Upcoming organ recording: suggestions?

Discussion in 'Location Recording' started by hughesmr, Feb 6, 2006.

  1. hughesmr

    hughesmr Guest

    Hi folks,

    I have an organ recording gig coming up this week, and it will be the first time I use a 4-channel set-up for this kind of source (typically used 2 spaced omni). The organ stands high in a rear gallery of a VERY live stone edifice with a barrel-vaulted ceiling and about 5 seconds of verb. This organ carries a lot of LF info which has "overwhelmed" the balance in previous recordings I've made there using only 2 spaced Earthworks QTC1s. Also, my QTC1s lose a lot in HF reproduction in the diffuse field.

    So, as my first set-up, I intend to use two Schoeps CMC621 in a "Jerry Bruck" arrangement (14 inches, 120deg) as mains ... not too close, but close enough to get some detail. I can get my mics up to 23', so no problem there. I plan to support the mains with two spaced QTC1, between 5-8' apart, 15' high and maybe 15' behind the main array, faced with the organ off-axis, perhaps facing down the nave away from the organ, and mixed in somewhat lower (perhaps 3-6 dB) than the main array. (Mixing live to 2-track).

    Any comments on this approach? Should I expect mush, or should I just use my QTC1s as mains and move them closer than I did in the past?

    I noticed that John Eargle used some fairly unusual set-ups for his organ recordings ... especially the use of a coincident pair of cards as room pickups far behind a main array of either coincindent cards or spaced omnis. I have a pr of KM184, but am not too crazy about carting them out for this gig. Has anyone here tried such a thing?

    Just want to hear what you think about this approach as a starting point.

  2. BigRay

    BigRay Guest

    The technique that serves me best for that purpose is a Jecklin Disc as mains with Split Cardiod Flankers (meter apart). The Jdisc will help control the LF and give you some beautiful imaging. I have this LF problem in my archiving gig at the cathedral here. The Jdisc combats it nicely..and is a beautiful technique taboot. I have also done baffled fig8s with omni flankers with great results. I do these gigs about twice a week and am finally getting it dialled in.

  3. Simmosonic

    Simmosonic Active Member

    Adding more microphones might end up in more mush, not less. If your previous recordings are overwhelmed with LF energy (amplitude or reverberant, by the way?) AND lacking in HF detail, then perhaps your omnis are simply too far away?

    Or... maybe omnis aren't the right microphone for the job. Who cares for a 360 degree soundfield and extended LF response when it's giving you recordings that are overwhelmed by LF energy and lacking HF detail? Maybe you need to get more direct energy, less LFs and less room?

    I'd be trying to get the sound I wanted with just a single stereo pair of microphones, and if I'd had no success with the QTC1s in the past I'd see no point in continuing to use them on that particular job. There comes a point where it's prudent to move on...

    Grab the Schoeps MK21s and take the following methodical approach:

    1) put them flush together, facing forward and touching, and monitor one of them only;

    2) find a height and distance that gives the most desirable spectral balance and detail on the organ (ignoring reverberation, which will probably be drier than you want at this stage);

    3) from that position, move the mics apart, keeping them facing forward, until you get the desired stereo image (monitoring in stereo again) - they'll probably end up somewhere between 15cm to 30cm apart;

    4) finally, adjust the angle between the two mics, opening them outwards to dial in the appropriate direct/reverberant relationship. Obviously, keep the distance between the capsules as determined in step 3.

    Double check the organ's spectral balance established in step 1 because the pipes may now be audibly off-axis to the capsules, and tweak/compromise as necessary. (For any given mic/sound source relationship, the distance between the mics affects the stereo image, while the angle between them determines the direct/reverberant balance.)

    Also check the mono-compatibility, in case you've stumbled upon a comb-filtering disaster.

    That approach ought to create a near-coincident technique specific to that situation, rather than relying on one of the 'presets' such as ORTF, NOS, Jerry Bruck's suggestion, or whatever - all of which are ideal for some situations but not others.

