Skip to main content

Stereo recording from the center of a circle?

I’d like to make recordings of our group of musicians who play irish music. We get together every week and sit in a circle, in a private space (not a bar), and play Irish tunes. I’d like to avoid interfering with the group dynamic in our small circle and not be intrusive, so here’s my thought, let me know what you think. I could put two figure-eight mics in the center of the circle, head to head (one upside down above the other, at 90 degrees, similar to blumlein but we’re around all sides) and hopefully get a decent stereo recording. Other than making sure the guitarists are centered between both mics (its the only instrument with a low end) I’d just let everyone else end up wherever they happen to sit encircling the mics. Everyone maybe four or five feet from the mics in the center? We have two guitar-like instruments (really a tenor guitar snd a mandocello), three fiddles, and bagpipes (the mellow Irish kind of pipes called Uilleann pipes). The room might sound ok, it’s pretty big and we’re in the center maybe fifteen feet from any walls. The mics would be recorded through the two xlr inputs on my Tascam DR40 recorder. I’d cut out the sets that sound good and EQ them etc in Garageband (hey it’s free!) to be shared as mp3s. 

Is this a dumb idea for a recording setup? I ask because i will have to buy the microphones and that’s a pretty big financial misstep if my plan is flawed in some way. Let me know what you think.

The mics I’m considering could be chosen from the following models (unless you have a better option): the AKG P420, Audio-Technica AT2050, CAD m179, or Aston Spirit (all around $200 ea, all figure eight, keeping the project under $500 for two). 

I also am curious about the SE X1R ribbon mics though this would also require buying a preamp since the Tascam’s preamp is not great. And, I don’t really know if its worth the extra cost, but i love the idea of giving the whole project a warm smooth and detailed sound without the need for as much fussing on the computer afterward. that would be great! 

Our only experience with microphones is SM57s on a PA system for live performances, and honestly I think they sound pretty bad, especially on a violin. And, I’m not going to set up six individual mics like that every week for our sessions, anyway. 

Thank you for your time, and i look forward to reading your criticism and/or encouragement!

Comments

paulears Thu, 11/03/2022 - 00:28

This is really interesting. A Blumlein pair in the centre would give you coverage through the 360 degrees. Two ribbons would certainly ensure everyone gets recorded. In a rehearsal you would need to alter distance to deuce the loud instruments and perk up weaker ones.
 

I cannot see any reason to not try this, but there are a few important things to remember. It’s not stereo. Stereo, of course has a clearly defined left, right and centre. This will have a sort of reflected image. Imagine drawing the group onto a circle on paper. Then you’d fold the circle onto itself, making two people overlap in places they’re not really occupying. I tried it once with a small choir and it’s sounded really good, and it was in a small church so the acoustics were nice too. I only spotted the weird addition when I noticed two voices that stood out as a bit er, obvious, appeared to be very close together, annoyingly prominent. They were not together in the circle. 

it does produce good capture, but it’s not ‘real’ however, does that matter. I do know I would do it again but be more careful to monitor in closed headphones at a loud volume to try to spread people around. With your people you might accidentally overlay two instruments that shouldn’t be together (if you recorded normally).

sounds fun!

Thomas W. Bethel Thu, 11/03/2022 - 02:30

paulears wrote:

This is really interesting. A Blumlein pair in the centre would give you coverage through the 360 degrees. Two ribbons would certainly ensure everyone gets recorded. In a rehearsal you would need to alter distance to deuce the loud instruments and perk up weaker ones.
 

I cannot see any reason to not try this, but there are a few important things to remember. It’s not stereo. Stereo, of course has a clearly defined left, right and centre. This will have a sort of reflected image. Imagine drawing the group onto a circle on paper. Then you’d fold the circle onto itself, making two people overlap in places they’re not really occupying. I tried it once with a small choir and it’s sounded really good, and it was in a small church so the acoustics were nice too. I only spotted the weird addition when I noticed two voices that stood out as a bit er, obvious, appeared to be very close together, annoyingly prominent. They were not together in the circle. 

it does produce good capture, but it’s not ‘real’ however, does that matter. I do know I would do it again but be more careful to monitor in closed headphones at a loud volume to try to spread people around. With your people you might accidentally overlay two instruments that shouldn’t be together (if you recorded normally).

sounds fun!

One of the best recording I ever did was of a Bluegrass band in a circle. I used two M-49 Neumann mics on bi-directional and placed them in the middle of the group. Members of the band stepped forward to "solo"and stepped back when they were done. It worked great! I say go for it...

