Comparison of SASS-P and PBB2 Localization & Depth - Race Cars

Download folder with QuickTime Comparison Movie & Select Pairs [ 38mb .zip ]

I wanted to study some outdoor recordings made with Rich's stock Crown SASS-P with ones from my PBB2 array (a DIY version of the SASS Model B but using flush-mounted 21 mm AT4022 omni mics).  A trip to a race track to see my brother race provided me an opportunity to do some simultaneous, side by side recording. The dynamic and robust outdoor subject does not handicap the SASS-P for its lower self-noise performance.

The differences I can detect are striking and potentially instructive. The SASS-P produces higher levels of brighter toned sounds in the center of the stereo field. Take a listen to the separate pairs in the folder.  The proportionally higher volume in the center of the field  enhances the impression of depth though it comes with some loss in width and lateral localization performance. In Pair_1, the car moves deeper, more vividly into the center of the field in the first SASS-P sample, yet the car moves off further off to the right at the end of the of PBB2's rendition. In  Pairs 2 & 6, the middle of the SASS-P's stereo field feels substantially more populated and dynamic. In Pair 7 with the cars off to the right, the depth imaging of PBB2 seems wider and the  reflections picked up on left side feel more vibrant and airy. 

I personally feel that the tonal balance of the SASS-P borders on being too bright, but its depth imagery in the center of the stereo field is considerably more engaging. The PBB2 sounds too dark in comparison. Paul Jacob's interest in getting brighter response from flush-mounted AT4022's could be on target. See more discussion of this below.

In order to learn more about specific, contributing qualities, I tried to equalize the PBB2's recording to sound more like that made by the SASS-P.  One octave wide +4dB boosts centered at 4K Hz and 8K Hz helped, but the slurred, superimposed mid-range pitches of the cars created numerous, different peaks between 500-2K Hz in both arrays. Boosting the level of the center of the PBB2's Stereo field using M-S EQ helped the most but the number and varied distribution of peaks over 3K Hz made it clear that the two array have many complex differences. In the end, I opted to equalize the two recordings separately to hear how the EQ'd versions would sound next to each other. It seems that the under-expressed center of the PBB2 was easier to correct than the narrowness of the SASS-P.  Both arrays produce spatial "airy-ness" in the lower registers, but the PBB2, post-equalization, has a bit more of this "magic" to my ears. Taste plays a big role at this point.  Neither array produced results with this loud subject that were that were impressive as-is.  I also noticed that the SASS-P's lower mid range had more "body" to work with and smoother bass response. This was surprising to discover given the overall warmer sound of the PBB2.

As Paul Jacboson pointed out in relation to some confirming measurements he made on his SASS B-like rig, "Billingsley and Bartlett...noted that the SASS was designed to work with flat pressure response mics, rather than flat free-field response."  The SASS-P's mics are are flat pressure response type and the AT4022's are made to have flat free-field response in free air, not flat response when mounted in a boundary.  If this phenomenon is responsible, why is the SASS's better HF response mostly evident in the center of the stereo field? (I thought that flat pressure response mics had reduced HF response at 0 degrees). I need to dive into the Lipshitz and Vanderkooy AES paper again with these new insights.

The race car test is consistent with Vicki's recent localization tests and those I did with PBMB1, that the HF lift for the flush mounted AT 4022 rigs is from the sides around 3 - 4:30-- including the Knapp/Elliot MKH-20 SASS-P body mod.  This correlates directly with the boundary "grazing effect" which could be more influential with the flush-mounted AT 4022 rigs than we have grasped up to now.

Assuming the goal is to get more HF response in the center of the stereo field using low noise mics, here are some possible variables to explore: 

(1) Make AT40022 mics into flat pressure response type mics and mount them in a PBB2-like body. Paul says we can buy the AT4022 capsules separately. 

(2)  Take the above flat pressure response modded 4022's and mount them PZM-style closer to the boundary than I was able to do with the stock 4022's in the Barn test. 

