Topic: Oddball microphone technique (OMT) - part 3

[Gutbucket]

Have you ever tried omnis on all four sides of that box?

Do it! In my experience this works extremely well.  It is in essence a loose description of my "other" OMT rig which I don't post about, is the setup I run most often, and why I commissioned Chris Church to build a custom 4-ch CA-UGLY2 preamp years ago for use into a DR2d.

Differences are use of a rectangular rather than square "chest/torso box", rotated 45° to form a Left/Center/Right/Back configuration rather than 4-corner quad-like arrangement, with four DPA 4060 in the DPA boundary mounts, one at the center of each face.  This was the inspiration for my initial experimentation with the original OMT4 array that extended a basic pair of spaced omnis to a sort of diamond-shaped Decca-tree rather than the traditional triangle-shaped Decca tree, using 1m spacing and sphere-mounted omnis rather than flying a big rectangular box atop a stand.  The effect of the wide rectangular baffle box became even better emulated by substitution of the directional front/rear-facing center pair between the wide-spaced omnis rather than using 4 omnis all around, and eliminated the need for as much front/back spacing.

I'm sure folks following this thread can envision what I'm getting at, but PM me if you'd like clarification.

[Gutbucket]

Off-center setup/config for specific room
^ I may have posted a link to this in the previous OMT thread, but am linking it here after Heathan cited it today in another thread.

Good discussion and examples of thoughtfully applied odd setup techniques for the common halfway back and off-center recording position in a hardwalled room problem.

[EmRR]

With a PAS shotgun approach in a fairly small indoor room like I tried, I gotta wonder if there is value in using only one, pointed directly at the closest stack rather than at the center of the stage, then mixing it in the center of another stereo array.  It's usually vocal register stuff I'm missing, and dead on the stage adds more drums than anything, which is NOT the deficit in the other arrays. 

I noticed a Rycote mount meant for mounting to 1/4" camera points that looked like a possibility for this, combined with a clamp, since it has a swing arm with some slide positioning.  It's definitely fiddly to get positioned coincident but seems like a solution.   I can see I may end up machining a replacement swing arm of greater length and slide to really get this where I want it.   Full rig in a post below.

[Gutbucket]

I like where that is going.  I thought I had commented previously on the bit below, but maybe that was only in my own mind since we were talking about other things as well..

With a PAS shotgun approach in a fairly small indoor room like I tried, I gotta wonder if there is value in using only one, pointed directly at the closest stack rather than at the center of the stage, then mixing it in the center of another stereo array.  It's usually vocal register stuff I'm missing, and dead on the stage adds more drums than anything, which is NOT the deficit in the other arrays.

Yes!  My justification for a shotgun over a better behaved pattern is the push for maximum possible pickup of direct-arrival sound for good clarity, most often needed for vocal register stuff as you mention.  It seems odd to intentionally make the array asymmetrical by pointing the directional center microphone somewhat off-axis from the rest of the array, but one can think of it as a further "division of labor" or differentiation of tasks that OMT is built around to begin with.

Before I moved to a M/S center pair (which in my current setup is more or less fixed in a directly-forward pointing direction with respect to the other mics in the array), I had the center mic on a telescopic TV antenna that gimballed at its attachment point, and sometimes pointed the center microphone in a direction other than straight-ahead with respect to the rest of the array.  I usually did that when presented with an awkward recording location that did not allow the wide-dimension of the array to be perpendicular to the axis along on which I wanted the directional center microphone to point.  Or sometimes up on-stage or stage-lip for non-PA amplified instrumental acts where I'd orient the entire array with respect to the ensemble, then point the center mic up and over at the snare rather than at the kick-drum.  The first few times I did this I was concerned about causing an image imbalance, but was pleasantly surprised to find that wasn't a problem.  The stereo image is initially set by the orientation of the primary wide-dimension of the array, and strongly influenced by manipulation of balance level of the Right/Left assigned array channels in the mix.  Angling the single directional center microphone away from directly forward did not shift the imaging at all, it only affected the quality of sound populating the center of the playback image.  That was not initially intuitive, but makes sense considering image placement of a single center microphone is determined to some extent by its spacing relationship with respect to the other microphones, and to a greater extent by its panned balance position in the mix.  If panned to center, it's content is going to fill the center of the playback image regardless of where it's pointed.  It needn't even be a member of the array.. or even share the same content.  Consider a voice-over mixed to center.  It fills the playback center image but isn't otherwise related to the stereo content it overlays at all.

