M23R, the woofer SEAS W22NY003 mount on a box of 25cm width. I compare my measurement with SEAS datasheet. My measurement (red) is limited by the gate 4.2ms below 200Hz. Not bad at all.
Enjoy !
Plenty of images in this posts for fun.
Some tests on four microphones, two of them are calibrated and the others are not but it is an old laboratory and the other is cheap microphone for sound engineers.
Four microphones, Left to right on the photo below :
1. Superlux ECM-999 . No calibration file
2. SonarWorks XREF20 with a calibration file
3. EarthWorks M23R A calibration is supply but not need
4. A B&K 4155 capsule (1980) mount on an input stage B&K ZC 0020 coming from a B&K2231sound level meter. My first measurement microphone. Not have calibration. I made an enhancement to a XLR P48. You can find in the manual of the B&K2231 theses curves :
The photo
The DUT is the SEAS 27TAC/GB, on axis measurement 1m distance with a gate of 4.2ms.
The M23R reference measurement :
Now the comparaison between measurement starts :
Superlux ECM999. Here +2.5dB max but honestly for the price not sot bad
The XREF20, good calibration from Sonarworks 🙂 same response.
The B&K 4155. -1.5dB max of deviation but we are the tolerance of the curve above. Now I can make a correction curve.Note this microphone can go as high as the M23R. I think the signal noise ratio is better than the M23R.
My conclusion Is : you can do speaker measurements with cheap microphones in the audio range. Just need a good calibration file.
I should test distortion measurement to see the difference.
An other test I made with the M23R, the woofer SEAS W22NY003 mount on a box of 25cm width. I compare my measurement with SEAS datasheet. My measurement (red) is limited by the gate 4.2ms below 200Hz. Not bad at all.
Enjoy !
^My conclusion Is : you can do speaker measurements with cheap microphones in the audio range. Just need a good calibration file.
I should test distortion measurement to see the difference.
I came to the exact same conclusion. Cheap mics don't usually have low distortion, but you really do not need that for most speaker building.
https://www.a4eaudio.com/miscellaneous-studies/a-comparison-of-some-pre-amps-and-microphones
I should test distortion measurement to see the difference.
I came to the exact same conclusion. Cheap mics don't usually have low distortion, but you really do not need that for most speaker building.
https://www.a4eaudio.com/miscellaneous-studies/a-comparison-of-some-pre-amps-and-microphones
That calibration (technically verification) part is unfortunately the most challenging aspect. (not all provided cal files are accurate)
But yes, I agree overall.
As for distortion, it really depends.
No human being can reliably predict significant distortion issues just by intuition. (freq resp, impedance etc)
But understanding distortion is absolutely essential for determining appropriate crossover points.
Wich often can be determined "enough" with just cheap microphones.
The challenge is that affordable measurement microphones often don’t offer the resolution needed to capture those finer distortion details accurately.
So it kinda depends how far you want to go and important it is to you.
The websites of @A4eaudio and @mtg90 are very useful 🙂
And I forgot to write about my measurement system, it is a DIY JIG Box with a DIY microphone preamp P48 based on a INA217 and an integrated amplifier with a LM3886 30W/8. I would like make an integration of the soundcard in the box as I done when I have a creative sound card. It will improve the precision of the measurement. The sound card is a TASCAM US-1x2HR . I can do impedance and spl measurements just turning a knob...
The goal of my post is to show you can do basic measurements with a basic microphone, reinventing the wheels. It is useful you see what happen. Most of people don't want to spend money or don't have the technical knowledge behind measurement theories.
A ECM999 or a ECM8000 are useful tools for audio engineers to tune a sonorisation system. For the non professional, the hobby builder of speaker it could avoid some big mistake like have a driver in reverse polarity for exemple.
Yes I'm agree if you go further like distortion or measurement in the low frequency or both, there are big difficulties, serious technical challenges.
It is for me something to explore but I don't know if microphone measurements is enough ?
And I forgot to write about my measurement system, it is a DIY JIG Box with a DIY microphone preamp P48 based on a INA217 and an integrated amplifier with a LM3886 30W/8. I would like make an integration of the soundcard in the box as I done when I have a creative sound card. It will improve the precision of the measurement. The sound card is a TASCAM US-1x2HR . I can do impedance and spl measurements just turning a knob...
The goal of my post is to show you can do basic measurements with a basic microphone, reinventing the wheels. It is useful you see what happen. Most of people don't want to spend money or don't have the technical knowledge behind measurement theories.
A ECM999 or a ECM8000 are useful tools for audio engineers to tune a sonorisation system. For the non professional, the hobby builder of speaker it could avoid some big mistake like have a driver in reverse polarity for exemple.
Yes I'm agree if you go further like distortion or measurement in the low frequency or both, there are big difficulties, serious technical challenges.
Yes and you can do better matching of drivers. It is important to mix drivers with the same profil of distortion, you will have a much more homogeneous sound.
It is for me something to explore but I don't know if microphone measurements is enough ?
Well lucky us, our friend @DcibeL has purchased and reviewed something just like this.
An all-in-one box that includes a small amplifier, an audio interface, and jig, and microphone to allow impedance measurements and acoustic measurements for loudspeaker cabinet and crossover design.
Review of all-in-one box here
Last edited:
Very cool box, make things easier. The J-Box III look like a lot my JIG box, I put less options because I want it simple. I went back to four RCA cinch because I could change the sound card. I have a lot of holes in the panels because i have had several versions. All configurations/commutations are done with relays.