• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Influence of pressure-field, diffuse-field, and free-field microphones on microphone orientation

Keith_W

Major Contributor
Joined
Jun 26, 2016
Messages
3,290
Likes
7,676
Location
Melbourne, Australia
One thing that has bugged me for a long time is why we sometimes point microphones on axis, and why we point it at the ceiling. For convenience I will call this 0deg and 90deg orientation respectively. I thought that it had something to do with whether we wish to capture reflections and measure surround speakers (90deg) or the on axis sound (0deg). For all my measurements and corrections, I have always pointed the microphone at the speaker. But - diligent audiophiles must always question if what we are doing is wrong.

There is certainly a measurable difference at 90deg. Jochen Schulz compared three microphones at different orientations and found most variation at high frequencies:

1723788840592.png
1723788850713.png
1723788862336.png


Note that the nonlinear frequency response below 2kHz is not the microphone, but the speaker he was measuring. His conclusion - always point the microphone at the speakers.

This lead me to do further reading about pressure field, diffuse field, and free field microphones. Bruel & Kjaer have a nice document explaining the difference here. In brief, free field microphones are for measurement in a sound field where sound waves propagate without any reflections. They are suitable for anechoic chambers, or measurement of speakers outdoors. Diffuse field (sometimes known as random incidence) microphones are used in sound fields with many reflective surfaces, producing reverberant conditions - as in listening rooms beyond the critical distance. Pressure field microphones are used where sound pressure and phase are the same in any position within the field, such as in the case in small enclosed spaces (like artificial ears) and nearfield measurement of drivers. The problem is, we want to do all 3 types of measurement!

The difference between the three types is in the capsule design:

1723789102423.png


You will notice that the pressure field microphone has a hole in the centre, and the other two don't. Why is this hole needed? As Sound on Sound explains, if a mic is placed on axis to a speaker, sound will strike the diaphragm at a 90deg incidence. At long wavelengths, this is not a problem. But at short wavelengths, a pressure build-up in the cavity will occur, causing an error of up to +10dB at 20kHz. The purpose of the hole is to relieve the pressure.

1723789288909.png

Bruel & Kjaer published this graph showing the different response of their Type 4971 pressure-field microphone (with the hole in the top) when used in the pressure field (blue curve, the response is flat), random incidence aka diffuse field (red curve, about a 2.5dB rise in frequency response at high frequencies), and free field (grey and yellow curves) at 0deg and 90deg respectively. The graph can be found here.

This made me realize that my "omnidirectional" Earthworks M30 mic isn't as omnidirectional as I thought. Microphone orientation matters, depending on what type of microphone you own! Nowhere in their product page do they indicate what type of microphone it is. Their calibration file does not indicate whether it is for 0deg or 90deg orientation (I am aware that the iSemCon 7150 comes with two calibration files).

1723790224002.png


So I compared the capsule grid of my Behringer ECM8000 (left) and Earthworks M30 (right). The Earthworks has a mesh covering the capsule, with no side hole slits. I am guessing that both are pressure-field microphones.

At this point the only thing left to do is to take my own measurement and see if I can replicate the theory. I used my Earthworks M30, swept the left speaker only from MLP at 0 deg (red) and 90 deg (green). There is a dip in the treble because of a mistake I made with the mixer (only noticed it when I was processing the curves ... doh!!). Only the orientation was changed between the two measurements, the position was otherwise not moved.

1723791708816.png


The first thing we observe is that the gain in the vertical position is about 2dB compared to 0deg.

1723791256321.png


When the two curves are normalised, we see that the mic loses 2dB of treble in the vertical orientation, similar to what Jochen Schulz observed. We also see a marked difference in bass which I was not expecting. I do not have a good explanation for this - the wavelengths are so long that in theory, neither microphone orientation nor any inadvertent small shift in microphone position should produce this boost. Maybe bass is reflecting from the ceiling and striking the diaphragm, but that SOS article says that bass freqs should pass through the microphone as if it's not there, so the angle of incidence should not matter.

This made me realize that even omnidirectional microphones have individual polar response and the manufacturers should publish their own polar measurements, the same way we want speaker manufacturers to publish spinoramas. I have to say I am a little bit miffed that even a microphone as expensive as my Earthworks M30 does not say what kind of mic it is on their website, nor do they supply calibration files for horizontal and vertical orientation. There is no market imperative for them to provide more information, because awareness of the issue is so poor. I did a search for "diffuse field microphone" and came up with no results - so it looks as if this has not been discussed on ASR before.