Your setup above takes this a couple steps further.  1st by achieving maximum on-axis focus in that channel by using a shotgun mic, and 2nd by retaining a M/S or X/Y center pair as part of the rest of the array with the angled single PAS shotgun coincident to that.

[Gutbucket]

How important is that? Dunno, this is new territory.

It raises these questions in my mind:
Can we eliminate the center pair by just using the fig-8 in combination with the PAS-angled shotgun for M/S?
If so, should the fig-8 remain perpendicular to the rest of the array (which is the basis of stereo orientation for the array), or to the angled shotgun?
^
I'll have to think some more about that one.

[EmRR]

You couldn't find your original response because the quote started over in that ORTF + omni thread. 

I'll certainly listen to the shotgun as the M of the MS pair to see.  I am guessing the directionality may make the image weigh to one side, because of the weighting of information in the MS encoding. 

Separately I am curious to see how the spot shotgun addition to an MS pair blends.  I can see there being some dynamic bleed artifacts with things having temporary wider bursts that swim around.  Possible the shotgun will need LPF/HPF to tailor blend, on top of needing to have a minimal blend level to achieve spotting without overwhelming the main MS. 

Looks like a 7 mic array test coming up.....

[EmRR]

Theoretical 7 channel rig build.   THAT took awhile to configure.  I'm hoping the angled shotgun spot will prove useful at Ramkat in W-S, NC, and some other wide/shallow rooms.   I've got a couple things coming up there so should be able to try it soon. 

The dual mid-side is fore and aft as Rolo 46 has proven with classical music over at Gearslutz.  I've run it once that way and didn't note any obvious problems. 
Center coincident shotgun to check against the dual mid pattern options, or mix with. 
Second coincident shotgun for PAS use, thinking to spot some extra vocal clarity, etc.  That clamp definitely needs a safety chain. 
AB omni outriggers at 30" spacing.  I've run them this width a couple of times.

I'd love to have a way to mount the DMS over/under as is typical, with the shotguns in between them, keeping the capsules as vertically close as possible, but I don't see how to achieve that easily.  This was as close as I could get it all without hardware or shocks touching each other. 

If it all turns out to be a giant phasey mess, well......

[Gutbucket]

If it all turns out to be a giant phasey mess, well......

Difficult to maintain tight coincidence with 4 mics.  That's a lot up there.

Looks like it might be possible to put both shotguns on top to get the stack tighter vertically by extending the rear vertical support upward, flipping the top forward-facing shotgun upside down, and moving the PAS shotgun from the bottom to just under the top shotgun, clamped to the rear vertical support.  PAS angle would then be adjusted by rotation of the clamp on the vertical support.  Depends on the ability to position the PAS shotgun coincidently that way.

If you can't get the PAS shotgun well-aligned that way, just flipping the top forward-facing one upside down should allow you to get that one in very close in above the DMS pair.

But this is all little more than armchair photo analysis.

[EmRR]

If it all turns out to be a giant phasey mess, well......

Difficult to maintain tight coincidence with 4 mics.  That's a lot up there.

Looks like it might be possible to put both shotguns on top to get the stack tighter vertically by extending the rear vertical support upward, flipping the top forward-facing shotgun upside down, and moving the PAS shotgun from the bottom to just under the top shotgun, clamped to the rear vertical support.  PAS angle would then be adjusted by rotation of the clamp on the vertical support.  Depends on the ability to position the PAS shotgun coincidently that way.

If you can't get the PAS shotgun well-aligned that way, just flipping the top forward-facing one upside down should allow you to get that one in very close in above the DMS pair.

But this is all little more than armchair photo analysis.

Yep, armchair analysis here too.  I will say there may be advantage in the vertical distance that's counterintuitive.  I've found multiple close mics in studio could be better or worse opposite of normal expectations.