    I used that approach numerous times when recording pipe organs with DPA 4015 wide cardioids, although they have an on-axis boost at around 12kHz that is actually quite narrow, so they were a bit trickier to fine tune the angle and hence the direct/reverberant ratio. We were using DPA's fabulous (and fabulously expensive) calibrated stereo bar, which helped enormously.

    The process becomes simpler, and changes become more obvious, when you move to more directional microphones such as cardioids...
  4. Thomas W. Bethel

    Thomas W. Bethel Well-Known Member

    Sorry but I have to ask are your Omnis pointed toward the organ? The reason I am asking is because recently I did a recording session in a church and could not help but notice that the resident recording person had hung omni microphones pointing directly down and was using them to record the organ. I talked to him and he was complaining about too much LF information and lack of presence. (sound familiar?) I suggested that he aim the Omnis at the pipe organ and he took my suggestion and it much happier with the results. So I had to ask!
  5. hughesmr

    hughesmr Guest

    Yes, being in the diffuse field and using omnis without a HF lift (the Earthworks spec "ruler flat"), I've always aimed the caps directly at the instrument. I realize that a mic like this instead of something like MK2S or 4006's with attachments isn't necessarily the optimal mic for the job, but they're the only omnis I have (at the moment ... egad, out comes the credit card again!?!)

    A problem in the past at this venue was that I could only get 15' up, and in order to facilitate direct line of sight to the organ facade in the loft, it meant I had to be a bit further out in the room that I would have preferred. Now I can get up to 23', so getting closer and maintaining line of sight is possible now.

    Another thing is that I really wanted to hear how the Schoeps would do in there, but I don't want to sacrifice TOO much low end. That's why I thought I might dovetail the QTC1s lower into the mix as ambient mics to support the increased presence (and perhaps a bit more precise soundstage) using subcards as mains.

    I was always intrigued with the possibility of using a Jecklin or Schneider, but haven't gotten around to it yet.

    Thanks for the suggestions. Keep 'em comin.
  6. Cucco

    Cucco Distinguished Member

    I don't see you sacrificing much low end going with the Schoeps. (omnis that is - even the 21s) They are very robust on the bottom.

    The MK2s is one of my favorite mics ever. You've got to get a little distance or the top end will be VERY over the top, but the bottom is so smooth and accurate (yet beautifully musical) it's hard to imagine much better.

    If you're looking for the absolute lowest frequencies - the M296 picks up EVERYTHING.

    I almost always use omnis for pipe organ and my primary pair for this task is always the MK2s's. Usually a spaced pair well placed is sufficient - no spots needed.

    Just be careful of room resonances. (Do some math up front and it will save you LOADS of time later.)

  7. mdemeyer

    mdemeyer Active Member

    Agree you will not lose much bottom with the MK21's and you can always lift the bottom with a little EQ (oh my, did he say EQ?) if you feel it is a little light. If you can't get close enough for the right direct/reverberant balance with omnis (which it sounds like), they are an excellent choice.

    Also, the QTC1 is much less rolled off-axis on top (the beauty of that tiny capsule), so how you point it would have a lot less effect than other mics like the MK2s. If you are used to the QTC1's, pay attention to this with other omnis, since they are quite different in this regard.


    PS. I want to compliment Simmosonic (is it Greg?) for his excellent post. Providing procedural guidance is so much more valuable than just 'a solution' to the problem. We are all learning and trying to improve. This probably isn't the thread for it (maybe I should start another one), but I'm curious to know his thoughts on how this procedure relates to the published Williams work on stereo arrays. I think it is a good, basic setup procedure, but it probably oversimplifies the interaction between spacing and angle in the final result.
  8. Plush

    Plush Guest

    You're getting some good advice and I'll add something so that you can get a beautiful rolling (trailing) reverb.

    One of the things that people ignore in organ recordings in reverberant spaces is that they forget to mic the reverb!

    Also, many organs are voiced so that the proper sound is at congregation level, not really high up. Check to see if this is the case at your venue.