So now I have some questions!!

1. Is the Earthworks M30 a diffuse-field, pressure-field, or free-field microphone? Neither their website nor their manual says what it is. I have emailed them.
2. Anybody have a good explanation for the different bass behaviour at 0deg and 90deg?
3. Does this mean that if you own a pressure field microphone (hole in the top) you should be pointing your mic at 0deg to the speaker when doing nearfield measurements, and 90deg at the MLP where you are taking diffuse field measurements?

I assume that the Klippel @amirm is using has a pressure field microphone. Would you be kind enough to post a picture of the capsule?
 
Well some do offer different cal files for 0 and 90 degree. As for the low end I wonder about two things. One is did you have the microphone capsule in the same position? If you turned it up and this moved the capsule a bit higher, then you can get room modes effecting the result despite the length of the low frequency waves. Small changes near a node can make for differences greater than you think. Secondly, it could be low frequency noise from other sources. Low frequencies are very hard to block. So trains, trucks, etc. in your general area can corrupt this part of the frequency spectrum more than is often accounted for doing. One way to check for this is to do a few measurements a few minutes apart and see if they are consistent or not.

You can see from your B&K graph why they suggest making room measurements at 90 degrees. It reduces the effect of the on axis peak from direct sound from the speaker, and they can EQ out other effects in the cal file. You get similar issues when using omni SDCs for recording. I found using some Avantone CK-1 mics with the omni capsule that the treble could be off the charts when pointed ahead. I recorded a live performance once and had them pointed up and so many people kept asking why I would do that. A xylophone was one of the instruments used for some Christmas music. It sounded so much better than if I pointed the mics straight ahead. Some of the better companies take this into account for their omni SDCs.

Good topic BTW. Maybe some more knowledgeable people will comment.
 
MiniDSP UMIK-2. I have found several sites recommending it be used either way - so nothing definitive.
It does have a 0deg and a 90deg cal file.
IMG_3540.jpeg

Looks like it has the pressure field microphone hole.
 
Also looking at the mic graphs supplied I’m not surprised I’m finding it so hard to get 1k-4kHz range correct on my set up. How can anyone use a target curve when there’s that much variability.
 
One thing that has bugged me for a long time is why we sometimes point microphones on axis, and why we point it at the ceiling. For convenience I will call this 0deg and 90deg orientation respectively. I thought that it had something to do with whether we wish to capture reflections and measure surround speakers (90deg) or the on axis sound (0deg). For all my measurements and corrections, I have always pointed the microphone at the speaker. But - diligent audiophiles must always question if what we are doing is wrong.

There is certainly a measurable difference at 90deg. Jochen Schulz compared three microphones at different orientations and found most variation at high frequencies:

View attachment 386774View attachment 386775View attachment 386776

Note that the nonlinear frequency response below 2kHz is not the microphone, but the speaker he was measuring. His conclusion - always point the microphone at the speakers.

This lead me to do further reading about pressure field, diffuse field, and free field microphones. Bruel & Kjaer have a nice document explaining the difference here. In brief, free field microphones are for measurement in a sound field where sound waves propagate without any reflections. They are suitable for anechoic chambers, or measurement of speakers outdoors. Diffuse field (sometimes known as random incidence) microphones are used in sound fields with many reflective surfaces, producing reverberant conditions - as in listening rooms beyond the critical distance. Pressure field microphones are used where sound pressure and phase are the same in any position within the field, such as in the case in small enclosed spaces (like artificial ears) and nearfield measurement of drivers. The problem is, we want to do all 3 types of measurement!

The difference between the three types is in the capsule design:

View attachment 386778

You will notice that the pressure field microphone has a hole in the centre, and the other two don't. Why is this hole needed? As Sound on Sound explains, if a mic is placed on axis to a speaker, sound will strike the diaphragm at a 90deg incidence. At long wavelengths, this is not a problem. But at short wavelengths, a pressure build-up in the cavity will occur, causing an error of up to +10dB at 20kHz. The purpose of the hole is to relieve the pressure.

