Wednesday, October 06, 2010

Powys Localization Pink Noise Test -Capsule Size, Boundary & Baffle Effects

Click on image to enlarge it. 

Vicki played a pink noise sample from a small CD player at 15 positions at 4.5 meters from the test arrays in a open, flat area. The test results can be used to study a number of boundary micing variables because she made efforts to control boundary size, boundary materials, capsule positioning, baffling and capsule types:

 

Paul Jacobson Oct. 20, 2010 Polar Pattern Plots

Unfinished Wood Boundary; 25mm capsule; SASS-Dimensions; No baffle; 
Frequency Response relative to Free Air Response at 1K Hz, 0 degrees.
 
Click on image for more detail


High-Density Closed-Cell Foam Boundary; 25mm capsule; SASS-P Dimensions; No Baffle
Frequency Response relative to Free Air Response at 1K Hz, 0 degrees. 

Click on image for more detail
 

High-Density Closed-Cell Foam Boundary vs. Wood Boundary; 25mm capsule; SASS-P Dimensions; No Baffle
Frequency Response relative to Free Air Response at 1K Hz, 0 degrees.

  Click on image for more detail


Crown SASS-P modified with 25mm capsules  
Frequency Response relative to Free Air Response at 1K Hz, 0 degrees.

Click on image for more detail
 

Crown SASS-P vs. Unfinished Wood Boundary; SASS-P dimensions; 25mm Capsule;  no Baffle
Frequency Response relative to Free Air Response at 1K Hz, 0 degrees.

Click on image for more detail


= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =

Danielson Oct. 10, 2010 Polar Pattern Plots & Observations


Click to Enlarge - Pink Noise, Outdoors @ 4.5 meters - test conducted by Vick Powys

For the two boundary samples above, 10mm omnidirectional mic capsules were flush-mounted in 4-7/8" X 3-5/8" boundaries.  0dB RMS calibration at 0 degrees.

Click to enlarge - > 6K Hz White Noise, Outdoors @ 4.5 meters - test conducted by Vick Powys

For the two boundary samples above, 10mm omnidirectional mic capsules were flush-mounted in 4-7/8" X 3-5/8" boundaries.  0dB RMS calibration at 0 degrees.

Small boundary influences on 10mm capsules (Rob D. observations)

The gain produced by the boundary has been negated in the plots by using a common reference of 0dB RMS for the pink noise at 0 degrees. The first plot shows how the overall sensitivity changes with direction. Note that the frequency response of the 10mm capsule in free air is basically uniform from 0 to 90 degrees. When the 10mm capsule is flush-mounted in small boundaries, the resulting polar patterns exhibit the "lobes" like cardioid microphones. With both boundary types, sounds coming from 75 degrees are reproduced with about 3 dB less gain than those coming directly into the microphone's "0" axis.  At 90 degrees, there is a 6 dB loss with a wood boundary and 4.5 dB loss with the high density closed-cell foam boundary. The decrease in off-axis sensitivity with the foam boundary sets in at 30 degrees and at 50 degrees with the wood. You can think of the foam boundary producing a slightly wider cardioid polar pattern than the wood boundary.

The high frequencies in the pink noise sample played back drop off so quickly that the plots labeled ">6KHz" can be more accurately regarded as response around 6K Hz.  The off-axis response at 6K Hz for the 10mm  capsule in free air is still quite good-- down only 2dB at 90 degrees. For boundary mountings, at 75 degrees, we see a similar cardioid pattern with a 7dB drop for wood and a 6dB drop for foam. At 90 degrees, however,  ~ 6K Hz response for wood is down 9dB and 8dB for foam. Paul Jacobson has been studying HF response in greater detail and he finds that above 8K Hz, the response of small boundary mounted capsules drops off quite rapidly. Interestingly, for sounds above 7K Hz, the sensitivity is higher in the 60 to 90 degree range than it is for sounds coming in more directly.  This counters the usual pattern with cardioid mics where the highest frequencies drop-off very quickly for sources positioned more and more to the sides.

 = = = = = = = = = =

Click to Enlarge - Pink Noise, Outdoors @ 4.5 meters - test conducted by Vick Powys

For the two boundary samples above, 25mm omnidirectional mic capsules were flush-mounted in 4-7/8" X 3-5/8" boundaries.  0dB RMS calibration at 0 degrees.


Click to Enlarge - > 6K Hz White Noise, Outdoors @ 4.5 meters - test conducted by Vick Powys

For the two boundary samples above, 25mm omnidirectional mic capsules were flush-mounted in 4-7/8" X 3-5/8" boundaries.  0dB RMS calibration at 0 degrees.

Small boundary influences on the 25mm Sennheiser capsule with Diffuse EQ on. (Rob D. observations)

Again, the gain produced by the boundary has been negated in the plots by using a common reference of 0dB RMS for the pink noise at 0 degrees. Take a look at the the polar pattern for the 10mm capsule in free air again. You'll see that the polar pattern for the larger 25mm mic in free air  drops in comparison--  its -6 dB at 75 degrees compared to 0 dB for the 10mm capsule, full spectrum.  Though the 25mm mic capsule has less off-axis response to start with, mounting it in the boundaries  does not seem to cause more loss of HF response when measured with full spectrum pink noise. The performance in this test may be enhanced by the use of the Diffuse EQ circuit in the Sennheiser mic which creates a rising 6 dB boost from 2K to 8K Hz and above.

As with the 10mm capsules, discrepancies grow as we look at ~6K Hz response.  The sensitivity at 60 degrees is sustained between boundary mounting and free air, but towards 0 degrees, there's a 5-6 dB difference between the boundary mountings and free air. Towards 90" there's about a 4 dB difference between boundary and free air. Because the pink noise did not reproduce evenly across the audio spectrum when played with Vicki's small portable CD player, we'll have to wait for better tests to get a full picture of response above 6K Hz. It is likely, however, that response above 7K Hz will drop off at faster rates as frequency increases. We should also expect peaks in the 60-90 degree range with falling HF sensitivity towards 0 degrees. We have seen this last effect exhibited in DIY and modified SASS-P arrays with flush-mounted larger diaphragm mics.

This test was conducted in part because we wanted to know why Vicki's SASS-like arrays with small, 10mm capsules sounded so much brighter than the SASS-P array with 25mm capsules. The answer, as Mike and Paul predicted, is that small capsules have considerably better off-axis high frequency response. As to the question of whether high density close-cell foam or wood makes the best boundary material, foam has a little bit better response from 60-90 degrees but wood has flatter response from 15 to 60 degrees. The differences might be a little more dramatic if the wood had been smooth and hard-finished.