Topic: Wide cardioid/subcardioid spacing and angle

[wforwumbo]

To address OP first: I have run my subcards between 27 cm up to 1 meter, angled between AB (0 degrees) and 110 degrees. For amplified music, my ears prefer a spacing of 30 cm and anywhere from 30-75 degrees. The most commonly used mount in my gear bag is 30cm/60deg. I can try to get you some tapes if you'd like, off the top of my head here's one already on archive:

https://archive.org/details/fruitbats2023-05-17.mk21.flac16

My experience is that my subcards really want to be on axis with the PA. I got em dead on into a v3 for a phish show in 2022 and that tape came out sublime - I'll upload a song from that to give you an idea of what these narrower angles and spacing give with subcards.

^^^^^^^
Not to go too far OTS, I should have said: psychoacoustically vs acoustically
I was trying to explain to the OP why he may not perceive bass differences the way he does higher frequencies.

So I agree here with you and is what I was speaking to: "Rather, I think the issue is room acoustics behavior and human hearing perception.

As to what you say about human hearing capable of finding distinction, or "stereo imaging" lower than 100Hz, my experience agrees with that. However, there is a fair amount of literature which claims "trained listeners" will distinguish characteristics in similar musical pieces which "untrained listeners" may not. A very well regarded FOH engineer told me "it's not always what you hear, but what you don't hear I find fascinating"     

To allow a healthy dose of skepticism, that literature always uses stimuli we do not find in nature, in controlled conditions. Imaging below 100 hz absolutely exists - as evidenced by the fact that no engineer sums their mixes to mono below 100 hz. The idea that no imaging or spatial cues happen below 100 hz comes from mix engineers who discovered that panning bass and kick drums would cause weird phase cancellation in the stereo fields and appear to make those instruments weaker. That's more of a coherence thing, and binaural coherence is much more complicated of an argument than stereo/binaural cue perception at low frequencies. Keeping bass panned center helps prevent odd phase artifacts from muddying or confusing the mix.

[Gutbucket]

[...snip] For amplified music, my ears prefer a spacing of 30 cm and anywhere from 30-75 degrees. The most commonly used mount in my gear bag is 30cm/60deg.[...]

^Was thinking of this last week while posting, hoping you might join in.

There is a second aspect in play which that may hint at. What's uncertain is how much of each is attributable to one's ear and preference.

I feel one side of the mic pair spacing "coin" is image distribution, sense of space and how diffuse the reverberance sounds, and the other is how spacing effects frequency response in a secondary, stereo aspect kind of way.  That is, the frequency response of each individual microphone remains unchanged regardless of how far apart the pair is spaced, however, the response of the pair in sum is partly effected by the spacing between them via interaction with the vector of the wavefront.. and maybe with the distribution of modal pressure nodes. I say secondary stereo sense because unless some processing step along the way includes some sort of summation of the two independent stereo channels, this summing does not actually occur until stereo playback, and happens "in air" with speakers and maybe "in head" with headphones.  Its a ripple in the summed response due to comb filtering and is definitely audible, but is unlikely to be clearly identified as a frequency-response thing unless actively listening while making a change in spacing, in which case it is quite easily heard.  Its tangentially related to the mono-compatibility issue.

I've long wondered how wide the variance in taper preference might be in determining one's preferred spacing for a pair of mics, given this (conscious or not) weighing of the two different aspects.

tl/dr: The effect of mic pair spacing is tonal as well as spatial.

[wforwumbo]

So I can do one better for this thread... I've uploaded two recordings of the same show, from the same spot, in the same configuration (30cm/60 degrees), with different microphones/chains. The mk21 subcardioid recording is Schoeps mk21 (30cm/60 Degrees) > Schoeps KC5 > Schoeps CMCC5 > Grace Lunatec V3 > Sound Devices MixPre-6 (aux in, channels 5/6 @ 24/48k); the mk4 cardioid recording is Schoeps mk4 (30cm/60 Degrees) > nbob KCY > Naiant PFA > Sound Devices MixPre-6 (mic in, channels 3/4 @ 24/48k). The mics were clamped right next to each other, one atop the other. This is Phish's Walls of the Cave from 9/3/22.

https://drive.proton.me/urls/AMXPD3GHZR#uj1F0LbjCKbj

This is my favorite mk21 tape I've made so far to-date.

[Gutbucket]

That's certainly most helpful to the OP. That recording and our discussions about it are exactly what I had in mind!

But in regard to the summed response ripple thing, since both pairs used the same spacing that will produce the same frequency response effect on both of them.

