Setup for recording/mixing two choirs

[two2the8]

Hi all,

I'm new to the forums, so I hope this is the right place to post this. 

I'm a beginning ethnomusicologist, and I'm about to embark on a long field-recording project that will often involve recording two choirs that take turns singing to one another. The choir that is not singing at any given moment, though, is often very chatty (and I can do very little to control this). I've been thinking that I would like to record the two groups separately so that I can reduce the chatter when making a mix later. I have done some previous work with similar choirs, and I really like how they sound in stereo, so I would also like to make stereo recordings for each group. 

I am just beginning to think through my kit and approach, and here's what I'm thinking so far:

Set up two stereo rigs, one for each choir (maybe using a single-point stereo mic like the rode nt4 or bp4025, for simplicity). Set the rigs up facing opposite one another, to maximize rejection of the 'off' group. Use a 4-track recorder to capture each stereo pair separately. Later, when mixing, duck the volumes on the 'off' group with each alternation, and mix down to a clean stereo recording. 

Assuming I've explained my problem clearly, does this approach seem feasible? Are there other/better/smarter approaches that I'm missing?

Also, a related question: say I've recorded each group as an XY stereo pair, with each channel mixed hard-left and hard-right. When mixing later on, can I narrow the respective stereo field of each group? For example, can I place group A on the left side by mixing them hard left & centre, and then place group B on the right side by mixing them hard right & centre? Or will narrowing the field like that destroy the stereo effect for each group?

Thanks very kindly for any and all advice you can give. 

[Philip Perkins]

If they will let you record the choirs separately (like with the other choir not in the room) then that would be favorite for sure.  Even with directional mics you'll never be able to not pickup chatter coming from the other choir.  Question: are the choirs really that undisciplined that they can't be respectfully quiet while the other group sings?    My other question is if there is video involved, which usually limits your flexibility in mic placement and might require that the non-singing group be present?   Mic etc-wise here are many ways to skin the cat--my fave for choirs is either ORTF or (if you have more gear), an ORTF pair with omni outriggers.  

[jhharvest]

Is the 'chatter' from the other choir part of the narrative? When I lived in Tanzania the maasai had these callbacks, little quips they kept making about the other boys and if I had been filming it I definitely would have recorded that as part of the whole experience. But if it's just chat about what they had for lunch, then maybe not.

I usually use MS + wide AB for a lightweight choir setup, the mid mic from the prior can often be used to patch the soloist (assuming they are in the centre). If you plan on mixing left-centre, centre-right for the two choirs I would also suggest taking a long strip of quiet room tone, although I feel like that kind of stereo imaging might be distracting especially if the pieces are long.

[two2the8]

Thanks for the replies!

@Philip Perkins, that really would be ideal, but unfortunately in this case that's not really an option most of the time. For the most part I'll be recording singing that is associated with ritual events, and the singing is largely improvised, so I can't really do it over again or record the groups one at a time. The choirs in these events are non-professionals, and the singing is not really a performance, it's more part of the festive ritual atmosphere, so the singers are not particularly concerned about quiet. 

I do love ORTF, though—I find it a bit of a pain to set up in the field, but I think it sure sounds *great*. My concern about ORTF in a setup like I describe is that, if the groups are sitting opposite one another, pairs of ORTF cardioids will each pick up more of the 'off' group than pairs of X-Y (or M-S) would. (At least, I *think* that's the case, when I plot it out in my head. It's a bit confusing).

@jhharvest, sometimes, ya! The chatter can range from being interesting commentary on other singers, to folks practicing bits to sing next, to singers conspiring together about a verse, to mundane talk about lunch. There's often coughing and the like in the off-choir, too, which I'd like to minimize. (I do usually shoot video, too, but I'm not too concerned about keeping the mics out of the frame. If it's easy to do, then sure, but since it's principally documentary I mostly don't mind). 