(3) Vicki's recent localization comparisons remind us that "Wedge" arrays have always exhibited stronger center of the field imaging. Her foam wedges have 5" setbacks and the increased grazing effect towards the front could be a bit too pronounced [see still image from the test showing waveforms with 12 o'clock impulses matched .gif  and movie  .zip ]  Its seems logical to make an array that retains head-spacing and extends the leading edges to create effective 1.5" to 2" setbacks. The drawback with the wedge arrays has been the disappearance of the lower mid range "airyness."  Maybe we can discover how the PBB2 body is creating this airy-ness while improving center gain and center HF response.  Using MS-EQ to boost the center of the field is always a post option, but getting more from the start would be great. Vicki has proposed making her Foam SASS Lite's baffle smaller but having a less boundary "nose" with baffle modification might move this intuition forward. I might try head-spaced capsules with a 1.5" to 2" setback in boundaries that are 4- 7/8" wide and  2.5" high to get more HF lift.

(4) There's also a chance that the painted close-cell foam boundaries that Vicki has been experimenting with have impact on the greater HF response she's getting (with 10mm capsules). Paul J has suggested it might have more to do with the capsule size, but its probably time to do an exhaustive test of boundary materials on flush mounted AT4022 mics.  We might choose to add thin, metal plates after the results are in. The test I did of this variable before was with Perp2boundary with large diaphragm mics.

Vicki Powys Foam SASS & Wedge Array Tests- Part 2

Download QuickTime Movie [.zip ]

For this second sets of test with the Foam Wedge mic set at 70 degrees, I selected segments with bird motion nearer to the arrays with more distant calling to better study stereo imaging performance. I assumed that the Wedge70 would have poorer imaging but I'm not so sure. What do you think?

[Test conditions: SD702 record level indication @ "60" dB; Olympus LS-10 at Low Setting "10", MKH-20's Diffuse Field EQ ON which produces noticeable high frequency "graininess"]

The next test features similar material comparing my Foam SASS and my SASS Model B with MKH-20's.

Download QuickTime Movie [.zip ]

There is certainly more noise with the mini PIP mics, but to my ears the sound is more open and natural. I would like to measure the Wedge70 directly against the Foam SASS but I only have one LS-10 recorder. Does anyone know how I could use the minPIP mics with my SD702?  [Rob D note: For using PIPmics on your SD702, Mike Wall used a battery box very similar to Richard's schematic. Mike Rooke suggests these resistor values: 8K resistor for normal sensitivity; 4.7K for higher sensitivity and 10K for lower sensitivity. Pin 3 on the XLR input should be left unused. There's more about the circuit on Richard's site.  Mike Rooke might know of other phantom powering options he can build for you.]

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

Rob D. Comments:

First the question Vicki posed, Which of the DIY arrays has better stereo imaging?

(A) The SASS-B mod with the MKH-20's produces more "body" or emphasis in the lower mid range from approx 200-600Hz. This gives the SASS-B array a fuller tone and more for the brain to deduce distance from. It also has a noise advantage even when boosted to match the response of the PIPmics.

As for the runner-up:

(B) The DIY Foam Wedge array seems to be bunching the sounds together more in the center of the field compared to the DIY Foam SASS.  Vicki used deeper setbacks (5') and changed the WEDGE angle in this test.

(C)  The DIY Foam SASS seems to be producing slightly less dynamic range than the SASS B. The louder bird calls, though brighter, seem to have less separation from the background sounds which results in a little less audible clarity. I'm not sure where this could stem from, but I have a hunch that a coat of acrylic paint on this foam might help improve the boundary effect.

(D) I thought the DIY arrays were imaging pretty similarly to the SASS=B until I compared some presence clips between the Wedge and the DIY SASS arrays [ Download QT movie  3mb .zip ]. There are no simultaneous moments to compare, but its pretty audible to me that the SASS possesses some lower mid-range "magic" that contributes dynamic "airyness" that the Wedge lacks.  Another reason to make sure that the foam boundaries are  truly not detracting from the SASS DIY array's performance? Wood is not that much heavier and its easier to assure a seal around the capsule where it protrudes from the boundary.