What I'd really love to hear is the same show, recording position, mics, and angle, but several different spacings.  Even then it might be difficult to clearly identify the difference in response attributable to the ripple in the summed response, since the change in spacing will also effect the imaging and diffuse field correlation simultaneously. 

Now if someone were to somewhat rapidly slide one of the mics along the bar during the recording, the summed response ripple thing would be immediately obvious to any listener.  I say that because I did this with omnis years ago to get a better empirical understanding of how differences in spacing effects the sound and image.  During the relatively rapid change of position the ripple becomes obvious from the flanging effect of the movement shifting its frequency centers. After becoming aware of it in that way, the difference in static response between various positions becomes more easily perceived.  But actually doing that means intentionally sacrificing a recording.  Maybe something to try at a meh opener.  ..or maybe it works the same when testing in front of the stereo, I've not tried that.  Best to make a relatively swift slide between the various spacing positions (that's what makes the flanging effect obvious) noting each position, either verbally on the recording or via time-stamps, and sticking with each position long enough to get a good feel for the sonic differences of the other aspects - the difference in static response, imaging, and diffuse correlation.

Curious about your thoughts about all that.

Years ago there was a tread here at TS about a few different stereo pair robot mic bars able to remotely adjust angle/spacing.  Not really practical use type things but more like university project type stuff.  I recall some youtube videos which included sound and video of one shifting back and forth between configs - you could see each config and hear the sound the entire time.  I may search for those and see if I can hear the summed response flange effect while the mics are moving and the static response differences when still.  If I find a good example I'll post a link here. As I recall most were changing angle at the same time though, complicating things.

[DuctTaper42]

     

Really appreciate the discussion, off topic and on, here. Lots to mull over and explore further but this is shaping up be a nice little repository of info that accomplished exactly what I'd hoped of not only helping me get a good idea of what you seasoned vets do in practice with this particular, perhaps less common, polar pattern, but also creating a easily searchable thread for future learners to find without having to dig through numerous threads in search of those nuggets of gold. Y'all rock 

[DuctTaper42]

And as my own attempt to add this repository, there is significant discussion of spacing and angle in the Team Line Audio thread here: https://taperssection.com/index.php?topic=172445.0 especially since the CM3/4 are much more commonly used on this board than the OM1/Omni1.

A particularly detailed response from Voltronic:

What's generally the preferred pattern for CM3? I have a pair I've never used but can pickup a Shapeways.

Most of the classical people on GS like NOS (30 cm / 90 deg), and that's what I started with as well. Now, my go-to is the SRS "Hybrid" which is the same 30 cm but opens the angle up to 110 deg.

If you use the Sengpiel visualization tool, you can see the SRAs for these arrangements. Select NOS, then change the mics to Subcardiod. For the "hybrid", select ORTF (note the SRA), change to subcardioid, then the spacing to 0.3 m. You'll see the SRA is very close to ORTF with cardioids.

To give you an idea what these sound like, here are some of my samples. NOS was used for Ubi Caritas. Hybrid was used for Road Home, Rise Up, Joyous Occasion, and Innisfree. I think I used "Wide ORTF" for the rest, which is only 21.5 cm at 110 deg. I don't use that spacing anymore, as I found the wider spacing to sound better with these mics.

Keep in mind that this what I recommend for acoustic ensemble recording, which is what I do. For amplified music, other people here will have better advice.

[wforwumbo]

I heartily agree with Voltronic at a high level, and like him I love using the sengpiel visualization tool when explaining stereophonic zoom. Stereo zoom in my opinion is the most effective tool I can effectively experiment with when taping. Once I have a mic pair pattern and set up under my fingers, it's simply a matter of deploying what I think works in a given taping set up.

There is a significant difference in the equation when we are discussing amplified music - and the room acoustics of spaces normally used for amplified music. Recording the sound field of unamplified music and how we binaurally perceive the performing ensemble/the room acoustics is a very different deck of cards than capturing what is effectively an oversized stereo in enclosed space (we are only rarely blessed to operate in a free field, and this is precisely why subcards and Omnis shine in big meadows).

I'll also reinforce that every taper has our own personal preference, and that's okay too. If there were a "right way" to do things, we would all agree and trust one taper to run their mics and the rest of us would be one large patch chain.

And as my own attempt to add this repository, there is significant discussion of spacing and angle in the Team Line Audio thread here: https://taperssection.com/index.php?topic=172445.0 especially since the CM3/4 are much more commonly used on this board than the OM1/Omni1.

A particularly detailed response from Voltronic:

What's generally the preferred pattern for CM3? I have a pair I've never used but can pickup a Shapeways.

Most of the classical people on GS like NOS (30 cm / 90 deg), and that's what I started with as well. Now, my go-to is the SRS "Hybrid" which is the same 30 cm but opens the angle up to 110 deg.