The pieces are typically quite long. Do you have thoughts about ways to mix the two choirs that wouldn't be so distracting? Left-centre, centre-right is really just an example I used to explain, but I actually have been thinking about starting with something more like L-CR & CL-R, giving a lot of overlap to the two groups, but keeping them just a little bit spatially distinct. In any case, from your comments it sounds like it's possible to muck about with the stereo width after the fact without making too much of a mess? My previous recording gigs have all either used two channels straight to stereo or multiple mono ISOs, so I've never really had to wrestle with adjusting stereo width in post. 

[Philip Perkins]

Are you after two separate stereo images of one choir each, or one big image that is the whole event?  The ORTF pair will not pick up significantly more of the other group (behind it) vs XY--it allows you to get a broad stereo field while not being all that far away--possibly an advantage here.  Since there isn't anything you can do about the chatter I'd be tempted to mic the whole deal up as one event instead of 2.  As you said, a 1 point stereo mic can be very fast to set up (or 2).

[jhharvest]

2 hours ago, Aaron Pettigrew said:

In any case, from your comments it sounds like it's possible to muck about with the stereo width after the fact without making too much of a mess? My previous recording gigs have all either used two channels straight to stereo or multiple mono ISOs, so I've never really had to wrestle with adjusting stereo width in post. 

Yep, you can adjust stereo width in post. Especially if you end up doing MS, you can alter the pickup pattern entirely electronically in the decoding. ORTF does sound more natural than MS but gives you less flexibility after the fact. Having that (omni) AB pair would let you widen the image and add more space if you should need it though. Depends on how large the choirs are as well and the acoustics of the space, of course.

It's hard for me to give you a suggestion on the mix without knowing the material - but probably if it was me I'd just try a normal stereo image on both choirs first, with fade ins and outs and test out if they sound distinct enough, without using this kind of stereo trickery to separate them. I don't think I've ever heard a recording with long pieces mixed off to the side, even in the beginning of stereo recording when people experimented a lot more.

Either way, sounds like a cool project. Hope you have a great time!

[two2the8]

1 hour ago, Philip Perkins said:

Are you after two separate stereo images of one choir each, or one big image that is the whole event?

Well, I guess I'm trying to use option a) to get to option b), but maybe that's silly. I want to record both groups relatively cleanly—like mic'ing single instruments for a live band, but in stereo—and then use those clean recordings, created separately, to create a mix that gives an impression of the overall event.

I have done a bit of experimenting with throwing up just one ORTF on the whole event, and while the stereo image is really nice, the chatter in the off channel gets really overwhelming at times. 

 

1 hour ago, Philip Perkins said:

The ORTF pair will not pick up significantly more of the other group (behind it) vs XY--it allows you to get a broad stereo field while not being all that far away--possibly an advantage here.

That's good to know, thanks!

@jhharvest, thanks! Great to know, and thanks for the advice on mixing. I guess I'll do some experimenting before I go out & see what sound right.  

[The Immoral Mr Teas]

For relative simplicity, whilst being sure of getting a reasonable stereo pickup of each group, and assuming they're facing (against) one another, I'd go with recording four-track (onto a Tascam DR70 or better) with four capsules - starting in an IRT type cross (so two opposing stereo pairs) and if necessary widening these (ie to two ORTF) or narrowing, to alter the recording angle and/or pick-up/null.

In post, you could then easily 'mix' down to 2 track stereo, using the 'quiet' pair as 'ambience-response' bleed into the other channels, or mute completely if desired, or indeed leave the lot open as surround quad.

BUT - I'd ask myself all sorts of questions first on what you're trying to achieve, is picture involved, and if any release or presentation of the material is expected at the end of the project (CD? Gallery?)

It also depends somewhat on what mics/recorders/accessories you already own, what budget if any is available and whether you're willing or indeed wishing to spend to 'upgrade' kit to make the project possible. My first 'upgrade' (if necessary) would be to get a 'decent' 4 track recorder - which would be the DR70 at least or something better (ie not an H4 or DR60 where you can't easily set up 4 matched channels).