View attachment 386779
Bruel & Kjaer published this graph showing the different response of their Type 4971 pressure-field microphone (with the hole in the top) when used in the pressure field (blue curve, the response is flat), random incidence aka diffuse field (red curve, about a 2.5dB rise in frequency response at high frequencies), and free field (grey and yellow curves) at 0deg and 90deg respectively. The graph can be found here.

This made me realize that my "omnidirectional" Earthworks M30 mic isn't as omnidirectional as I thought. Microphone orientation matters, depending on what type of microphone you own! Nowhere in their product page do they indicate what type of microphone it is. Their calibration file does not indicate whether it is for 0deg or 90deg orientation (I am aware that the iSemCon 7150 comes with two calibration files).

View attachment 386781

So I compared the capsule grid of my Behringer ECM8000 (left) and Earthworks M30 (right). The Earthworks has a mesh covering the capsule, with no side hole slits. I am guessing that both are pressure-field microphones.

At this point the only thing left to do is to take my own measurement and see if I can replicate the theory. I used my Earthworks M30, swept the left speaker only from MLP at 0 deg (red) and 90 deg (green). There is a dip in the treble because of a mistake I made with the mixer (only noticed it when I was processing the curves ... doh!!). Only the orientation was changed between the two measurements, the position was otherwise not moved.

View attachment 386783

The first thing we observe is that the gain in the vertical position is about 2dB compared to 0deg.

View attachment 386782

When the two curves are normalised, we see that the mic loses 2dB of treble in the vertical orientation, similar to what Jochen Schulz observed. We also see a marked difference in bass which I was not expecting. I do not have a good explanation for this - the wavelengths are so long that in theory, neither microphone orientation nor any inadvertent small shift in microphone position should produce this boost. Maybe bass is reflecting from the ceiling and striking the diaphragm, but that SOS article says that bass freqs should pass through the microphone as if it's not there, so the angle of incidence should not matter.

This made me realize that even omnidirectional microphones have individual polar response and the manufacturers should publish their own polar measurements, the same way we want speaker manufacturers to publish spinoramas. I have to say I am a little bit miffed that even a microphone as expensive as my Earthworks M30 does not say what kind of mic it is on their website, nor do they supply calibration files for horizontal and vertical orientation. There is no market imperative for them to provide more information, because awareness of the issue is so poor. I did a search for "diffuse field microphone" and came up with no results - so it looks as if this has not been discussed on ASR before.

So now I have some questions!!

1. Is the Earthworks M30 a diffuse-field, pressure-field, or free-field microphone? Neither their website nor their manual says what it is. I have emailed them.
2. Anybody have a good explanation for the different bass behaviour at 0deg and 90deg?
3. Does this mean that if you own a pressure field microphone (hole in the top) you should be pointing your mic at 0deg to the speaker when doing nearfield measurements, and 90deg at the MLP where you are taking diffuse field measurements?

I assume that the Klippel @amirm is using has a pressure field microphone. Would you be kind enough to post a picture of the capsule?
When I review mics, I try to have the polar pattern data included. I can't find my EMX-7150 data at the moment, but here's the polar pattern for the M23R, which should be close to the M-30.

1723825628773.png
 
When I review mics, I try to have the polar pattern data included. I can't find my EMX-7150 data at the moment, but here's the polar pattern for the M23R, which should be close to the M-30.

View attachment 386836

Thanks, I was hoping you would reply :) According to B&K's graph, the polar pattern depends on the sound field you are using it in. So, how far away is the sound source (pressure field), and are you measuring it in a room (diffuse field), or are you measuring in an anechoic chamber or outside (free field)?

Do you have a website BTW?
 
One thing that has bugged me for a long time is why we sometimes point microphones on axis, and why we point it at the ceiling. For convenience I will call this 0deg and 90deg orientation respectively. I thought that it had something to do with whether we wish to capture reflections and measure surround speakers (90deg) or the on axis sound (0deg). For all my measurements and corrections, I have always pointed the microphone at the speaker. But - diligent audiophiles must always question if what we are doing is wrong.

There is certainly a measurable difference at 90deg. Jochen Schulz compared three microphones at different orientations and found most variation at high frequencies:

View attachment 386774View attachment 386775View attachment 386776

Note that the nonlinear frequency response below 2kHz is not the microphone, but the speaker he was measuring. His conclusion - always point the microphone at the speakers.

