Screwed up a M/S recording. What's the fix?

MichaelPatrick · 8th March 2010

Help! Recently when making an MS recording I aimed the stereo ribbon mic the wrong way. The picture below shows just what I did.

Trying to figure out how to put it right is giving me a headache; I just can't solve the puzzle.

Has anyone done this before or know how to resolve it?

John Willett · 8th March 2010

Quote:

Originally Posted by MichaelPatrick

Help! Recently when making an MS recording I aimed the stereo ribbon mic the wrong way. The picture below shows just what I did.

Trying to figure out how to put it right is giving me a headache; I just can't solve the puzzle.

Has anyone done this before or know how to resolve it?

This is a simple fix - in a normal matrix to stereo you put the side mic in the left channel and polarity reversed in the right.

You just put the polarity reversed to the left and normal to the right.

panatrope · 8th March 2010

Just to clarify - was the mic output recorded with or without the normal M-S matrixing? That is, are you trying to fix the matrixed signal as originally recorded, or did you record the mic signals unmatrixed and are you trying to work out how you should matrix the signals in post?

David Spearritt · 8th March 2010

If you recorded M and S, just reverse the phase of S before decoding to L and R. If you recorded L and R, just reverse the channels L > R and R > L.

Tommy-boy · 8th March 2010

It doesn't matter which way the side mic faces, just flip (phase invert) the side that has the negative (assuming that the front lobe of the center mic is positive). In a normal setup (your “what you should have done” pic) with the + lobes on front and left, you flip the right channel because when a signal comes in from a 45 degree angle on the right, the center mic will produce a positive signal and the side mic will produce a negative signal. If you don't flip the right channel, the signals from the center and side mics will cancel as the negative and positive signals combine. The left channel is left as is because sound coming from the left will create positive signals in both the center and side mics.

So in your “this is what I did” scenario – just leave the right channel as is and flip the left. This will ensure that negatives and positives don’t cancel out. Sounds coming from 45 degrees on the left will make the center mic produce a positive signal while the side mic produces a negative signal – so the left channel of the side mic has to be flipped to avoid signal cancellation. The right side gets left alone because sounds coming from 45 degrees on the right produce positive signals in both the center and side mics.

-Tom

karp47 · 8th March 2010

I dont get it.. could you post a picture of how the mics actually were positioned? and what mics were used?
the picture suggests that you just have to switch the channels, but im sure it has to be more technical than that. im assuming you used a figure-8 for mid and side from looking at the pic?

i can see a few possible problems but nothing that seems too intense...

1. you recorded 2 channels of mid and 1 of side? Solution: delete a mid and duplicate a side, and then phase invert and pan.

2. perhaps your mid was cardioid (though the diagram suggests not) and you pointed it 90' off axis by accident? Solution: record again

this baffles me, im keen to find out a bit more. looking at the diagram it seems like an easy fix, but im sure there has to be something more.

MichaelPatrick · 8th March 2010

Quote:

Originally Posted by panatrope

Just to clarify - was the mic output recorded with or without the normal M-S matrixing? That is, are you trying to fix the matrixed signal as originally recorded, or did you record the mic signals unmatrixed and are you trying to work out how you should matrix the signals in post?

That's an important question! I did not decode the M-S when recording. I need to decode it the DAW and fix my mistake with the array too.

MichaelPatrick · 8th March 2010

Quote:

Originally Posted by karp47

I dont get it.. could you post a picture of how the mics actually were positioned? and what mics were used?
the picture suggests that you just have to switch the channels, but im sure it has to be more technical than that. im assuming you used a figure-8 for mid and side from looking at the pic?

i can see a few possible problems but nothing that seems too intense...

1. you recorded 2 channels of mid and 1 of side? Solution: delete a mid and duplicate a side, and then phase invert and pan.

2. perhaps your mid was cardioid (though the diagram suggests not) and you pointed it 90' off axis by accident? Solution: record again

this baffles me, im keen to find out a bit more. looking at the diagram it seems like an easy fix, but im sure there has to be something more.

It was a stereo AEA R88 ribbon. AEA R88 Stereo Studio Ribbon Microphone – Products – AEA Big Ribbon Mics™ and Mic Preamps

Mic #2 on the system should have been in the middle as the picture shows, but instead I pointed Mic #1 to the middle.