    Back to micing the reverb---
    use your normal stereo pair and then place a spaced cardioid pair 1/3 to 2/3
    the way back the length of the nave. Face them away from the organ
    towards the back of the church. Take this pair into your console and crank the gain to 45-50dB. Blend this in with your main pair until you hear the reverberation of the church rolling back onto your main pair sound.

    The result is: upon playback, the listener is *there.*
  9. hughesmr

    hughesmr Guest

    A few questions:

    What kind of main pair do you use?

    Spaced cards: what kind of cap spacing/angle? (Recall Eargle's coincident ambient pickups...)

    Have you used subcards for your reverb mics?

    Do you time-align the ambient pr?
  10. Plush

    Plush Guest

    Michael Hughes wrote:

    "A few questions:

    What kind of main pair do you use?

    either omni 4006 or cardioid or wide cardioid--depends on acoustic and repertoire--for example, Bach is perfectly captured by cardioids while Reger (with montrous bass) would need omni or subcardioid.

    Spaced cards: what kind of cap spacing/angle? (Recall Eargle's coincident ambient pickups...)

    spaced cardioids spaced the width of the aisle of the church--angled towards the stone walls of the church--I've never used the strange idea of a coincident pair for reverb.

    Have you used subcards for your reverb mics?

    yes, sub cardioid or cardioid would be fine

    Do you time-align the ambient pr?"

    make an experiment--time alignment is not always necessary or desirable. The ambient pair is sometimes used to smear the sound and envelop the listener in the acoustic.
  11. Simmosonic

    Simmosonic Active Member

    You hit the nail on the head!! It is indeed an oversimplification of Williams work. I had initially started to write a huge tome (as I often tend to do, dammit!) quoting from The Stereophonic Zoom and also Bartlett, but then figured I'd just reduce it to a methodical approach that, if your ears are half-decent and you know what you're going for, ought to do the job - at least to land in the ballpark and know where to go from there.

    I personally think it is important to realise that ORTF, NOS and all of those near-coincident techniques are really just subsets of a huge variety of possibilities, and won't be ideal for everything.
  12. BigRay

    BigRay Guest

    does anyone here use a jecklin disc??ive never heard it mentioned here that I can remember, and yet i find it great as a mains solution..

    ive got a recital saturday, gonna try the reverb thing out..
  13. larsfarm

    larsfarm Active Member

    Could you please elaborate a little bit on that.

    best regards
  14. Cucco

    Cucco Distinguished Member


    Simply finding out what modes exist in the room by determining the resonant frequency will help to assist in placement as well as a potential eq needed later.

    To determine, simply find out what the length, height and width of the room is and find out what frequency has a wavelength of that value. Then avoid placing mics in areas with reinforcement planes. (eg - if the room is 124 feet long, you'd want to avoid putting the mic at 31, 62, or 93 feet intervals within the hall as these are points of STRONG resonance for certain frequencies. Breaking the room in 3rds or 5ths is far less aggregious and often the best scenario.

    The math to determine the wavelength of a frequency is to divide the speed of sound in feet per second by the frequency in Hz. For example a frequency of 1130 in a room at 20 degrees celcius will have a wavelength of 1 foot. A 20 Hz wave will have a length of 56.5 feet.

  15. hughesmr

    hughesmr Guest

    Hey all,

    Thanks for all the suggestions.

    I did the gig last night and decided to take Plush's approach as a starting point. For some reason, I just didn't have the willpower :roll: to record in there without omnis, so I ended up using the QTC1s as mains and MK21s as ambient mics. Ran everything through Benchmark MPS420 and Lucid AD9624 recording at 24/88.2.

    I toyed a LOT with placement and levels (mixed live to 2-trk)...final placement had the QTC1s at 23' (2' higher than the balcony rail) with 0.85m spacing, 0 deg angle, placed at about the 7th pew from the rear of the church (the last pew is basically right under the balcony rail). The MK21s ended up at the 15th pew from the rear (approx 27' behind the mains), 15' up and facing away from the organ, but angled 30 deg up and 45 deg to the outer walls. No time-alignment applied.