 = = = = = = = = = 

Another looming question posed by these comparison tests for me is why is the tonal balance for the small 10mm diameter PIPmics so much brighter than the larger MKH-20's capsules?  This is true for both foam boundaries.  I used parametric EQ on the SASS B-MKH20 samples to approximate the frequency response of the PIPmics in the DIY Foam boundaries: 

Download QuickTime SASS Arrays Matched with EQ movie [4mb .zip ] 

Download QuickTime SASS vs Wedge Arrays Matched with EQ movie  [5mb .zip ] 

The PIPmics have significantly more mid-range to high frequency respons when mounted in these foam boundaries. The most pronounced addition being a +9dB peak centered around 5KHz that extends all the way down to 1K Hz with some additional lift around 400 Hz-1K Hz. There might be a little too much HF emphasis in the PIPmics on the small boundaries but it does adds clarity to sounds in the distance as Vicki observes.

As to what can be causing this "lift" in the mid-range and high frequencies, Paul Jacobson, has suggested that the softer nature of the foam boundaries might result in less of the HF roll-off associated with boundaries and flush-mounted mics. Paul cited formulas in the Long & Wickersham  patent behind PZM technology that one can look at .pdf  Paul's is the best explanation I have though I'm wondering how reducing amount of HF roll=off would lead to a lift this dramatic that extends under 1 KHz.  Thanks Paul for this interesting lead to pursue!

If the softer boundaries have this effect on Hz response,one might be able to apply this variable beneficially-- such as making the lift a little less pronounced and more,.. This I propose a test with the below test samples to provide some evidence. I've included the larger diaphragm MKH-20 mics to eliminate the possibility of mic capsule size playing an unsuspected role.

(1) Pink Noise on boom box at 10' outdoors at 3:00 o'clock -> MKH20's in SASS-B  -> SD702 (max gain)

(2) Pink Noise on boom box at 10' outdoors at 1:00 o'clock -> MKH20's in SASS-B  -> SD702 (max gain)

(3) Pink Noise on boom box at 10' outdoors at 3:00 o'clock -> MKH20's in free air 13" apart facing front (removed from array)  -> SD702 (max gain)

(4) Pink Noise on boom box at 10' outdoors at 1:00 o'clock -> MKH20's in free air 13" apart facing front (removed from array)  -> SD702 (max gain)

(5) Pink Noise on boom box at 10' outdoors at 3:00 o'clock -> PIPmics in Foam SASS -> Battery Box Powering ->  SD702  (max gain)

(6) Pink Noise on boom box at 10' outdoors at 1:00 o'clock -> PIPmics in Foam SASS -> Battery Box Powering ->  SD702  (max gain)

(7) Pink Noise on boom box at 10' outdoors at 3:00 o'clock -> PIPmics in free air 13" apart facing front -> Battery Box Powering->  SD702  (max gain)

(8) Pink Noise on boom box at 10' outdoors at 1:00 o'clock -> PIPmics in free air 13" apart facing front -> Battery Box Powering ->  SD702  (max gain)

(9) Pink Noise on boom box at 10' outdoors at 3:00 o'clock -> PIPmics  flush-mounted in 3-5/8" X 4-7/8" hard finished wood boundary with 1" setback -> Battery Box Powering ->  SD702  (max gain)

(10)  Pink Noise on boom box at 10' outdoors at 1:00 o'clock -> PIPmics  flush-mounted in 3-5/8" X 4-7/8" hard finished wood boundary 1" setback -> Battery Box Powering ->  SD702  (max gain)

Avoid use of  wind screening materials for this test. 

Affects of PZM, Flush-Mounted & Boundary-Facing Orientation on Tonalty & Spatial Imaging

Download QuickTime Comparison Movie 19mb .zip

I would like to thank Paul Jacobson, MIchael Billingsley and Bruce Bartlett for their invaluable input on this test and probably several more to follow. Thanks also to Rich Peet for the kind loan of his SASS-P microphone.

The test is an initial study of tonality, localization and depth imaging properties that accompany three ways of mounting mic capsules within the pressure zones of the small boundaries of a SASS-like enclosure:

A. Flush-mounting of medium size omnidirectional microphones (21mm AT3032's) similar to the SASS Model mods made by Walt Knapp and my PBB2.

B. "PZM" mounting with small diameter omni capsules facing the boundary (as in the SASS MKII -P)

C. Boundary-facing positioning similar to PZM-mounting except with medium size omnidirectional microphones (21mm AT3032's) (I refer to the orientation as "downward-facing" in the current movie graphic).