Two Rode NT4s (or similar XY mics) piggy-backed in opposite directions would give you a similar useful recording setup - not precisely set as IRT / ORTF or custom positioned pairs and quads, but probably perfectly decent for the job in hand (and with the added benefit it could fit permanently in a fat rycote). It's also the kind of setup that I would consider B-format for: this type of ethno' choir (group rather than soloists) is already a 'wall of sound' and you also have the surround element ... but I expect this is way out of your budget (and if you aren't well versed already in ambisonics, MS, recording and mixing the results may end up patchy). I'd go with what you originally suggested: two pairs to four track.

Jez

[The Immoral Mr Teas]

On 04/06/2016 at 10:19 AM, Aaron Pettigrew said:

I actually have been thinking about starting with something more like L-CR & CL-R, giving a lot of overlap to the two groups, but keeping them just a little bit spatially distinct. In any case, from your comments it sounds like it's possible to muck about with the stereo width after the fact without making too much of a mess? My previous recording gigs have all either used two channels straight to stereo or multiple mono ISOs, so I've never really had to wrestle with adjusting stereo width in post. 

L-CR and CL-R sounds interesting ... but you'd really have to experiment with the CL and CR to get nulls against the 'wrong' side (perhaps with hypercardioids or bi-directionals), the recording angles natural, and the whole to work together. This would be difficult at best just with one fixed recording position ... trying to set it up with several different recording spaces could be terrifying.

Re stereo width and post - you should really look into (and experiment with some existing unmixed XY recordings) deriving M-S channels out of X-Y recordings. Of course the best course of events is to record the 'perfect' stereo angle and reproduce it, but I rarely ever worked on a studio music project without some MS appearing at some point, especially for disc (whilst many here know of my loathing of MS in theatrical film!)

Jez

[two2the8]

Wow, Jez (@The Immoral Mr Teas), thanks for the great response. Reading about B-format is blowing my mind!

2 hours ago, The Immoral Mr Teas said:

BUT - I'd ask myself all sorts of questions first on what you're trying to achieve, is picture involved, and if any release or presentation of the material is expected at the end of the project (CD? Gallery?)

It also depends somewhat on what mics/recorders/accessories you already own, what budget if any is available and whether you're willing or indeed wishing to spend to 'upgrade' kit to make the project possible. My first 'upgrade' (if necessary) would be to get a 'decent' 4 track recorder - which would be the DR70 at least or something better (ie not an H4 or DR60 where you can't easily set up 4 matched channels).

Well, my primary aim is to create good documentation for archival purposes and for music analysis, but I am also interested in working toward some kind of limited release CD/DVD publication if I can manage to get good recordings. Galleries/museums are an option, too. Most of the market for this stuff will be fairly local to the area where I'm working, which is pretty low-tech—not much likelihood of rendering ambisonic recordings in surround-sound. 

Fortunately I don't own much equipment yet, so I can kind of build up my kit with this project in mind. All my previous work in field-recording has been working with other folks' gear, which has been a wonderful luxury & a great learning experience. For myself, I'm looking at multi-track field recorders & mixers: I'm seriously considering the F8, maybe with a 422 in front for the nice preamps, or if I can get the scratch together maybe a used Zaxcom Maxx or 744 or something like (I wish the 552 recorded four tracks!) If you have thoughts about others to consider, I'd be interested to hear them. 

For mics, I'm considering two of the NT4s or BP4025s, for ease of use as I mentioned, but I'm a bit torn as to whether to go with integrated stereo mics or to pick up single mics which could prove more versatile. A hard decision—I'll have to do a bit of testing to see how I like the sound of the stereo mics before I can decide.

2 hours ago, The Immoral Mr Teas said:

L-CR and CL-R sounds interesting ... but you'd really have to experiment with the CL and CR to get nulls against the 'wrong' side (perhaps with hypercardioids or bi-directionals), the recording angles natural, and the whole to work together. This would be difficult at best just with one fixed recording position ... trying to set it up with several different recording spaces could be terrifying.

Re stereo width and post - you should really look into (and experiment with some existing unmixed XY recordings) deriving M-S channels out of X-Y recordings. Of course the best course of events is to record the 'perfect' stereo angle and reproduce it, but I rarely ever worked on a studio music project without some MS appearing at some point, especially for disc (whilst many here know of my loathing of MS in theatrical film!)

