Mid Side for extra channels!

Storyville · 6th December 2009

I had a session today where I only had 8 inputs on an interface, and a twelve mic setup. I was in a room that was extremely reflective, with all over the place bass. I realized for the drums I could set the OHs tight enough that I wouldn't need spots on the snare or toms. I submixed the kick into both OH channels. Now in the mix, I can use m-s encoding and decoding to mix the kick separately from the OHs. 3 channels for the price of 2. Just wanted to share this because I thought it was pretty crafty.

Ozpeter · 6th December 2009

For MS to work without gross phase errors, the mic capsules must be as closely aligned as possible. With three mics in significantly different places, I suspect you'll end up with a mess. And you've not actually stated that you reversed the phase of one of the OH mics before mixing in the kick, without which you'll end up with I don't know what exactly.

Storyville · 6th December 2009

Your right, I should elaborate.

Mics were XY SDCs, and LDD close kick. OH L and Kick signal are grouped to Group 1, OH R and Kick signal are grouped to Group 2. This way, when Group 1 & 2 are panned out (hard left and right) the kick stays as a mono image in the center, with different OHs on each side. This leaves OH L & R as strictly side information, and the Kick as strictly mid information. Unless I f'd up.... which is totally possible.

boojum · 6th December 2009

It doesn't sound like a good idea to me either but try it! It costs nothing to try it.

Ozpeter · 6th December 2009

Ah, I see (actually, while cooking the dinner I suddenly realised what you meant).

If you have an X/Y recording of a stereo image of the kit from above with the kick added in at the centre, then use something like the MSED VST in its encode/decode modes (from Voxengo) to add or remove more or less from the middle - which equals the kick, mainly - as you reduce the middle you'll be widening the image of the whole kit, or as you increase the middle you'll be narrowing it. And there's not much you can do about the change of image - because if you try to pan it back towards the centre (if it's sounding too wide) then you'll in effect be adding the kick back in, I suspect. Still, as previously said, if it seems to work, go for it!

**DrDeltaM** · 6th December 2009

It should be possible yes to do something with it, if only for the fact that your kickmic signal is totally not in the side-signal. Question will be how workable the mid signal is, but probably you can EQ out some of the overhead signal there if needed to isolate the kick more. Of course, you'll need to compensate for this loss of mid-signal of the overheads. You could double the mid-channel and EQ one to get the kick out, and the other to get the OH's out. Preferably use linear phase EQ then. Or you could re-add the side-channel-information into the mid-signal (this is a shameless plug for my DrMS

).

David Rick · 6th December 2009

I vaguely remember that John Eargle used to do something similar when making orchestral recordings with limited channels. I can't quite dredge up the details, but I think it involved using matrixing to allow the main pair and flanks to be rebalanced later, despite having been mixed to two channels on location.

Can anyone provide more details?

David L. Rick
Seventh String Recording

Ozpeter · 7th December 2009

Once you have a stereo recording, trying to 'unbake the cake' and change the balance of the centre vs the sides (etc) always leads to some kind of downside. In essence it's not unlike the hoary old chestnut of "vocal removal" which used to be the holy grail sought endlessly and tediously on internet discussion forums. One of the more sophisticated tools to attempt this kind of thing is part of Adobe Audition 3.0 (and possibly similar methods have been provided by others) - they call it the "Centre Channel Extractor" and it uses far more sophisticated techniques than hitherto, but with new downsides (depending on the content of the recording) eg phasiness and swishing in the result.

But Audition goes even further with its displays of phase and pan. In these you can highlight any spatial area (not just the centre) and then process the highlighted area, either simply changing its level, or applying a slew of effects in a chain.

When demoing Audition to students, I do a "party trick" which is to take a Norah Jones song - with typically a clear dry vocal in the centre - and with a few keystrokes in the pan display, drop the level of the vocal and add a largish reverb, to make it quite different from the original. Gets cries of ooh and ahh every time.