The above mounting variables are configured in arrays that are very similar in size and construction. All arrays have 3-5/8" wide X 4-7/8" high boundaries. The capsules in arrays A & C have a 1" setbacks from center of mic capsules to the leading edges.  The SASS-B has a .8" setbacks to the center of its capsules. The capsule positioning of A & C are offset vertically close to 1/3 proportioning. Both of the SASS-P's capsules are  2" down from the top boundary edges, The PBB2 array has hard finished wood boundaries. C has hard smooth metal boundaries and the SASS-P has slightly dimpled, hard plastic boundaries. The void behind the foam baffle in A & C is filled with deadening material (rubber-coated carpet underlayment). The void behind the SASS-P is in the original state manufactured by Crown. The foam baffles for A & C are made of high density dark gray acoustic foam (1.8 lb./sq.Ft). 

Rather than wait a month or more for this Summer's robust insect crop to wane and conduct the test in my usual outdoor site, I decided to try this test in my rectangular pole barn. In addition to pink noise, I used two audio clips, "Low Hz" to address response <1000 Hz and time of arrival imaging and "Mid Hz"  for >1000Hz to correspond with amplitude difference imaging. The pink noise results are presented in another movie (4 mb .zip download)

Download QuickTime Comparison Movie of Pink Noise Samples ( 4 mb .zip )

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

Rob D. Follow-up & Observations August 30, 2010.

The pink noise test sonogram differences are interesting to look at for pronounced differences such as the HF drop-off for on-axis sounds with the boundary-facing configuration (also referred to as "down-firing"). Also, the overall brighter response of the SASS-P in conjunction with less bass. I think one asset of the SASS arrays that could distinguish them from many other (or all?) arrays is their depth perspective and the "room" character that one can sense in the nature of the reflections.  Here are some of my initial observations along these lines and then a follow-up test.

(A) The SASS-P is too bright but its sense of depth and "room" character is quite striking compared to the other arrays. (I suspected this would become evident and its why I included the boundary facing larger mic with low self noise to see if there was a similarity). Small diaphragm mics supposedly work much better in PZM-like applications. This creates a bind as they best of the small capsule mics have ~6 dB(A) more self noise than the mics I'm used to for recording natural ambience.  A good application of the small mics in PZM mounting might be my PBB3 cap plan but I might want to re-think the boundary sizes as the extra brightness might become excessive .(I wonder if the small capsule size itself is somehow contributing to the HF boost Vicki is getting with her small capsule flush-mounted boundary experiments.)


(B) Bruce Bartlett predicted some excessive resonance with a larger diaphragm mic facing the boundary in PZM like fashion, and this test seems to confirm this phenomenon Though the impact of the depth imaging has some pleasant, unusual qualities, there's a dramatic peaking in the mid-range that creates dissonance.  The 4022's have stock "free-air" grilles in place and there's a chance this could change a bit if the mics had none, but the lower frequency location suggest to me that the resonance problems are associated with the larger capsule size.


(C) The Free Air mics have very different tonal character and they seem to exaggerate the resonances in the large barn space a lot more than the other capsule mounting configurations. Some people like this quality but from me they mask other complexities, nuance and overall clarity of the space and contained sounds.  The forward facing or  "A/B" style array with  13" separation does a fairly good job of localization but the depth is much flatter. For example, the 1 o'clock position feels closer,, more in the same plane as the 2:30 sample compared to the other arrays.

(D) The PBB2 seems to be a compromise, and a pretty good one. It has some of the faster depth drop-off of the SASS-P, more room character than the Free-Air and is tonal balance is much warmer and more pleasant than the SASS-P.  I just wish is had more "room" character of the SASS-P.


Next, I equalized the low (mostly <1KHz) and mid (mostly >iKhz) clock position samples (2:30, 1:30 & 1:00) with parametric EQ like I might If was trying to modestly bring-out spatial clarity. 

Download QuickTime Comparison movie 14mb [ .zip ]

I'll give folks a chance to listen to this test while I work on my observations.

Vicki Powys Light-Weight Closed-Cell Foam Mic Arrays

Vicki Powys, of New South Wales, Australia, has begun testing some experimental boundary mic arrays she constructs from foam. One is a SASS-shaped construction made of "Yoga" EVA foam blocks (ethylene-vinyl acetate). This material is described as having "barrier properties." Thee other rig is less dense, slightly dimpled, closed-cell foam that is used for making life preservers that she paints with acrylic paint.