Jez

This is all very helpful, too. I will look into making MS out of XY—that's an intriguing option for fixing imperfect setups later. Is the idea that you re-encode the two mics to approximate a centre and a figure 8 instead of the X & the Y? That's neat. If you're up for it, I'd be interested to know more about the challenges of L-CR & CL-R that you mentioned.

Thanks again!

[The Immoral Mr Teas]

The F8 is probably an excellent piece of kit for what you have in mind, and capable of quite a bit more. One of the few things it lacks for its pricerange is any digital input, but since it records 8 tracks up to 192k this probably isn't a problem for you. By accounts here, the preamps are probably as good or on a par with the SD442, except probably not the limiters or headphone amp - but I don't know. The 552, whilst just being stereo, has five direct outs as well as some kind of digital out (which might help sync to something like a DR680).

Not really sure I can advise on mics - I tend to use Sennheiser MKH 8040 cardioids but they're beyond your price range. I used to have an AT822 which is a decent well built single point stereo (cheaper model than the one you mention). The audio tourism guru David Lewiston used Neumann KM74 (precursers of the eventual KM184) and Electrovoice RE50s for loud Bali stuff (I'm watching the BBC recent doc 'Zoo Quest in Colour' set in Bali/Komodo as I type). I'm a believer in practicing the art with 'decent cheap' (said AT822, Oktavas perhaps) then eventually forking out for real quality - but buy what you're comfortable with and will do the job well. The MKH ranges for instance are battletanks in all conditions as well as sounding excellent, which is why I take them anywhere. I love also the DPA 4060 miniatures.

Ambisonics is fascinating and the Soundfield mics are very musical, but it is what it is (and expensive at that). It is not a magic box nor a universally useful tool. And MS is NOT an 'option for fixing imperfect setups later' (unless perhaps you're mastering a disc).

I strongly recommend you log into the Rycote site and look for the pamphlet 'The Stereophonic Zoom' which to my mind is the simplest and clearest available introduction to stereo XY recording technique - polar patterns, recording angles etc. It will help you understand exactly what is going on with your fixed point mic and setups such as ORTF, IRT and everything in between Blumlein to spaced omnis.

One (or two in your case) fixed point stereo mics might be an easier (and thus better) tool for you at this stage but a pair of cardioids and a K&M stereo bar are the bread and butter of recording when you really start getting into it. That and the ingenuity of coming up with your own methods of mounting, windproofing, and carrying them about the place! Are you able to borrow either a stereo mic or a pair of mics for the duration of your project (perhaps with the proviso that you return the favour with your own gear afterwards)? That way it could be cheaper to buy an identical mic set for yourself and thus have the 4 channel possibilities without a massive early outlay. I'm still a long way off getting my LsRs 8040 set ....

Jez

[JackHenry]

You'll probably get a better result to your query over on GearSlutz in the Remote Possibilities in Acoustic Music & Location Recording

I'm not saying that the answers you are geting are incorrect, but this is the 'bread and butter' on the  Remote Possibilities forum

[two2the8]

Hey, thanks again @The Immoral Mr Teas! (Sorry to leave you hanging, I was off traveling for the week). I'm reading the stereophonic zoom now & thinking on all of this advice. I also came across some ideas this week that might give me another reason to go F8 or 680 or something like to use the extra inputs. I'm not sure yet, there's lots to think about. In the meantime, that article is really great, and I sure appreciate your help!

@JackHenry thanks for the tip... I'll pop by there & see what I can learn. 

[PhillipWestbrook]

If you already have a 552 you could consider adding a zoom H6n to supplement your channel count.  As far as mics go have you considered the Superlux S502?  It is a 2 channel fixed ORTF mic that is only $199.  I used it on some ambient and music recordings (the music recordings were not planned) in Cuba and am pleased with the results after getting everything back into the studio.  Also what is your goal of the recording? To put the listener in the perspective between the two choirs (one ORTF pair panned left and one panned right) or is it to create a front and back experience? (This could be accomplished using the new waves virtual surround mixing room which recreates an immersion experience using regular headphones http://www.waves.com/plugins/nx)

[two2the8]

Hey @PhillipWestbrook, sorry to miss your comment. That Superlux sounds cool, I'll check it out.

