• 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!

Human Ear Polar Patterns

OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
When I compute vector average for the measurements, it's exactly the same as (A+B)/2. Looks like somehow, the phase of the backward facing mic was the same as the forward facing one.
However, the backward facing mic was quieter, and had a different response curve, as expected.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,198
Likes
16,981
Location
Riverview FL
When I compute vector average for the measurements, it's exactly the same as (A+B)/2. Looks like somehow, the phase of the backward facing mic was the same as the forward facing one.
However, the backward facing mic was quieter, and had a different response curve, as expected.


Vector Average Example for me:

UMIK-1 facing centerline of speakers

Red - Left
Green - Right

Blue - Both speakers - that's addition in the air, adds 6dB

Orage - Vector Average - doesn't add 6dB, so I adjusted it upwards by 6.3dB - so the Vector Average and the measruement of both speakers playing can be compared.

The Vector Average of the left and right speakers almost exactly equals the actual L+R in-air measurement. Note the hole at 48Hz in the Both and Vector average that doesn't appear in the L/R traces, due to phase trouble around that frequency. 1/12 smoothing.

1589151554450.png




The Vector Averaged Phase also matches the phase of the "both speakers" measurement. Orange (calculated) covers the blue (measured) almost perfectly here. No smoothing, no windowing.

1589151854040.png



Sometimes I'm really impressed when an indirect calculation equals the results of a direct measurement.

Earlier, Impulse Response and Step Response calculated from a sine sweep equalled a recording of a single-sample impulse and a step waveform played by the speakers, Still amazed by that one.

I'll add this one to my list...
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
I think what Kvalsvoll said is causing this. There are in reality only two mike patterns. Omni which is a pressure microphone. Pressure peaks are velocity nulls. And bi-directional or figure 8 which in theory measures velocity. Pressure nulls are velocity peaks. A cardioid is a combination of omni and figure 8. In the old days they were literally that. An omni and ribbon would be mounted close together inside the microphone. Modern condenser cardioids use a trick to act like a cardioid down to lower frequencies while made like a pressure mike. Something of a pressure gradient microphone.

I suppose a good next step is to move your measuring position for both mics two a second place. Maybe a couple feet further from the wall to see if the dip still fits with distance to the rear wall or if something else is going on.
 
Last edited:
OP
MRC01

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,422
Likes
4,029
Location
Pacific Northwest
I think what Kvalsvoll said is causing this.
Seems reasonable. The key question becomes, how does human hearing perceive this case, having particle velocity and direction with a pressure null? My entirely subjective self-experiment suggests it detects a sound, the particle velocity triggers some response, not "nothing" which a pure omni detects.

... I suppose a good next step is to move your measuring position for both mics two a second place. Maybe a couple feet further from the wall to see if the dip still fits with distance to the rear wall or if something else is going on.
I've done this. With the UMIK-1, as I move further forward into the room, the 70 Hz null shifts to a lower frequency. That's how I know it's LBIR from the rear wall. However, the NT1A mics (and my H4 recorder's built-in mics) never show this null, even when I face 1 forward and 1 right behind it facing back.

I suspect that human hearing is not completely cardiod, nor omni, but something in between. The article I posted earlier says it's cardiod-like at mid to high frequencies (with diagrams) becoming "more" omni in the low frequencies. But still not completely omni even at low frequencies, because I can hear the 70 Hz dip but it's perceptually quite small, much smaller than the UMIK-1 measures.

If true, then omni measurements are still useful for showing peaks & nulls for tuning rooms. But we need to find out where the reflections causing them are coming from. If they're coming from behind, as in my case, then do not correct it fully, but only correct it partially. Because we don't hear reflections coming from behind, the same way an omni mic does. Our head & ears attenuate it somewhat, while an omni mic doesn't attenuate it at all.
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Seems reasonable. The key question becomes, how does human hearing perceive this case, having particle velocity and direction with a pressure null? My entirely subjective self-experiment suggests it detects a sound, the particle velocity triggers some response, not "nothing" which a pure omni detects.


I've done this. With the UMIK-1, as I move further forward into the room, the 70 Hz null shifts to a lower frequency. That's how I know it's LBIR from the rear wall. However, the NT1A mics (and my H4 recorder's built-in mics) never show this null, even when I face 1 forward and 1 right behind it facing back.

I suspect that human hearing is not completely cardiod, nor omni, but something in between. The article I posted earlier says it's cardiod-like at mid to high frequencies (with diagrams) becoming "more" omni in the low frequencies. But still not completely omni even at low frequencies, because I can hear the 70 Hz dip but it's perceptually quite small, much smaller than the UMIK-1 measures.

If true, then omni measurements are still useful for showing peaks & nulls for tuning rooms. But we need to find out where the reflections causing them are coming from. If they're coming from behind, as in my case, then do not correct it fully, but only correct it partially. Because we don't hear reflections coming from behind, the same way an omni mic does. Our head & ears attenuate it somewhat, while an omni mic doesn't attenuate it at all.
I may be wrong, It was my understanding our eardrums are pretty much like omni capsules responding to pressure. Being in a long canal with resonances and having a head in the way as well as directional disruption at higher frequencies from the outer ear makes them directional, but that would still be a pressure phenomena I think. Your listening experiment indicates otherwise.

One recording method is to use a pair of omnis spaced 7.5 inches apart. Normally you get little stereo effect, but you place a round 12 inch disc between them (Jecklin Disc) and they are due to that directional at least a little like your ears are shaded by your head. Still they are pressure microphones.
 
Top Bottom