For several years, she's has been using an expensive and fairly heavy SASS Model B unit modified by Walt Knapp with flush-mounted Sennheiser MKH-20 mics in a Crown SASS MKII-P body. Her goal is to make a lighter-weight, comparable quality array that is easier to hike with.  In this test, she has incorporated small (10mm diameter) microphones made by PIPmics with a self-noise of 14 dB(A)--about 4-6 dB higher than the Sennheiser 20's.

From the test materials there are three QuickTime comparison movies and links to photos of the arrays. She also sent some photos of her DIY Lycra wind screens for her Wedge and SASS arrays which could be of general interest to recordists.

A.  Blind comparison QuickTime movie (4mb download .zip)  to assess the performance of the two DIY arrays in relation to her Crown SASS Model B:

    Powys Adjustable Wedge Closed-Cell Foam Array (construction,  furry windscreens)

    Powys SASS-shaped EVA Foam Array (construction)

    SASS Model B - Knapp Mod with MKH-20 omni mics.  DIfusse EQ and HP filter OFF.

The self-noise of the PIPmic "Mini" mics -> LS10 combination in the DIY arrays proved to be fairly audible in comparison to the MKH-20->SD702 combination [download 4mb comparison QuickTime movie .zip].  The Olympus LS-10 recorder [measured EIN -121 dBu (A)] has noise that extends lower into the spectrum and could be adding a touch of input noise. In any case, I feel that its likely the PIPmics will create a quality hit when recording in lower sound level settings. Its unfortunate, but not surprising.

To aid in the evaluation of the stereo imaging performance of the arrays, I used parametric "notch" EQ to lessen some the of the "hiss" from the PIPmics between 4-7Khz in the test movie samples.

The samples in the movie graphic are unlabeled. The dark samples are the SASS Modle B reference.  Of the green and red samples representing the two DIY arrays, which sounds like a Wedge array and which sounds like a SASS-like array?  Do these DIY arrays have potential? 

B.  The outcome of the first comparison surprised me as the imaging of the DIY Wedge array (bottom, red regions) with its sizeable, 3-1/4" setback compared very favorably with the SASS B mod.  I would have preferred  more spaciousness in the imagery as a whole, but Vicki commented that a running stream nearby or distant coal trains might be masking some lower mid-range acoustical cues.  We checked to make sure her MKH-20 capsules were mounted flush to the boundary and they were. The tonal character of the DIY rigs was noticeably brighter to good effect and we noted that the "Diffuse EQ" option for high Hz emphasis on the MKH-20 mics had been turned OFF.  [She has done some preliminary tests with "Diffuse EQ" turned on after this test and finds the increased HF response beneficial.]

I compared her SASS-B's recording with some I made with my PBB2 and like her DIY arrays, it had more response above 3K Hz. Wondering if Vicki's SASS Mod B array was performing up to snuff, I made movie comparison of her recordings to a recording made by Andrew Skeoch with another SASS Model B Knapp Mod and Vicki's DIY Wedge. Note that the "Diffuse EQ" was OFF on Andrew's array as well.

 

QuickTime Movie (12mb download .zip)

 

Obviously, Andrew's recording was made in a different location but it seems to have more HF response than Vicki's SASS Model B array given these samples. If the two rigs truly differ, I'm not sure why. Vicki reports that she'll continue to use the "Diffuse EQ" setting on the reference SASS array for her tests in any case.

The DIY Adjustable Closed-Cell Foam Wedge array holds-up quite well stereo image-wise. The array seems to benefit from more amplitude in the center of the stereo field compared to her DIY EVA Foam SASS array. When recording very distance sounds in open spaces such as these, more amplitude in the center of the stereo field can increase the sense of depth. This effect can also be realized to some extent in post using M-S processing so it might be more instructive to conduct careful localization performance testing on a few DIY array options before assuming the Wedge is superior. I know she has conducted tests with the wedge at set at a 70 degree angle. Maybe I can convince her to post the results of this test on the blog.


It doesn't seem like the multiple bumps that can be seen the painted foam boundaries of the wedge (inherited from the life preservers) would help the boundary effect but how much negative impact they have is unknown. One is tempted to assume, not much. 