If you use the Sengpiel visualization tool, you can see the SRAs for these arrangements. Select NOS, then change the mics to Subcardiod. For the "hybrid", select ORTF (note the SRA), change to subcardioid, then the spacing to 0.3 m. You'll see the SRA is very close to ORTF with cardioids.

To give you an idea what these sound like, here are some of my samples. NOS was used for Ubi Caritas. Hybrid was used for Road Home, Rise Up, Joyous Occasion, and Innisfree. I think I used "Wide ORTF" for the rest, which is only 21.5 cm at 110 deg. I don't use that spacing anymore, as I found the wider spacing to sound better with these mics.

Keep in mind that this what I recommend for acoustic ensemble recording, which is what I do. For amplified music, other people here will have better advice.

[Gutbucket]

^ And that's what Improved PAS is all about.  It's just Stereo Zoom applied to the taper situation of recording amplified music from an audience position.

All it really does is suggest the optimal spacing between a pair of mics that are pointed directly at the PA speakers. 

[Gutbucket]

All it really does is suggest the optimal spacing between a pair of mics that are pointed directly at the PA speakers.

To tie this in to the somewhat obtuse blather I was spouting earlier in the thread - "optimal" in this case refers to clarity, direct/reverberant ratio, image distribution and sense of space.  Stereo Zoom does not address that secondary stereo summed response thing at all, which is going to vary along with the different spacings it suggests. 

I personally feel clarity, direct/reverberant ratio, image distribution and sense of space are more important (putting me squarely in the Stereo Zoom camp), but different ears may value different aspects, which is why I mentioned the secondary effect on response.

[checht]

Very interesting theoretical basis for various techniques. I'm gonna just answer the initial tldr:

Howdy folks,
TL;DR I was wanting to inquire the hive mind on techniques y'all use for wide cardioids.

Outdoors and indoor venues with good sound/room, I run my mk22s OMT4, 4' spaced with mk41s x/y PAS in the center.
Indoor venues with questionable sound, I run the 22s wide ORTF.
Love the results. Simple enough even for me.

[vantheman]

This may be veering off topic, but since I use subcardioids a lot, and stage lip recording was mentioned, I’ll go for it.
How do you know when you should space your mics much wider that a common stereo config like DIN or ORTF? Someone mentioned spacing the mics several feet and aiming for the snare. Are there certain factors that should be considered such as the distance of the instruments/amps on stage? I’ve generally been happy with my “wide ORTF” stage lip recordings with the CM4s, but the center part of the stereo image can get a bit loud to my ears and I’d like branch out. I’ve been afraid to try anything else because as someone else said, on the day of the show I kinda just want to nail it. I’ll probably run on stage for Jazz is Dead in a few weeks and this might be a perfect opportunity to try out a more distant spread.

[wforwumbo]

You have two pairs of mics, and four inputs on your deck, per your sig. Use one pair as a known baseline, and the other to experiment. This is why I always run more than one pair of mics: I know I'll always like the mk4 in NOS even from compromised locations, and I can experiment with getting the 21 and/or 2 right for unknown venues, acts, and locations. Any experiment involves a control, otherwise we are shooting in the dark.

If you feel your recording is bunched up in the middle, you need to widen the coherence (similar to crosstalk) between the left and right channel. This is accomplished by narrowing the angle between mics and increasing the distance between your microphones. I too found wORTF to sound a bit bunched up in the middle as well as lacking in bass response compared to NOS, which is why I use the 30cm distance as my standard. I find that going larger than 30 cm increases bass response but at diminishing returns, so I call 30 cm "good enough for government work" and vary angle to taste based on experience. I encourage you to give NOS (30cm/90deg) a shot, it's a standard for a reason and I suspect you will pull a recording to be proud of with it even on stage or stage lip. No need to point a mic at the snare, simply center the auditory image.

One thought for stage lip is that you want to be far enough away from instruments as well as monitors such that they don't dominate the response too much, and so that they don't end up hazing your image. It seems counter intuitive, but there is such a thing as "too close" and sometimes a dash of room rather than direct sound can lock in and smooth over an auditory image.

[vantheman]

Thanks, this is probably the best course. I could buy a second set of CM4s to be able to make real apples-to-apples comparisons and sell them off once I've gotten the gist.

[checht]

While reading this thread, I'm listening to a rough mix of yesterday's Los Lobos show at at 300 seat outdoor venue.

Ran omt4 with mk22s 4' split plus mk41s x/y pas. Back home, I ran the 41s track through rx to separate out the vocals.
Rough mix w everything normalized: 22s even, 41s 3db down, vocals 2db down.