6th December 2009	#1
Storyville Lives for gear Join Date: Jan 2008 Location: Philly/New York Posts: 4,765 Thread Starter	Mid Side for extra channels! I had a session today where I only had 8 inputs on an interface, and a twelve mic setup. I was in a room that was extremely reflective, with all over the place bass. I realized for the drums I could set the OHs tight enough that I wouldn't need spots on the snare or toms. I submixed the kick into both OH channels. Now in the mix, I can use m-s encoding and decoding to mix the kick separately from the OHs. 3 channels for the price of 2. Just wanted to share this because I thought it was pretty crafty. __________________ I have a new website - check it out: www.Weiss-Sound.com Member of The Pyramid Recording Collective. Grammy Nominations, Platinum & Gold credits, yeah... we got that.

6th December 2009	#4
boojum Lives for gear Join Date: May 2007 Location: Astoria, OR, US&A Posts: 2,132	It doesn't sound like a good idea to me either but try it! It costs nothing to try it. __________________ Nov schmoz kapop.

6th December 2009	#6
DrDeltaM Moderator Join Date: Sep 2003 Location: Belgium Posts: 4,299	It should be possible yes to do something with it, if only for the fact that your kickmic signal is totally not in the side-signal. Question will be how workable the mid signal is, but probably you can EQ out some of the overhead signal there if needed to isolate the kick more. Of course, you'll need to compensate for this loss of mid-signal of the overheads. You could double the mid-channel and EQ one to get the kick out, and the other to get the OH's out. Preferably use linear phase EQ then. Or you could re-add the side-channel-information into the mid-signal (this is a shameless plug for my DrMS ). __________________ Mathijs Indesteege aka Mathew Lane mixing - mastering - audio restoration - plugins http://www.mathewlane.com DrMS. Focus on your stereo field. - NEW v3.2 OUT NOW! DrMS spatial processor - native RTAS/AU/VST plugin » Digital Audio Product Support Joystick Audio - Benelux High End Distributor http://www.joystick.be

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread starter	Forum	Replies	Last Post
DIY mid side technique for lowering mid info on stereo track	Blast9	So much gear, so little time!	12	22nd April 2011 08:31 PM
Mid Side won't null?	Storyville	Mastering forum	11	14th November 2009 10:44 AM
omni mid side	aracu	Remote Possibilities in Acoustic Music & Location Recording	13	19th July 2009 06:11 PM
mid side side effects	1ManBand	Mastering forum	2	2nd November 2008 02:42 AM
Mid - Side - Two or Three distinct channels?	BouncyJones	Mastering forum	6	16th October 2006 01:27 PM

6th December 2009	#2
Ozpeter Lives for gear Join Date: Oct 2008 Posts: 580	For MS to work without gross phase errors, the mic capsules must be as closely aligned as possible. With three mics in significantly different places, I suspect you'll end up with a mess. And you've not actually stated that you reversed the phase of one of the OH mics before mixing in the kick, without which you'll end up with I don't know what exactly.

6th December 2009	#7
David Rick Lives for gear Join Date: Mar 2004 Posts: 555	I vaguely remember that John Eargle used to do something similar when making orchestral recordings with limited channels. I can't quite dredge up the details, but I think it involved using matrixing to allow the main pair and flanks to be rebalanced later, despite having been mixed to two channels on location. Can anyone provide more details? David L. Rick Seventh String Recording

7th December 2009	#8
Ozpeter Lives for gear Join Date: Oct 2008 Posts: 580	Once you have a stereo recording, trying to 'unbake the cake' and change the balance of the centre vs the sides (etc) always leads to some kind of downside. In essence it's not unlike the hoary old chestnut of "vocal removal" which used to be the holy grail sought endlessly and tediously on internet discussion forums. One of the more sophisticated tools to attempt this kind of thing is part of Adobe Audition 3.0 (and possibly similar methods have been provided by others) - they call it the "Centre Channel Extractor" and it uses far more sophisticated techniques than hitherto, but with new downsides (depending on the content of the recording) eg phasiness and swishing in the result. But Audition goes even further with its displays of phase and pan. In these you can highlight any spatial area (not just the centre) and then process the highlighted area, either simply changing its level, or applying a slew of effects in a chain. When demoing Audition to students, I do a "party trick" which is to take a Norah Jones song - with typically a clear dry vocal in the centre - and with a few keystrokes in the pan display, drop the level of the vocal and add a largish reverb, to make it quite different from the original. Gets cries of ooh and ahh every time.