The effectiveness of the EVA material on the boundary effect remains unknown, in part because of the question about the HF response of the reference mic and in part because the performance of this array was less impressive in this test. It could be that the paint coating is important. 

To determine if these materials work as well or better than traditional, hard smooth surfaces, one would have to make up three arrays with the exact same dimensions out of (1) EVA "Yoga" block foam, un-coated, (2) Closed-Cell foam, painted (3) Wood, fine sanded and sealed. It might require a careful localization test done outside at 100'  to produce definitive differences. In the past, I found that high density foam (probably EVA) produced a different lift compared to wood, but this test hasn't been repeated.  The "hardness" requirement may not mean that the boundary surface be perfectly smooth because as Vicki points out, the boundaries on the Crown SASS array have a dimpled surface that is not too dissimilar to the texture of her painted, close-cell foam. Its tempting to forge ahead and assume that there is no significant change in at least using the painted foam, but one could be wrong.

Knapp SASS B Mod vs. Crown SASS MKII-P Blind Comparison Test

Going through Walt Knapp's SASS B mod instructions, I noticed the comparison recordings he provides of insects for the stock MKII with PZM mics and his mod. I thought it would be useful to revisit this as a blind test in light of discussions Paul Jacobson and I are having about the lower HF response of flush mounted omni mics like the AT4022's that are designed for use in free air. The 7mb QuickTime comparison movie can be down-loaded here .zip.  I'll identify which recording was made with which array below witin a few days.

Listening Comparison Of Comb Filtering in Boundary Mounted & Free Air Mics

One of the criticisms of placing microphones next to boundaries is that it produces objectionable comb filtering. This comparison test demonstrates that any comb filtering produced is very subtle and occurs at very high frequencies. The use of moving pink noise  for detecting the phenomenon was suggested on the Nature Recordist list. I performed many "3 O'clock" to "12 O'clock" pans with a boom box speaker around the mics in an arc at a distance of 5 ft. and 10 ft. I selected one sweep from both distances where all three mics exhibited smooth amplitudes.  I used a 744T recorder with 45 dB of gain to record all three mics at the same time.  I had to use a mix of AT3032 and AT 4022 mics which are known to be very, very similar in terms of performance. 

A blind version of the QuickTime movie comparison is available for downloading (3mb .zip)  to see if you can match the samples with the right mic configurations. I'll soon post the ID's below. 

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

Here's the comparison test the samples identified [ download .zip 3mb]
Here's a larger image of the sonogram with more detail.


Follow-up discussion can be found on the NatureRecordist list as well as in Paul Jacobson's assessment below. Curiously, the Perpendicular to Boundary capsule positioning produced the least amount of comb filtering in this test. 

The guesses I received  suggest that any artifacts pink noise sweeping might reveal are difficult to accurately link to the arrays:


Sample A Guesses:  Free Air, Free Air, PBB, Perp2Air, Perp2Boundary
Sample B Guesses:  Perp2Boundary, PBB, Perp2 Boundary, PBB. PBB
Sample C Guesses PBB, Perp2Boundary, Free Air, Free Air, Free Air

PBB3 - Portable Stereo Boundary Mic Array Plan

Finally got around to seeing how PBBx components would scale for a baseball cap. Worn backwards, would it be a little less conspicuous?   I wanted a rig I could use hands-free and hear what the mics are "hearing" without needing headphones. The stereo imagery should have impressive depth and reach. Recordings should play well on speakers as well as headphones. There are only a few small compromises in quality as best as I can tell.

In solving the structural integrity aspects, make sure that the leading edges of the boundaries end where shown. Add no hard, reflective surfaces in between them, even behind the foam.  I'd consider adding a "V" wedges behind the boundaries that attach to the cap sides.  The angles can flex and change a bit, no problem.   You might also want to clip off any part of the brim that sticks out beyond the foam baffle that is under the capsules-- that would create another boundary. Off-set center of mic capsules in the vertical  boundaries .75" from top on one channel and .75 from bottom on the other. For wind-screening, I'd start with stretchable fabric like a sheer Lycra? pulled across all of the mic components. No obstacles allowed on the boundaries themselves. Make sure  the diaphragms of the mics are flush with the boundaries.