1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

MS Recording balance issue

Discussion in 'Recording' started by johnski, Nov 8, 2011.

  1. johnski

    johnski Active Member

    Hi,

    when recording acoustic guitars (or other instuments!) using MS technique Im getting mixed results in respect of the balancing of the S channels.

    We use good mics, AKG 414 / U87i etc and are dong it through decent analog desk (Soundcraft ghost), but get strange things when balancing for mix.

    As we add the M mic to the S and increae the volume, it usually pulls the stereo image off to the left, so we tend to have to compensate by panning the M Mic slightly right if want a fairly wide mix, which doesnt seem right.

    It doesnt happen all the time, occasionally its fine and all sits central.

    Im always careful to ensure the mics are dead close and square and that when recording the 2nd S channel levels match the first (by centring the pan on both and adjusting volumed until one cancels out the other completely) so Im a little stumped.

    Any ideas what could be causing this?

    Another issue regaring the same matter, instead of recording the 2nd S channel out of phase through the desk, we have tried doing it in Logic by simply duplicating the first S channel and adding a "Gain" plugin which has a phase switch. In theory its exactly the same, but never works. Im presuming due to small delays caused by the software and plugins ?

    Many Thanks
     
  2. bouldersound

    bouldersound Real guitars are for old people. Well-Known Member

    Well, if you place the mics close at the 12th fret most of the guitar's sound is coming from the left and the mic array is simply representing that accurately.

    I wouldn't use a plugin to invert polarity for the reason mentioned, plugin induced delay. If Logic doesn't have a polarity switch on the channel then use a non-real-time process to flip polarity without adding delay.
     
  3. Boswell

    Boswell Moderator Distinguished Member

    The usual MS mixing technique is to start with the M channel at good listening level and then add the S pair to achieve the required width. If you do this and the image still shifts off to one side, then there is something awry with the S mic position or settings. You are sure the pattern was fig-8 on the S mic? Was the axis of the S mic parallel to the body of the guitar? Because of the lop-sided nature of the sound source, when setting up an MS pair in front of an acoustic guitar you need to experiment with different positions that would look odd for a mono mic.

    The MS decoding should work through Logic, but if you are going to use a "gain" plug-in, I would put all the channels through copies of the plug-in and invert the phase on the R channel of the S pair. This will balance the delays and should ensure that the images match up to give good stereo.
     
  4. IIRs

    IIRs Well-Known Member

    MS matrixing in the DAW should actually give better results than with an analog desk as you can match gain settings perfectly.

    I tend to do it using Reaper's routing. I prefer to record M and S to separate tracks: I then put both tracks inside a folder track, which automatically becomes a sub-group. The M channel is now set up. For the S channel I then turn off master/parent send and manually route it to the folder track twice: once to the left channel, and then again to the right channel but with the phase flipped.

    Voila: easy Reaper-native MS matrixing, with the AB stereo version conveniently bused to a track for further stereo processing.

    Another method which may be more convenient if your DAW's routing options suck: record M and S together as a stereo track, then drop in the Voxengo MSED plug to matrix to AB stereo.

    Or record separate mono M and S tracks, then pan them hard left and right, and bus them to a group with MSED inserted.

    I have a perfect solution for the lopsided stereo image: simply use a phase rotation plug (like Voxengo PHA979) on the S channel, and rotate the phase till the image is centred. thumb
     
  5. IIRs

    IIRs Well-Known Member

    Forgot to mention: I only actually set up this routing once. I then saved the whole arrangement as a track template, which I load whenever I record MS stereo.

    And MSED is free in case you didn't realise!
     
  6. johnski

    johnski Active Member

    The position of the guitar has no effect. Ive experimented with that
    Obviously It's not using a stereo mic, but a mono mic fig 8 which is duplicated and one phase inverted and panned off hard Left/Right to give the width so positioning of the instrument left or right of the S mic is irrelevant.
     
  7. IIRs

    IIRs Well-Known Member

    It certainly is relevant. But if you got it wrong you can fix it with the phase rotation trick I described above: it works really well, try it!
     
  8. johnski

    johnski Active Member

    Sorry, I dont understand. If the S mic is a mono mic (fig 8), that is put into a mono desk channel, duplicated etc, then where the instrument is relative to L/R of that shouldnt make any differene to the final mix.
    Surely what is relevant is the position of the M cardioid mic against the fig 8 mic

    If it were to be a stereo mic splt into 2 channels L/R then of course the position would shif if the guitar were moved slightly.

    Can you clarify why you belive the relative position of the instrument to the array makes a difference to the stereo image?

    Thanks
     
  9. IIRs

    IIRs Well-Known Member

    The stereo image is created by level and phase differences between the M and S mic. If you move the guitar to the left you change the relative level and phase between the two mics, so therefore the image shifts.

    Another way to think about it: when I record a choir with an MS pair of M160/M130 ribbons I get a beautiful stereo image, with each singer perfectly pinpointed in the soundfield. So how does the mono S mic know which singers are over to the left and which are over to the right?

    Lets imagine a singer positioned dead centre, exactly in the null of the S mic: this voice is picked up in phase by the mid mic, and is rejected by the S mic, so ends up panned centrally.

    Now imagine a singer over to the left: the M mic is still in phase, but his time the S mic also picks up some in-phase signal as well. After matrixing this signal will re-inforce the L channel, and cancel the R channel, so that singer ends up panned left.

    Finally image a singer over to the right: the M mic is in phase again, but this time the S mic picks up some reversed-phase signal from the rear of the mic, so now the L channel cancels when matrixed, but the R channel is re-inforced, so the singer is panned right.
     