Tom McC · 8th March 2010

Quote:

Originally Posted by Tommy-boy

......So in your “this is what I did” scenario – just leave the right channel as is and flip the left. This will ensure that negatives and positives don’t cancel out. Sounds coming from 45 degrees on the left will make the center mic produce a positive signal while the side mic produces a negative signal – so the left channel of the side mic has to be flipped to avoid signal cancellation. The right side gets left alone because sounds coming from 45 degrees on the right produce positive signals in both the center and side mics...

Tom, don't overthink it - just follow the clear, succinct instructions of Mr David Spearritt.

If you still want to visualize it pictorially: a normal M+S summing generates a dumbell virtual mic pattern with a northwest plus-lobe and a southeast minus-lobe. That is, the L channel - a Fig-8 pointing 45 deg left. But because you screwed up and reversed the iniitial S alignment, your M+S summing has generated a dumbell with the plus-lobe now northeast - that is, the R channel.

karp47 · 8th March 2010

so you captured the wrong signal with the wrong mic?

essentially you wish to change the mid to side?

as far as i'm aware its not possible to do this, but from what you described you should have achieved the desired result, perhaps just with the wrong mics?

John Willett · 8th March 2010

OK - so you recorded (1) as mid (on the left) and (2) as side (on the right) and so doing polarity reversed the side mic.

So everything is correct other than the side mic. is polarity reversed.

You can correct as I suggested in my earlier post.

Or - you can polarity reverse the right channel only in the DAW and then you have a correct MS file for matrixing to stereo.

Or - you matrix to stereo as normal and then left/right reverse the channels.

I hope this helps.

MichaelPatrick · 8th March 2010

Thanks all. I'll spend time this evening reading all the helpful comments and I'll post something to say how it went.

Thanks again!

Matti · 8th March 2010

It is normal for field recording to flip the phase of the side mic when
having the array upside down and thus reversing L-R

Matti

MichaelPatrick · 8th March 2010

Quote:

Originally Posted by MATTI

It is normal for field recording to flip the phase of the side mic when
having the array upside down and thus reversing L-R

Matti,

That's not what I did in this case but your point makes sense. I hope I'll get better at this with more experience doing M-S arrays. It isn't second-nature just yet!

Tom McC · 8th March 2010

Quote:

Originally Posted by MichaelPatrick

Thanks all. I'll spend time this evening reading all the helpful comments.... !

Michael, I'd made comments - and wrongly directed them at Tommy-boy (apologies) - before full details of your array became clear.

If all you want is a quick L and R, then your mic1 + mic2 => R, and mic1 - mic2 => L.

Presumably, though, you want to retan the M and S channels for possible later remanipulation, I would urge you to fix and rename them now for posterity: mic1 = "M channel"; mic2 (after DAW phase-reversal) = "S channel". Therafter, of course, you work with M+S => L; M-S => R.

But surely you can listen to the output imaging and judge whether it's in broad agreement with what you experienced live at the venue?

Roland · 9th March 2010

Just for clarity, whichever mic was facing forward (in this case the second capsule) this is now your centre mic. On capsule one (which is now your side mic) split the signal and flip the polarity of the left hand channel and you should be alright. If you still haven't got it right by reversing your final MS'd channel pans it's all good.

Regards

Roland

MichaelPatrick · 10th March 2010

I looked at the advice you all provided and realized that, as several have said, this problem really isn't difficult. I was making it bigger than it really is.

I do not need to think about the middle mic, it's just #1 instead of #2 and perfectly OK.

The side mic is easily "fixed" with pan and phase changes as you guys explained. Thanks for helping out!

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
How do I fix Low Live Recording Volume?	jonshin	Remote Possibilities in Acoustic Music & Location Recording	5	12th October 2009 04:18 PM
Using a Calrec talkback mic for homestudio recording - how to fix to a stand?	Santiago	Low End Theory	3	7th June 2009 06:47 PM
"ssssssss" sound after recording - how ti fix it ?	sylwia666	So much gear, so little time!	2	7th January 2009 01:13 PM
I screwed up	shproductions03	Remote Possibilities in Acoustic Music & Location Recording	10	13th September 2007 05:29 AM