Gotta say this method is yielding my personal favorite sound of 40 years of recording. Wide, solid soundstage. Detailed, present vocals and guitars, and the whole thing ends up sounding quite natural, just right.
Of course, this is only personal taste, but I'm doing the work, might as well have it come out to my spec.

The band kindly offers a patch to known tapers, and though I record it, I often don't use it b/c this config/method is so pleasing to me.

[Gutbucket]

I’ll probably run on stage for Jazz is Dead in a few weeks and this might be a perfect opportunity to try out a more distant spread.

Delicious opportunity!

You said, "the center part of the stereo image can get a bit loud to my ears and I’d like branch out".
^ Okay, working with that, and differentiating between the level of the center in comparison to the rest of the stereo image, as contrasted with image distribution, which would be more like instrument sources sounding overly clumped together in the center rather than being spread out more evenly..

If you feel the center has a bit too much level, you might try any one of the following to correct for that:

1) Record using the same mic config, then try making a slight mid/side ratio adjustment afterward, reducing the level of Mid and increasing the level of Side.  This quite literally decreases the level of the center portion of the image.  It will also make it sound wider because there is less center content than before.  Tread lightly and listen carefully when you try it though.  Too much is going to sound bad (due to the classic non-coincident pair mono-compatibly problem), but just enough might do what you want without apparent audible problems.  This option keeps your mic config unchanged, making it easy to nail it in the same way you've done previously on the day of the show, but first give it a try on one of your past recordings using this same mic config that sound a bit too center heavy to you to determine if it is going to do what you want without making things worse.

2) Angle the mics apart more but keep the spacing the same.  This reduces the level of the center due to less pickup pattern overlap - sound sources near the center arrive at each of the two microphones somewhat more off-axis where they are less-sensitive. This is actually similar to 1) in what it does, except for being implemented at the physical microphone level, which avoids the potential mono-compatibility problems of changing the M/S ratio afterward. The change in the microphone configuration itself produces less Mid content and additional Side content. This should work well for you when recording on-stage or at stage lip with the CM4s for a couple reasons: The off-axis response of subcards tends to be smooth and natural, so even though the mics are pointed farther away from the sources in the center the center part of the image should still sound clear and natural.  And, when recording on stage or at stage-lip, the sources will tend to to be spread out across a wider angle, so a wider mic angle makes sense because the mics are still mostly on axis to the ensemble, unlike a recording position far back in the room where the mics would end up pointing at the side walls.  This is may be the best way to go because it's a better way to approach it than 1), yet doesn't alter your current microphone config much, making it simple to walk in and nail it the day of the show.

3) Space the mics apart significantly more, and angle them less.  This actually increases pattern overlap, yet can still reduce the level of the center because each microphone is further away from the central sound sources.  It will also change the image distribution, pushing sources that were near center a bit farther out toward to the sides of the playback image, which perceptually can sound like there is less level from the center.  This one sort of takes advantage of the proximity and wide positioning of the sound sources on stage - a quite different acoustic situation than a recording position out in the audience.  The decreased angle between mics will also reduce level of audience and room sound somewhat.  You will definitely need to use more spacing between the mics to compensate for the narrowed angle between them, because otherwise the center image would get louder rather than quieter [edit to correct oops] as with 2).  Additionally, the wider spacing will provide improved audience and room sound - not only will the mics be somewhat less sensitive to audience pickup due to their angle, the larger spacing will help de-emphasise pickup of individual audience members, and the room/reverberant pickup will tend to be more open and natural sounding.  Implementing this is going to be more of a gamble though, because you are changing both angle and spacing, balancing one against the other.  It's also a bit more complicated in that it may require two separate mic mounts (two clamps or short stands).  But when you do nail it, it can pay off nicely in this kind of situation. 

4) Get the advantages of 3) and a lot more, with way less gamble, by using 3 or 4 microphones total, placing the third mic or a 2nd pair in the middle - this is what Checht is talking about.  It does require more stuff and work, but dramatically increases the likelihood of nailing it.  To properly integrate the additional mic(s) in the center you'd need to significantly increase the spacing and/or angle of the wide pair, more so than in 3).  The additional mic positions and increased spread/angle of the wide pair will provide a more even pickup of all sources on stage AND provide you with a small but welcome degree of control over instrument and image balance afterward.  IME, in this way you gain a lot more leeway in setup since the actual spacing and angles between microphones are less critical, while achieving both a tight, well-anchored center AND a nice wide image width and sense of space at the same time (aspects that are otherwise often contradictory), as well as gain the ability to make the center louder or softer afterward, allowing you to dial it in just right.