  10. BobRogers

    BobRogers Well-Known Member

    What's happening is that the two mics are basically capturing the same signal in phase. M+S on the left is adding and M-S on the right is canceling. I've had this happen on quiet fingerpicked guitars when I moved the mics (Beyer 160/130) very close (3-4") to the guitar. IIRs trick will center the image. But the pair of signals I had when this happened to me were so close that I don't believe there was much stereo information there. I ended up just using the M track as there were a lot of other elements in the mix.
     
  11. johnski

    johnski Active Member

    Ok thanks - that makes sense now. I presumed that a mono channel (being fed by a fig 8 mic) could only either be in or out of phase (as its a single channel), but youre sayiong of course it can actually capture both, which is then decoded once split.

    We've experimented with muliple guitar positions and moved aboit whilst recording as initially thought that to be the problem, and so will try some of the other techniques you suggest and I'll report back!
    Thanks for your time :)
     
  12. TheJackAttack

    TheJackAttack Distinguished Member

    Which mic are you using as the fig of 8? A C414 is designed by nature to have a slightly darker sound on the back side. The U87 is not designed that way but there are still some slight differences from front to back side. In the case of the C414 this could on the surface appear to created S balance issues.

    Also, it is not out of the question that the S mic is not functioning correctly or at least that the polar selector switch is sketchy.
     
  13. johnski

    johnski Active Member

    Hi we have various fig 8 mics including a few 414's, u87, SE Electronics and have used different ones at different times but havent related it to this balance issue, so something for us to try I guess. The Fig 8 patterns are definitely selected (weve done this countless times) and used different 414's.
    What Fig8 mics do you recommend for MS on acoustic guitars and similar ?
     
  14. Boswell

    Boswell Moderator Distinguished Member

    MS guitar miking is one of the applications where the Beyer M130/M160 combination can work well, but not necessarily for every instrument and player. My most successful recordings using this technique have been done using a pair of Rode NT2-A condensers. They are a little large and slightly clumsy to position, but give excellent results.

    Depending on the acoustic environment, I sometimes also use the technique I call MS-Blumlein. This uses the same physical positioning and decoding as standard MS, but both mics are in Fig-8. The M pattern has, of course, a rear lobe of the same sensitivity as the front lobe, so room reverberation plays a larger part in the recordings than with standard MS, but it can avoid the uncomfortable closeness effect you sometimes get on a guitar with a cardioid M mic.
     
  15. IIRs

    IIRs Well-Known Member

    Yes, its worth noting that the Mid mic can be any polar pattern you want. I've had good results recently using my omni Earthworks QTC30 for the mid, with an SE Z5600A or U87AI as the side: looked a little odd, but actually sounded very nice.
     
  16. johnski

    johnski Active Member

    The MS test results are in !

    And the winner is.... JackAttack !

    OK, we've spent a while checking this out by recording a sound source using MS with the following Mics for Fig 8 (S) .... AKG 414B Xii (2 of), U87Ai, Rode NT2A, Se Electronics Z5600ZA.

    All was fine on every occasion so long as we didnt use the 414's for S... either of them, they has the same effect of when the M fader is increased.. the stero image is pulled off centre to the side.
    Rotate the mic 180 degrees and it gets pulled off the other way !!

    Also, whether put through our desk and phase flipped on the desk or recorded as 2 tracks and the S duplicated and phase flipped in Logic, the results were the same.

    The only other problem we encountered was using the direction mixer plugin in logic, it just sounds awful, but the Logic "gain" plugin with phase flip switched on or the Voxengo MSED (Thanks for the suggestion IIRs!) work just fine.

    So - who'd have thought it ! Result - seems to be dont use 414's for MS recording ! Nice one JackAttack! Shame though as out of all the combinations we tried, the 414 sounded the most detailed for the S mic !

    Thanks All :)
     
  17. TheJackAttack

    TheJackAttack Distinguished Member

    I have certainly used a pair of C414's for MS recording. I also have used the phase rotation trick on pairs other than C414's. Really this one reason MS is supposed to use mics as identical as possible. Also, you might consider having the mic checked out unless it is EVERY 414 in your collection behaving in identical fashion. Nowadays I use a Royer SF12 or a pair of R101's for MS. A pair of U87's is out of my budget and if I were to drop dime on Neumann, it might rather be for a trio of M50's for an orchestral decca tree.
     
  18. RemyRAD

    RemyRAD Well-Known Member

    MS can provide as much directional information or more than XY, ORTF, spaced omnis. But a lot of the MS magic happens because of the acoustics of the room. No acoustics will present much more difficult even stereo imaging. Especially when working close to an acoustic guitar whose large tubby body is off to one side and it's skinny non-resonant neck off to the other side. How is that supposed to sound centralized? It can't. It shouldn't. Although part of your description makes me believe that something is awry in your routing of the S channel? Software MS decoding is far more accurate than you believe. I get identical results whether I'm using software or the old Neve. And frequently, I'll compare both. The only real difference I experience is either hearing that Neve or not hearing the Neve. But really what makes you believe that an asymmetrical instrument should sound symmetrical? In order to centralize your stereo image with MS you have to move the performer further to their left if they're right handed. This is so that you extend the body of the instrument into the central stereo perspective. And that's how you have to deal with MS. I have to deal with MS a little differently because not only do I use it, I have it. Thankfully it's in remission. Thankfully my technique isn't.

    MS technique. MS disease. MS I am... Remy Ann David
     
  19. audiokid

    audiokid Staff

    For John,

    Re M50's
    If only I was Santa
     
  20. johnski

    johnski Active Member

     

Share This Page