8th March 2010	#1
MichaelPatrick Lives for gear Joined: Jul 2006 Posts: 2,254 Thread Starter	Screwed up a M/S recording. What's the fix? Help! Recently when making an MS recording I aimed the stereo ribbon mic the wrong way. The picture below shows just what I did. Trying to figure out how to put it right is giving me a headache; I just can't solve the puzzle. Has anyone done this before or know how to resolve it? Attached Thumbnails

8th March 2010	#3
panatrope Gear addict Joined: Apr 2009 Location: Blackburn, OZ Posts: 351	Just to clarify - was the mic output recorded with or without the normal M-S matrixing? That is, are you trying to fix the matrixed signal as originally recorded, or did you record the mic signals unmatrixed and are you trying to work out how you should matrix the signals in post?

8th March 2010	#4
David Spearritt Lives for gear Joined: Nov 2005 Location: Australia Posts: 1,323	If you recorded M and S, just reverse the phase of S before decoding to L and R. If you recorded L and R, just reverse the channels L > R and R > L.

8th March 2010	#5
Tommy-boy Gear addict Joined: Aug 2006 Location: Suburbs of Philly, PA Posts: 432	It doesn't matter which way the side mic faces, just flip (phase invert) the side that has the negative (assuming that the front lobe of the center mic is positive). In a normal setup (your “what you should have done” pic) with the + lobes on front and left, you flip the right channel because when a signal comes in from a 45 degree angle on the right, the center mic will produce a positive signal and the side mic will produce a negative signal. If you don't flip the right channel, the signals from the center and side mics will cancel as the negative and positive signals combine. The left channel is left as is because sound coming from the left will create positive signals in both the center and side mics. So in your “this is what I did” scenario – just leave the right channel as is and flip the left. This will ensure that negatives and positives don’t cancel out. Sounds coming from 45 degrees on the left will make the center mic produce a positive signal while the side mic produces a negative signal – so the left channel of the side mic has to be flipped to avoid signal cancellation. The right side gets left alone because sounds coming from 45 degrees on the right produce positive signals in both the center and side mics. -Tom

8th March 2010	#6
karp47 Lives for gear Joined: Aug 2008 Location: Brisbane, Australia Posts: 726	I dont get it.. could you post a picture of how the mics actually were positioned? and what mics were used? the picture suggests that you just have to switch the channels, but im sure it has to be more technical than that. im assuming you used a figure-8 for mid and side from looking at the pic? i can see a few possible problems but nothing that seems too intense... 1. you recorded 2 channels of mid and 1 of side? Solution: delete a mid and duplicate a side, and then phase invert and pan. 2. perhaps your mid was cardioid (though the diagram suggests not) and you pointed it 90' off axis by accident? Solution: record again this baffles me, im keen to find out a bit more. looking at the diagram it seems like an easy fix, but im sure there has to be something more.

8th March 2010	#11
John Willett Lives for gear Joined: Feb 2008 Location: Oxfordshire, UK Posts: 5,291	OK - so you recorded (1) as mid (on the left) and (2) as side (on the right) and so doing polarity reversed the side mic. So everything is correct other than the side mic. is polarity reversed. You can correct as I suggested in my earlier post. Or - you can polarity reverse the right channel only in the DAW and then you have a correct MS file for matrixing to stereo. Or - you matrix to stereo as normal and then left/right reverse the channels. I hope this helps.

8th March 2010	#13
Matti Lives for gear Joined: Dec 2004 Location: Finland Posts: 3,756	It is normal for field recording to flip the phase of the side mic when having the array upside down and thus reversing L-R Matti

9th March 2010	#16
Roland Lives for gear Joined: Oct 2002 Location: St Leonards on Sea, England Posts: 2,133 2 Reviews written	Just for clarity, whichever mic was facing forward (in this case the second capsule) this is now your centre mic. On capsule one (which is now your side mic) split the signal and flip the polarity of the left hand channel and you should be alright. If you still haven't got it right by reversing your final MS'd channel pans it's all good. Regards Roland

10th March 2010	#17
MichaelPatrick Lives for gear Joined: Jul 2006 Posts: 2,254 Thread Starter	I looked at the advice you all provided and realized that, as several have said, this problem really isn't difficult. I was making it bigger than it really is. I do not need to think about the middle mic, it's just #1 instead of #2 and perfectly OK. The side mic is easily "fixed" with pan and phase changes as you guys explained. Thanks for helping out!