 

jjw Thu, 11/03/2022 - 04:24

paulears wrote:

This is really interesting. A Blumlein pair in the centre would give you coverage through the 360 degrees. Two ribbons would certainly ensure everyone gets recorded. In a rehearsal you would need to alter distance to deuce the loud instruments and perk up weaker ones.
 

I cannot see any reason to not try this, but there are a few important things to remember. It’s not stereo. Stereo, of course has a clearly defined left, right and centre. This will have a sort of reflected image. Imagine drawing the group onto a circle on paper. Then you’d fold the circle onto itself, making two people overlap in places they’re not really occupying. I tried it once with a small choir and it’s sounded really good, and it was in a small church so the acoustics were nice too. I only spotted the weird addition when I noticed two voices that stood out as a bit er, obvious, appeared to be very close together, annoyingly prominent. They were not together in the circle. 

it does produce good capture, but it’s not ‘real’ however, does that matter. I do know I would do it again but be more careful to monitor in closed headphones at a loud volume to try to spread people around. With your people you might accidentally overlay two instruments that shouldn’t be together (if you recorded normally).

sounds fun!

Thanks, yeah it wouldn’t be “true” stereo as the people 90 degrees on either side of you are on the other side of the “stage” and the person across is on the same side. Or you could all four end up center stage if you’re all in a crotch of the x pattern. I’ll have to come up with a seating chart! 

Boswell Thu, 11/03/2022 - 09:24

I've recorded string quartets using a crossed-pair of fig 8 ribbons (AT4081s) in Blumlein configuration. You have to be careful to place the players so that if they were to look in a mirror mounted axially at the microphones they do not see their opposite number in the middle of the mirror. In that way, you can get them spaced around a quadrant in the resulting stereo image, even though half of them will be out of phase (which matters not in this case).

You can also experiment with a small rotation of the microphone pair (not more than about 10 degrees) to fine-tune the positioning. A greater rotation than that starts to intrude into M-S Blumlein territory, which is indeed another way to go, but needs M-S decoding before mixing.

Try drawing it out on paper using an imaginary mirror with corresponding reflections. For an odd number of players, you could use marker pens to make lines from each player to the centre of the circle. You then fold the paper over through the centre point and find a crease position that keeps the lines maximally apart as they show through the paper. The crease should be 45 degrees from the axes of your microphones.

One advantage of the AT4081 is that it has a phantom-powered built-in buffer with some gain, so its output is similar to that of a quieter capacitor mic. A pair of 4081s would probably feed your Tascam without further external gain. However, they are not in the $200 category.

jjw Thu, 11/03/2022 - 12:53

Boswell wrote:

I've recorded string quartets using a crossed-pair of fig 8 ribbons (AT4081s) in Blumlein configuration. You have to be careful to place the players so that if they were to look in a mirror mounted axially at the microphones they do not see their opposite number in the middle of the mirror. In that way, you can get them spaced around a quadrant in the resulting stereo image, even though half of them will be out of phase (which matters not in this case).

You can also experiment with a small rotation of the microphone pair (not more than about 10 degrees) to fine-tune the positioning. A greater rotation than that starts to intrude into M-S Blumlein territory, which is indeed another way to go, but needs M-S decoding before mixing.

Try drawing it out on paper using an imaginary mirror with corresponding reflections. For an odd number of players, you could use marker pens to make lines from each player to the centre of the circle. You then fold the paper over through the centre point and find a crease position that keeps the lines maximally apart as they show through the paper. The crease should be 45 degrees from the axes of your microphones.

One advantage of the AT4081 is that it has a phantom-powered built-in buffer with some gain, so its output is similar to that of a quieter capacitor mic. A pair of 4081s would probably feed your Tascam without further external gain. However, they are not in the $200 category.

Thanks, that's super helpful!  Yeah I see what you mean about making sure we don't end up with a bunch of people on the same side of the stage accidentally! I like the idea of folding a paper circle to figure out where to point the mics in order to separate everyone's angle on stage, that's brilliant!

I won't be able to afford two AT4081s but I think I will try this with less expensive ribbons, even though they'll probably need a preamp - the Tascam can pass phantom power to something like a cloudlifter. As a fiddle player myself, the mic I own at home ought to be a ribbon, anyway. Thanks!

kmetal Thu, 11/03/2022 - 19:50

Is there any advantage to two figure 8 mics instead of a single mic in Omni? Or even a pair of disimilar mics in Omni? 

My concern with the figure 8 is it may lead to some parts of the circle that are nulled to some degree. If players move around this could cause some weird things, and possibly phasey stuff if I'm thinking about it correctly.