One problem with the clamped extension is it's just a couple degrees off of straight, so there's another angle to fight in positioning.  Getting the shotgun capsule coincident is harder than expected, with a fight between available mount position/obstruction and mic position within the mount.  No one would be throwing a rig like this up in a hurry, this is really more the kind of thing I can get away with in an official capacity, starting at soundcheck. 

I realize I neglected to try mounting the angle gun off of a right angle stud adapter connected to the center stud....and that might require a short extender to get the mic bodies higher. 

A right angle mount for the DMS that allowed both to mount horizontally with front/rear running left/right and side running front/back orientation would collapse the height requirement substantially. 

[Gutbucket]

I also find the coincident thing sometimes behaves oddly outside of a straight coincident stereo pair.  I've tried omnis coincident with supercards in a standard near-spaced stereo arrangement in an attempt to extend bass response a few times but that never worked right without low-passing the omnis (yeah I know that's the "right way" anyway), regardless of omni level, even though the mics were very closely coincident and theoretically shouldn't have had significant phase conflicts above the low-pass.

I've seen photos of a L/C/R arrangement which intentionally used a very large vertical spacing.  It was built upon a standard near-spaced 2-channel L/R pair with a center mic placed directly above it something like 2 meters  higher.  The intent was to have relatively close near-spaced coincidence for direct-arrival in the horizontal plane, combined with significant decorrelation for diffuse reverberant sound arriving from all directions.  No idea how well it worked though.

[EmRR]

Interesting, 2 meters is huge.  I've had problems at times.  OTOH, you might remember I did a MS thing with the side between a card and an omni, and found I could treat them as a single variable pattern mic.   MKH though, very well behaved in the first place, and much closer than my theory above.  I definitely plan to do more 8 and omni combos for picking pattern later. 



I've run a pair of omni together a lot lately in studio world, so I could pick the preferred sound later.  In a couple cases I've run them 50/50 in the mix with no artifacts.  That's a KM131 with an MKH20, both ruler flat on paper, but very different sounding.  The Neumann is soft up top but has a great lower mid range quality, and the Sennheiser has more articulation in the treble and lower bass. 

[Gutbucket]

I definitely plan to do more 8 and omni combos for picking pattern later.

^
This remains high on my list as a modification of the wide A-B omnis in an OMT setup to dial in variable front/back sensitivity. 

May prove useful for stereo mixes as well, but the primary motivation is better use of those channels as discrete surround feeds by providing control over how much forward direct sound is picked up by them.. I once determined I needed at least 6dB reduced forward sensitivity in the omnis when they were routed to the surround speakers in order to keep the front image from bleeding into the surrounds (and for that reason have considered rear-facing subcards in place of the omnis as a simpler if non-variable alternative).  If that works well enough I may be able to eliminate the seperate near-spaced rear-facing supercard pair, and keep the channel count the same.  [edit- although it works really well to route the near-spaced rear-facing supers to the rear surrounds and the omnis (or better, rear-facing subcards) the the side surrounds in a 7 ch surround playback arrangement..]

To do that I need to overcome the practical challenge of supporting the 8's and their cabling on the feather-light telescopic bar I use (again, rear-facing subcardioids may be the more practical answer).. and I'd need to pickup another sd 8 since I only have 2 on hand and don't want to give up the one used in the center MS pair.  Would be more easily doable in a few joints where there is a long railing and I can just clamp the 8's.

[EmRR]

Take 2.  Closer, easier, better.....in theory.  Could flip the Two MS mics position and get the center shotgun a bit closer. 

[Gutbucket]

Checked David Griesinger's website this morning, which I do a few times a year [http://www.davidgriesinger.com/] and found the portion quoted below in a recent post.  My bold emphasis on his assertion which I take as confirmation of an important aspect I hear and continued to pursue via OMT arrays-
 
"10/31/19 I have been so busy I have not had time to update this website with the new code for LOC, our measure that predicts from a binaural impulse response whether a particular seat will provide “Proximity”. The experiments shown in the paper for that talk showed that reflections that arrive up to 5ms after the direct sound augment the direct sound, and increase the likelihood of hearing proximity. But reflections that arrive after 7ms decrease proximity. I added a cross-fade window centered at 6ms in LOC to account for these perceptions."[snip]