This lead me to do further reading about pressure field, diffuse field, and free field microphones. Bruel & Kjaer have a nice document explaining the difference here. In brief, free field microphones are for measurement in a sound field where sound waves propagate without any reflections. They are suitable for anechoic chambers, or measurement of speakers outdoors. Diffuse field (sometimes known as random incidence) microphones are used in sound fields with many reflective surfaces, producing reverberant conditions - as in listening rooms beyond the critical distance. Pressure field microphones are used where sound pressure and phase are the same in any position within the field, such as in the case in small enclosed spaces (like artificial ears) and nearfield measurement of drivers. The problem is, we want to do all 3 types of measurement!

The difference between the three types is in the capsule design:

View attachment 386778

You will notice that the pressure field microphone has a hole in the centre, and the other two don't. Why is this hole needed? As Sound on Sound explains, if a mic is placed on axis to a speaker, sound will strike the diaphragm at a 90deg incidence. At long wavelengths, this is not a problem. But at short wavelengths, a pressure build-up in the cavity will occur, causing an error of up to +10dB at 20kHz. The purpose of the hole is to relieve the pressure.

View attachment 386779
Bruel & Kjaer published this graph showing the different response of their Type 4971 pressure-field microphone (with the hole in the top) when used in the pressure field (blue curve, the response is flat), random incidence aka diffuse field (red curve, about a 2.5dB rise in frequency response at high frequencies), and free field (grey and yellow curves) at 0deg and 90deg respectively. The graph can be found here.

This made me realize that my "omnidirectional" Earthworks M30 mic isn't as omnidirectional as I thought. Microphone orientation matters, depending on what type of microphone you own! Nowhere in their product page do they indicate what type of microphone it is. Their calibration file does not indicate whether it is for 0deg or 90deg orientation (I am aware that the iSemCon 7150 comes with two calibration files).

View attachment 386781

So I compared the capsule grid of my Behringer ECM8000 (left) and Earthworks M30 (right). The Earthworks has a mesh covering the capsule, with no side hole slits. I am guessing that both are pressure-field microphones.

At this point the only thing left to do is to take my own measurement and see if I can replicate the theory. I used my Earthworks M30, swept the left speaker only from MLP at 0 deg (red) and 90 deg (green). There is a dip in the treble because of a mistake I made with the mixer (only noticed it when I was processing the curves ... doh!!). Only the orientation was changed between the two measurements, the position was otherwise not moved.

View attachment 386783

The first thing we observe is that the gain in the vertical position is about 2dB compared to 0deg.

View attachment 386782

When the two curves are normalised, we see that the mic loses 2dB of treble in the vertical orientation, similar to what Jochen Schulz observed. We also see a marked difference in bass which I was not expecting. I do not have a good explanation for this - the wavelengths are so long that in theory, neither microphone orientation nor any inadvertent small shift in microphone position should produce this boost. Maybe bass is reflecting from the ceiling and striking the diaphragm, but that SOS article says that bass freqs should pass through the microphone as if it's not there, so the angle of incidence should not matter.

This made me realize that even omnidirectional microphones have individual polar response and the manufacturers should publish their own polar measurements, the same way we want speaker manufacturers to publish spinoramas. I have to say I am a little bit miffed that even a microphone as expensive as my Earthworks M30 does not say what kind of mic it is on their website, nor do they supply calibration files for horizontal and vertical orientation. There is no market imperative for them to provide more information, because awareness of the issue is so poor. I did a search for "diffuse field microphone" and came up with no results - so it looks as if this has not been discussed on ASR before.

So now I have some questions!!

1. Is the Earthworks M30 a diffuse-field, pressure-field, or free-field microphone? Neither their website nor their manual says what it is. I have emailed them.
2. Anybody have a good explanation for the different bass behaviour at 0deg and 90deg?
3. Does this mean that if you own a pressure field microphone (hole in the top) you should be pointing your mic at 0deg to the speaker when doing nearfield measurements, and 90deg at the MLP where you are taking diffuse field measurements?

I assume that the Klippel @amirm is using has a pressure field microphone. Would you be kind enough to post a picture of the capsule?
Excellent post. I always like to learn something useful, especially when I didn’t even realize that I didn’t know about it. Didn’t recognize you at first. You’re not a dragon any more.
 