    I'm really diggin' the results. My main problem from earlier was QTC1 placement .... just couldn't get close/high enough with my previous rig. But, the addition of the Schoeps really got things "rolling", as Plush would say, without obscuring clarity of line from the main pr.

    I'll try to post a short piece from last night soon, along with some pics of the venue I shot for the CD I'm putting together for the client.

    Thanks also to Jeremy for the standing wave info. The church I'm talking about has some notorious LF nodes: once I settled on mic placement last night, I had the organist play all passages from the program with sustained pedal points just to make sure I wasn't hitting an especially crucial one dead center!

  16. larsfarm

    larsfarm Active Member

    Thanks! Then I can one expect canceling of that frequency (and it's multiples?) halfway between the reinforcement planes too?

    so for all three dimensions avoid...

    0 -- strong at the wall/floor/ceiling
    1/8 -- weak
    2/8 -- strong
    3/8 -- weak
    4/8 -- strong

    ... hmmm, many instruments are high up on the rear wall, centered on the back wall. At least around here. Then a tall stand in the aisle, halfway between the long walls looking back at the instrument would violate this recomendation.

    best regards
  17. Cucco

    Cucco Distinguished Member


    And yes, if you are truly mounting the stand at the half way point, you will definitely be violating this one...

  18. DavidSpearritt

    DavidSpearritt Well-Known Member

    I think its sensible not to represent problems of standing waves in a church to these grossly simplified rectangular prism theoretical formulas. A large church rarely has standing waves, because of the pockets and side passages in the main surfaces, also the modes if they can develop will be so low in frequency that they will not be a problem, ie they will be below any music excitation.

    In most normal sized performance rooms, even in the unrealistic theoretical rectangular prism, modes are so abundant and close in freq that they bunch together above about 100Hz, to become a broadband response and uniform.

    Be careful tossing these formula around, they are really of little practical use in most rooms, a far better way to establish if there are one or two LF modes that may influence mic placement is to do a sine sweep on site with the measurement mic in position. Its also a fun way to find loose light fittings and other rattling hardware.

    Here is a plot of a sine sweep of the governess's bedroom sans furniture in Old Govt House in Brisbane.


    The freq scale at the end of the distinct modes area is about 100Hz and this would be considered a very small room for recording music. This was a pure rectangular prism, with no furniture and windows closed. A larger room would have even lower freq distinct modes if any. No jokes, please, about the governess.

    I should add that if there are modes in a room, which could be avoided by moving the mic stand a bit, it will be disastrous for performing or recording in, far better to pack up and go home.
  19. Cucco

    Cucco Distinguished Member

    I'm afraid you may be FAR more fortunate in other parts of the world than we are in America as to how your churches/places of worship are designed and built. Throughout other parts of the world, churches are a place of architectual beauty. In America, they're brick boxes (lucky if they're brick - often steel structures or "make shift" structures.) In my town (the town is actually quite small with only around 10,000 folks), there are over 300 churches!!! (Some are small shacks or houses converted into churches). Yup, that's one for every 35 people!!! The outlying areas are far more developed than the small town, but they also have numerous churches.

    Most of the churches are nothing spectacular and most of them have a simple rectangular sanctuary. Only a few of the churches are worthy of being called an architectual spectacle and still, most of those are based around a rectangular sanctuary.

    Also - VERY true that the fundamental pitches which are problematic in a room of this size are WAY too low to cause any real problems. However, their direct multiples are potential problems. (Particularly in the low frequency as they tend to dissipate in the higher frequencies).

    The intent of the "Math" portion wasn't to say that by moving the mic an inch or two would destroy the recording. Instead, it was to point out that certain places are, by their nature, bad places to put a microphone. Halves, quarters, etc...

    If one were to get bogged down with too much of this math, it could easily drive them crazy! The main point is, eyeball the placement, if it's roughly half or quarter of the hall, you're going to have problems. However, if this is the only placement option, do the math and you can easily figure out how to correct it.