Boswell Fri, 11/04/2022 - 03:38

If you were happy with a mono result, using a single omni would certainly avoid any potential phase problems that might arise from using a Blumlein pair. But it won't get you a spatially-separated image, however contrived that may be.

paulears Fri, 11/04/2022 - 05:57

I've been scratching my head with a little diagram to see if there's obvious problems with the coverage, and I think it's just the mirror image of the rears, which sort of makes the crossover between the two mics tickle my brain. I suppose I need to actually put up a pair of fig-8s and walk around the circle and see what that sounds like. If it goes left, right, stays right then goes left, that is the real situation, but in my head it will go left right, left right ....... (one or the other). I'm off site at the moment, but next time I'm in the studio I will record it and see!

jjw Fri, 11/04/2022 - 10:10

Ordered two used SE X1r mics and a used cloudlifter, total cost only slightly over what I was comfortable paying lol, and we'll see what happens!

Yeah, I know the stereo image is contrived but I'm not replicating reality, I just want the listener to be able to tell the instruments apart. Irish tunes are unison melody, so good stereo will help a lot in separating all the fiddles and pipes so they don't sound like one bizarre loud instrument. I considered an omni but feel like it would come out muddled for this particular group and the genre in general. Part of the excitement is perceiving the unison melody lines with their own slight variations and how they all fit tightly together.

I'll test to see if there's a dead spot in the middle of the pattern but if people use this config for regular Blumlein (90 degree angle) with ribbons frequently, I'm hoping that means a full circle should also have full coverage. Also hoping that the ribbons I ordered have even tone/EQ at different subject angles - that is supposed to be better in ribbons than in cardioid large condensers (?) I'm told.

Thanks everyone, this is a really active and thoughtful forum!!!

 

audiokid Fri, 11/04/2022 - 12:35

Hey jjw, if you are comfortable doing so, when you get it all sorted….  please share any audio examples related. It’s the most fun part of the posts for me 😊
 

fun topic!

jjw Fri, 11/04/2022 - 13:32

Wait no, there might be a problem. Is this right? You can be in front of a Blumlein or behind it, but you shouldn’t be in either of the two “side” quadrants because then you’d be playing into the front side of one mic and the back side of the other and if they were ever blended to mono you’d  cancel yourself out! Is that right? You either have to play into two backsides or two front sides, right?

… On further reflection perhaps i should put the two guitars on the sides  and not tell them that there will be a secret trick to obliterate them!! 

jjw Fri, 11/04/2022 - 14:01

audiokid wrote:

Hey jjw, if you are comfortable doing so, when you get it all sorted….  please share any audio examples related. It’s the most fun part of the post for me 😊
 

fun topic!

Lol, I have very little shame and even less taste, so I will gladly share whatever mediocre product i manage to create! 

Boswell Mon, 11/07/2022 - 08:10

jjw wrote:

Wait no, there might be a problem. Is this right? You can be in front of a Blumlein or behind it, but you shouldn’t be in either of the two “side” quadrants because then you’d be playing into the front side of one mic and the back side of the other and if they were ever blended to mono you’d  cancel yourself out! Is that right? You either have to play into two backsides or two front sides, right?

… On further reflection perhaps i should put the two guitars on the sides  and not tell them that there will be a secret trick to obliterate them!! 

No, that's not correct.

Imagine you have a Blumlein pair of fig-8 microphones possessing a symmetrical response (equal amplitude front and back - some Royers don't) and set them up at right angles at the centre of a circle. Now set them to record while you carry a constant sound generator round the circumference of the circle, starting by facing the front of the L microphone. Then listen to the recorded result in stereo. The sound will move smoothly from L to R and back to L as you complete the tour. This is independent of whether you treat the mics as L-R or M-S (with decoding), although the acoustic starting point will appear as being in the centre for the M-S configuration.

This test shows that it is valid to place your performers anywhere around the circle and they will all be treated equally in the stereo recorded output of the microphones. Mixing the stereo image to mono is the same as not using the S mic in the M-S configuration, so any sounds that are purely at right angles to the mono M mic would not be heard because they are arriving directly side-on to that mic's fig-8 response. Because this is simply a property of a fig-8 mic, the same holds for X-Y, but shifted through 45 degrees in that case.

In saying all this, I'm ignoring the phase inversion from the rear sides of the microphones, because it plays no part in the explanation. However, if you were also recording a DI signal via a trailing lead from your constant sound box, you would be able to see the signal from each mic separately fade to zero and come up again with inverted polarity relative to the DI. Sines and Cosines!

 

kmetal Mon, 11/07/2022 - 20:00

Boswell wrote:

jjw wrote:

Wait no, there might be a problem. Is this right? You can be in front of a Blumlein or behind it, but you shouldn’t be in either of the two “side” quadrants because then you’d be playing into the front side of one mic and the back side of the other and if they were ever blended to mono you’d  cancel yourself out! Is that right? You either have to play into two backsides or two front sides, right?