I always assumed you pointed the mic toward the ceiling because we are trying to measure the sound in the room that's hitting our ears, not just the speaker. Our ears are "directional" but the directionality is the result of two ears and a brain.

I also assume the calibration for a measurement mic is done at 90 degrees. Although it might be helpful to have an on-axis calibration if you are testing/measuring speakers in an anechoic chamber (or outdoors).

I'd consider an omnidirectional mic to be "not particularly directional" :p
 
I got the miniDSP umik-1. It comes with 0 degree and 90 deg cal files. IIRC, they recommend using 0deg orientation if measuring a single speaker at a time. When setting up a multichannel system, it’s more convenient to use 90 deg orientation. I notice the little measurement mic supplied with a Denon AVR uses 90 deg orientation.
 
This made me realize that my "omnidirectional" Earthworks M30 mic isn't as omnidirectional as I thought. Microphone orientation matters, depending on what type of microphone you own! Nowhere in their product page do they indicate what type of microphone it is. Their calibration file does not indicate whether it is for 0deg or 90deg orientation
While not stated explicitly, IMHO one can safely assume that it's your normal free-field type and the frequency response shown (and cal file) is for 0deg incidence. If otherwise they would have stated it. Pressure field applications are as special and rare as are diffuse field applications, compared to free field.
But indeed a professional company should offer precise documentation.
 
And the (almost) hassle-free solution to the general issue: use a good 1/4in capsule like Microtech Gefell MK301. It only 2.5dB down at 16kHz/90deg, and completely flat polars in a +-30deg window up to 20kHz.
 
When I review mics, I try to have the polar pattern data included. I can't find my EMX-7150 data at the moment, but here's the polar pattern for the M23R, which should be close to the M-30.

View attachment 386836
I had managed to get a 90 degrees calibration offset from Earthworks for M-series mics after some stalking. Their reply with the attachments:

"We do not capture 90-degree ECFs for each microphone, but the attached frequency plot and ECF indicate the 90 degree response of a typical Earthworks M-Series measurement microphone."
90_Degree_M_Series_Typical_Response.png

I also suggest using a minimum phase version of the calibration file with 90 degrees as the phase shifts are quite significant.
 

Attachments

  • 90_Degree_M_Series_Typical_Response (1).txt
    4.6 KB · Views: 34
use a good 1/4in capsule
Yes, it's important to note that, ceteris parabus, a 1/4" will give a better polar pattern than a 1/2" or 1", the tradeoff being noise.
 
Yes, it's important to note that, ceteris parabus, a 1/4" will give a better polar pattern than a 1/2" or 1", the tradeoff being noise.
Damned physics again. We would be using great 1/8 th inch mics if it weren't for sensitivity and noise.
 
Damned physics again. We would be using great 1/8 th inch mics if it weren't for sensitivity and noise.
MEMS Mics!
FWIW, I've used Knowles SPU0410LR5H-QB in a mic array (for research on bats) which is only 3x3.6mm and has excellent and predictable HF response.
Noise can be dealt with in various ways, like using very long log sweeps and/or averaging.
 
  • Like
Reactions: 617
I also suggest using a minimum phase version of the calibration file with 90 degrees as the phase shifts are quite significant.

But how does it create "significant" phase shifts with practical use?

umik-a.png umik-b.png umik-c.png

Screenshot from 2024-08-16 14-26-28.png Screenshot from 2024-08-16 14-26-44.png Screenshot from 2024-08-16 14-26-55.png
 
Yes, it's important to note that, ceteris parabus, a 1/4" will give a better polar pattern than a 1/2" or 1", the tradeoff being noise.

I thought that 1/8" mics are preferable because they are small compared to the wavelength. Not true?

Also, can we make a generalisation that "hole on top" = "pressure field mic"?
 
I thought that 1/8" mics are preferable because they are small compared to the wavelength. Not true?
True. Almost no baffle step with this small size.
Also, can we make a generalisation that "hole on top" = "pressure field mic"?
I don't think such generalizations are valid.
In the end its all just EQ what defines the class. Pressure field basically means flat response when flush mounted, no baffle step to compensate.
 
I thought that 1/8" mics are preferable because they are small compared to the wavelength. Not true?
The smaller the diameter, the better the polar pattern and the worse the noise (and sensitivity). Tradeoffs.
 
Back
Top Bottom