It's a great question you ask, about the goal of the stereo image. It's one I don't have a good answer for yet (though I'm all ears if anyone wants to make a suggestion!). I think my goal is probably more like your first option, though likely with some significant overlap between the two spatially. I'll have to listen to see what sounds OK.

Thanks again to everyone who has pitched in. For anyone following along, here's where I'm at so far:

I think I'll pick up four mics (4x Shure KSM 141s is my best guess so far) & two stereo bars. That will give me lots of options for setups, and having four of the same mic in the field will be great for redundancy & repairs, too. Plus the KSM 141s are switchable to omni, so they're versatile for all sorts of other situations I might encounter. Anyway, I'll run these into a 552 (in the process of purchasing now) and out to a minimum 4-track recorder, probably the F8. I'll try to do some experiments before I go to see how the various setups sound.

After I read the stereophonic zoom, I did up some diagrams to picture the different setups for my application. In theory, for two choirs seated opposite each other, back-to-back spaced pairs appear to align the 'null' areas of each stereo pair to create the best rejection. I'll attach the diagrams. They're keyed to the pickup pattern of the MKH 40, roughly. I know there are problems with thinking about the nulls, but it was fun to make anyway, and maybe it's a little bit instructive? There's a little guide to the diagram below, if anyone's interested.

If it *is* the case that spaced pairs would work decently well for this application, will I have problems when I go to mix? If I narrow the image of each AB pair, am I likely to introduce phasing problems?

---

Guide to diagram:

1: an approximation of the cardioid pickup pattern. 

 

2-6 (ORTF)

 

2&3: What each mic picks up in ORTF. Note the darker purple & brownish-green areas correspond to the SRA from the stereophonic zoom article: sources inside that zone will record in both mics; sources outside will record in just one. 

 

4: What all mics are picking up. Note the areas of bleed. The next slides project out to show how this bleed encompasses both groups. 

 

5: Projecting outward using the same angles to show how the bleed works for the top mics. Same is true for the bottom mics. 

 

6: Projecting outward using the same angles to show how the bleed works for all mics together. It gets too messy here to show what's bleeding where. The upshot is that, with this arrangement, there are basically no null areas. Every mic catches about 2/3 of the off-channel group. 

 

7-11 (XY)

 

7&8: What each mic picks up in XY. Again darker purple & brownish-green areas correspond to the SRA from the stereophonic zoom article. The considerable overlap is part of why XY sounds less interesting than ORTF—XY looks wider, but it sounds narrower because there's nothing in your sound source that will sit on the far side of either channel. 

 

9: What all mics are picking up. Note the areas of bleed as above. The next slides project out to show how this bleed encompasses both groups. 

 

10: Projecting outward using the same angles to show how the bleed works for the top mics. Same is true for the bottom mics. 

 

11: Projecting outward using the same angles to show how the bleed works for all mics together in XY. Again, it's too messy here to show what's bleeding where. Bleed is not really any better here: with this arrangement, there are still basically no null areas. Every mic catches about the same amount of the off-channel group. 

 

12-16 (AB)

 

12&13: What each mic picks up in AB. Again darker purple & brownish-green areas correspond to the SRA from the stereophonic zoom article. Overlap is comparable to ORTF. 

 

14: What all mics are picking up. Note the smaller areas of bleed as compared to those above. The next slides project out to show how this bleed encompasses both groups. 

 

15: Projecting outward using the same angles to show how the bleed works for the top mics. There's almost none. Same is true for the bottom mics. 

 

16: Projecting outward using the same angles to show how bleed works for all mics in AB. It's relatively easy to see what's bleeding where: almost nothing of your desired sound source is bleeding. Null areas align so that you get maximum rejection. Every mic catches only about 1/8 of the off-channel group, if that. 

Microphone_Nulls_V2.pdf