… On further reflection perhaps i should put the two guitars on the sides  and not tell them that there will be a secret trick to obliterate them!! 

No, that's not correct.

Imagine you have a Blumlein pair of fig-8 microphones possessing a symmetrical response (equal amplitude front and back - some Royers don't) and set them up at right angles at the centre of a circle. Now set them to record while you carry a constant sound generator round the circumference of the circle, starting by facing the front of the L microphone. Then listen to the recorded result in stereo. The sound will move smoothly from L to R and back to L as you complete the tour. This is independent of whether you treat the mics as L-R or M-S (with decoding), although the acoustic starting point will appear as being in the centre for the M-S configuration.

This test shows that it is valid to place your performers anywhere around the circle and they will all be treated equally in the stereo recorded output of the microphones. Mixing the stereo image to mono is the same as not using the S mic in the M-S configuration, so any sounds that are purely at right angles to the mono M mic would not be heard because they are arriving directly side-on to that mic's fig-8 response. Because this is simply a property of a fig-8 mic, the same holds for X-Y, but shifted through 45 degrees in that case.

In saying all this, I'm ignoring the phase inversion from the rear sides of the microphones, because it plays no part in the explanation. However, if you were also recording a DI signal via a trailing lead from your constant sound box, you would be able to see the signal from each mic separately fade to zero and come up again with inverted polarity relative to the DI. Sines and Cosines!

 

Great post Boz!

paulears Wed, 11/09/2022 - 06:54

setup2setup1I put upa couple of 414's 90 degrees to each other, one above the other and recorded the two outputs. One panned left and the other right. Walking around the two mics, starts left, and rotates through to right in 180 degrees of travel and then just repeats. Looking at the file in the phase meters in Audition shows that if you walk around the two mics in  360 degreemotion, the fig-8 pickup repeats twice, so it goes around two complete left to right cycles for one complete walk around - so recording instruments in a circle means you would have to be quite careful where they are positioned or the aural perspective could be of two sources sharing a single location. I'd not thought of that. with speakers and headphones Blumlein positioning of two figure 8 mics does mean that you can hear positions changing, plus see them easily on stereoscope/phase type displays interesting how the room sounds too. my feet on the hard floor and cables are also captured, and the voices are actually upstairs and much more obvious in the recording than in the room, and there is a fan noise clearly captured.

I'd love to try recording this with musicians. It also made me wonder what you could do with a cluster of four cardioids, in the centre, recorded to a 4 channel mix?

 

  • rotations-blumlein.mp3

Boswell Wed, 11/09/2022 - 08:06

Thanks, Paul. That was a good demonstration of what I was saying earlier, both in terms of the amplitude of the outputs and in the instrument positioning. 

Did you try listening to the result on headphones? I'd be interested to know if the sound appeared to go across behind your head in the 90 to 270 degree half of the traverse, where both phases are inverted in the sector from 135 to 225 degrees.

kmetal Wed, 11/09/2022 - 16:58

Interesting stuff!!!!

 

Would a perhaps "safer" option be one mic or pair of Omni mics? Using just the single Omni mic you could use post processing like delays, verbs, eqs that can pan based on frequency, even some audio to midi for sample replacement, and iZotopes music rebalance, all to isolate/reinforce/pan the instruments a bit after the fact, while retaining whatever the natural phase relationships are between the instruments at the mic location?

paulears Fri, 11/11/2022 - 23:44

Boswell wrote:

Thanks, Paul. That was a good demonstration of what I was saying earlier, both in terms of the amplitude of the outputs and in the instrument positioning. 

Did you try listening to the result on headphones? I'd be interested to know if the sound appeared to go across behind your head in the 90 to 270 degree half of the traverse, where both phases are inverted in the sector from 135 to 225 degrees.

Yes, and I tried it on DT100 and some open back Sennheiser, and it stayed in front. It did surprise me that it did the same thing sort of twice. I had, in my head expected the reverse side to be reversed, but it went left right left right in the complete circle. This makes sense but i sorry of thought it would do left right right left, if you get me? Watching the stereoscope was interesting using audition, not Cubase as the 360 display has a defined area and a spot where it thinks the location is and the area shape constantly changed, BUT the spot  did go the 360 circle, as in as you said, behind the mic. It also shows a location as the spot close to centre or to the edge. If you use pan and level to create location, this spot is closer to the edge of the circle, mono in the centre spot, but a Blumlein shows as a rotating circle more mono, than nearer the edge, presumably the room?

x

Register