Greisinger has been interested in identifying the auditory perception of what he calls "Proximity" and developing quantifiable measurements for it, in order to provide more useful acoustic tools for the design and optimization of acoustic performance spaces.  I've found one way OMT arrays differentiate themselves from more typical stereo recording microphone setups is that they are often able to provide a stronger perception of proximity as each pair is brought up in the mix.  Although my application is in some ways the inverse of what he is focused on, I see it as providing similar perceptual cues to the listener.  While DG is describing a perceptual attribute of the soundfield and how it varies at specific positions in a hall, and I am building microphone arrays partly intended to improve perceived sound quality when recording from a location that is somewhat more distant than would be desirable, I suspect OMT arrays may be leveraging the same perceptual attribute.  They may be in effect synthesizing proximity cues via the geometry of the microphone array itself.  For instance, the OMT8 array I'm using currently samples the soundfield across 7 positions in space within a 5-6 foot left-right horizontal-axis and about a 1-1/2" front-back axis.  That spacing correlates to a maximum time-of-arrival difference of 5 to 6ms for sound arriving from the sides.  Ignoring the rear-facing pair (intended to exclude front arrival sound), there are 5 sampling positions arrayed horizontally left-right which correspond to 3 stereo pairs contributing to front sound pickup: A center M/S pair, 5-6' spaced omni pair, and mid-spaced 90 degree angled supercardioid pair.  I am repeatedly struck by how muting of any one of those 3 pairs in the mix significantly reduces the perception of "proximity" and is one primary reason I've willingly if somewhat begrudgingly shifted from 6 to 8 channels.

I also see a possible parallel here with Tony Faulkner's 4 microphone phased array setup (not his older parallel 8's setup) and other "multiple microphones arranged in a line" techniques some tapers are using.

[Gutbucket]

Also noted on this visit to David Griesinger website is increased focus on making his headphone equalization technique I've mentioned before more widely available (DGSonicFocus now mentioned prominently at the top of the homepage) via an application used to help equalize headphones to match the hearing response of the listener. It allows individuals to find the precise equalization for their headphones which matches the sound pressure they hear through the phones to the sound pressure from a flat-frequency loudspeaker positioned directly in front of them.  I've mentioned this here previously, requested the program and received it along with other information from him last year with the intent of starting a discussion about it here at TS, yet still haven't found the time to explore it myself, much less start a new thread dedicated to it.  I remain excited to try it as it very much falls in line with my thinking on what is important with respect to the problem of accurate headphone reproduction.

"Our apps use a simple loudness comparison test to find an individual’s frequency response at their eardrums, both from a free-field source at the front, and from headphones and earbuds. The app then provides the precise equalization needed to give the listener accurate timbre and frontal localization without head tracking.   The difference in image and timbre with our equalization is startling.

The apps also provide a time domain transaural mode that reproduces the sound of [personally equalized] headphones [over speakers] without individual equalization. It was developed to demonstrate our binaural recordings of concert halls at acoustics conferences world-wide. When reproduced this way the recordings are startlingly real. ..One visitor who had hearing aids asked if it would still work. I said take out the aids and try the app without them. He was ecstatic. 'I can mix again with this!'"

In a similar if more complex mulitchannel vein, the Smyth A16 Realizer Kickstarter project finally reached completion a few months ago and is now shipping out to backers.  I'm something like backer number 325 of the original 330 or so from 3 years ago.  Word is they are shipping about 10 units out per week and are currently around number 120, so I'm hoping to receive one next spring perhaps.  Despite the long delay I've been looking forward to this ever since undergoing a personal HTRF calibration and demoing the original A8 unit at a CanJam headphone convention in 2008, and have gushed enough about that experience previously in these threads.

Although for it to work optimally as intended one must place miniature microphones in the ears and make a series of physical HTRF measurements of actual speakers placed in the intended positions around the listener in an actual listening room, this unit now also includes a simpler headphone equalization mode which appears to be based directly upon Griesinger's approach.  Some rather droll youtube videos of its operation, and less droll videos of demonstration reactions can be found here- https://www.youtube.com/channel/UC0yePwI_x1q8oACTVHpYxkw/videos