    Does that help to clarify...?
  20. Simmosonic

    Simmosonic Active Member

    Chiming in (hopefully constructively) on one of my pet topics...

    Regardless of the shape of the room, once it approaches certain dimensions resonance is not an issue for our purposes - regardless of whether it's fundamental or harmonic. This is why books on concert hall design make little or no reference at all about resonance - if they do, it is usually about stage resonance, a different problem...

    To find out if resonance is going to be an issue or not you need to calculate the room's 'critical frequency', which I'll call 'fc' from here on. This is the frequency in the middle of the transition from wave behaviour (room modes or resonance) to ray behaviour (reflections).

    Wave behaviour, which is resonance, occurs when the wavelength is relatively large in comparison to the room's dimension (not necessarily larger, just large).

    Ray behaviour, which is reflections, occurs when the wavelength is considerably small compared to the room's dimensions.

    So... the larger the room, the sooner ray theory kicks and the less problem you have with modal resonances.

    There are a couple of ways to calculate fc. The first relies on measurements of the room's Rt60 and is quite accurate but difficult. An easier way calculates it theoretically, less accurately but good enough for our purposes. How to do it?

    First, you have to calculate the room's Mean Free Path - this is the average distance that a sound ray can travel in the room without encountering a surface. It's useful because it allows us to consider the interaction of the individual dimensions, rather than focusing on one dimension only. The formula is:

    MFP = (4 x Vm)/S

    where Vm is room volume in cubic metres, and S is total surface area of room in square metres.

    When the MFP is known, the critical frequency can be calculated with the formula:

    fc = 1.5 x (V/MFP)

    where V is velocity of sound in metres/second (344m/s at 21 degrees Celsius), and MFP is Mean Free Path (metres).

    Example: A room measures 8m long x 5m wide by 3.5m high. What is its fc?

    To solve fc, we must first solve MFP.

    MFP = (4 x Vm)/S

    Vm is the room’s volume, which is equal to its length x width x height.

    Vm = L x W x H = 8 x 5 x 3.5 = 140m3

    S is the room’s total surface area, which is the addition of the area of each individual surface within the room.

    S = 2 x ((L x W) + (L x H) + (W x H))
    S = 2 x ((8 x 5) + (8 x 3.5) + (5 x 3.5))
    S = 2 x (40 + 28 + 17.50)
    S = 2 x 85.5
    S = 171m2

    With Vm and S solved, we can calculate MFP:

    MFP = (4 x Vm/S)
    MFP = (4 x 140m3/171m2)
    MFP = 560m3/171m2
    MFP = 3.275m

    Now that the MFP is solved, we can calculate fc:

    fc = 1.5 x (V/MFP)
    fc = 1.5 x (344/3.275)
    fc = 1.5 x 105
    fc = 157.5Hz.

    Therefore, the critical frequency for a room measuring 8m x 5m x 3.5m is 157.5Hz. Wave theory would apply to all frequencies below 157.5Hz, and ray theory would apply to all frequencies above 157.5Hz.

    As the room gets larger, fc gets lower and resonance becomes less of a problem. In a large concert hall, fc falls to below 20Hz and so resonance isn't a problem. In a church or similar, however, you're sometimes on the borderline.

    To simplify all of this, I have a spreadsheet that only requires you to enter the room's length, width and height. If the room is not a 'cuboid' (i.e. shoebox or similar, 90 degree corners, parallel surfaces blah blah blah), just do your best at averaging the dimensions into a cuboid. For the purposes we're discussing here, it will put you in the ballpark. It's very small (13.5k). Perhaps Jeremy or someone can post it somewhere?

    It is currently loaded with the data used in the example above...

    If fc is above the lowest note being played in the music, then maybe you ought to start thinking about where the modes exist in the room, where their nodes and anti-nodes are, and so on...

    And, after all of this, remember that no matter what the theory says, if it sounds bad, it IS bad (to paraphrase Duke Ellington). Then you've got to do something about it. If you can't figure out what to do, just move the microphone(s) elsewhere (anywhere!), re-assess, and successively approximate from there.

Share This Page