Mid Side technique question

[geeek]

Hello..

I've been recording acoustic guitar all day using the m/s configuration.

The setup I've been using is cardiod Beyer MC930 into DAV BG for the Mid, and Neumann m269 figure 8 pattern into neve 1084 - then into pro tools where I have split the side into stereo hard panned and phase swapped the right channel.


I believe I have done everything right, and to me, it sounds sweet..

My question is..

The left channel in the mix is the louder one when I introduce the side element..

Is this normal?



Thanks in advance

[Baithak]

From your description it seems you have not decoded the MS signal correctly. If ALL you have done is phase changed the R channel then, then you just have a pair of tracks which are M and -S. So, the loud signal you are hearing in the LEFT channel is the original M signal.

You have to duplicate the R track and change the phase of one of the duplicated tracks. Now pan the original M dead center (equal in both output channels), pan the original R track (which corresponds to the S signal) hard left and the duplicated phase changed R channel hard right.

So your master bus will now have the Left output as M+S and Right output at M-S which a complete decode of the MS signal.

Good luck,
Baithak

[geeek]

Well.. the M signal (cardiod) facing source has remained untouched..

The S signal (figure 8) at 90 deg (side of mic facing source) has been duplicated, panned hard left and right.. then the phase flipped on the right hand channel..

Is this not correct?

Cheers

[Shadow_7]

Sounds a little bit like Blumlein. Minus the dual figure 8's. It's my understanding that the S in M-S is not figure 8. But I've never done any M-S stuff myself. So I could be wrong. The left is probably louder because of the phase.

[Corran]

The side mic in an M-S setup MUST be figure 8. The mid mic can be whatever.

Was the left of your acoustic guitar simply louder? I've thought this before but it was simply a product of the acoustics. If the left side of your array was pointing towards the sound hole and the right side to the bridge, it would stand to reason that there would be more sound to the left, correct?

[Vocalvoodoo]

Originally Posted by binary

Hello..

I've been recording acoustic guitar all day using the m/s configuration.

The setup I've been using is cardiod Beyer MC930 into DAV BG for the Mid, and Neumann m269 figure 8 pattern into neve 1084 - then into pro tools where I have split the side into stereo hard panned and phase swapped the right channel.


I believe I have done everything right, and to me, it sounds sweet..

My question is..

The left channel in the mix is the louder one when I introduce the side element..

Is this normal?



Thanks in advance

I'm glad you have posed this question because I'm getting the same thing. I've only just begun experimenting with m/s recording, so I also have no idea if this is normal.

[stuntbutt]

This may be a product of ITB latency.

[Banjo]

Hi, unsure what program you are using, but somehow using busses etc. add M to S for one channel and M to inverted S for the other. Using sends to an aux bus seems the easiest in 'tools. By adjusting the volume of either will adjust the stereo image.. Denis

[sonicdefault]

Originally Posted by Corran

The side mic in an M-S setup MUST be figure 8. The mid mic can be whatever.

Was the left of your acoustic guitar simply louder? I've thought this before but it was simply a product of the acoustics. If the left side of your array was pointing towards the sound hole and the right side to the bridge, it would stand to reason that there would be more sound to the left, correct?

This is most likely, exactly the reason for the variance. M-S for Acoustic is difficult for this very reason. Simply because most of us want the Mid mic to point at the sweet spot on the gtr, and this is usually not the sound hole. As a result, the Side image is shifted due to the offset.


-SD

[audiothings]

This is most likely, exactly the reason for the variance. M-S for Acoustic is difficult for this very reason. Simply because most of us want the Mid mic to point at the sweet spot on the gtr, and this is usually not the sound hole. As a result, the Side image is shifted due to the offset.

what he said!

plus, it could be because of mismatch between the two microphones, or because one side of the fig 8 mic responds differently from the other (this tends to happen with a lot of fig 8 mics over time).

.02,

[geeek]

Originally Posted by sonicdefault

This is most likely, exactly the reason for the variance. M-S for Acoustic is difficult for this very reason. Simply because most of us want the Mid mic to point at the sweet spot on the gtr, and this is usually not the sound hole. As a result, the Side image is shifted due to the offset.


-SD

I think this could be a reason for it.. but I don't see how, since the mic is figure 8 and isn't truly a stereo source, would it cause this issue?

"plus, it could be because of mismatch between the two microphones, or because one side of the fig 8 mic responds differently from the other (this tends to happen with a lot of fig 8 mics over time)."

This might also be a factor.. as the figure 8 mic is a Neumann M269 which was made in the mid sixties..

This is really really frustrating..

I've been tracking for an album, and had the artist play the same 2 tracks in twice now, then gone back and decided I've not been happy.

I've got the sound I want from the two mic's positioning, but now this massive veer to the left.

I found a link at Mid-Side Microphone Technique - WikiRecording

Near the bottom of the "theory" bit.. it mentions something about "But around the right side of the microphone the Side mic is 'incorrect' compared to nature, and therefore cancels the Mid, so you get much less signal on the right side.
In fact, if you draw out all the numbers, you end up with a hypercardioid microphone aimed 63.45º to the left of center.
At this point, add the two mics together and you get a hypercardioid signal aimed better than 45º off to the left."

It's difficult to understand what they are trying to say, but it seems to be telling me to do something further than I already have.


Does anyone else have experience of M/S at all using pro tools? I'm not sure where the issue lies here..

Is there any way I can rectify it at all!?

Thank you, sorry for the frantic nature of this post.. but it's driving me insane.

[geeek]

does anyone at least have an idea how to remedy this in post production?

or do I need to go back and re-track?

even then, how do i solve this happening?

thanks

[d_fu]

If all else fails, just use the mono track from the Beyer and work with that... Else, reduce the amount of S (reducing stereo width) and pan by adjusting the volumes of the two S tracks until you get what you want.

[Westmalle]

Use the S1 Shuffler or Imager of Waves.

[phil.]

Here is how I would decode MS in ProTools.

2 Mono Audio tracks for Mid and Side channel recording
2 Mono Aux for splitting the Side into S+ and S- (grouped together and Phase reversed on the S- channel with the EQ Plug In)
1 Stereo Aux as a MS Master Channel

Hope this helps.

Best,

Phil.

[mohthom]

To ensure that your matrixing is correct;

Turn off the centre channel. Remove the left/right panning (pan both dead centre). Do not adjust the phase flip. The sounds should cancel.

It's that simple!

MohThoM

[geeek]

Originally Posted by phil.

Here is how I would decode MS in ProTools.

2 Mono Audio tracks for Mid and Side channel recording
2 Mono Aux for splitting the Side into S+ and S- (grouped together and Phase reversed on the S- channel with the EQ Plug In)
1 Stereo Aux as a MS Master Channel

Hope this helps.

Best,

Phil.

Cheers.. I've done it exactly like that yehh.. I can't understand this shift to the left.. do you guys think it's probably die to more sound coming from the L of the S mic?

I really wanna keep the sound of the combined mics, and have wide imaging, but this left shift is really horrible!

[d_fu]

Originally Posted by binary

I really wanna keep the sound of the combined mics, and have wide imaging, but this left shift is really horrible!

Maybe you could upload short sample files of M and S tracks...

[geeek]

yup, i'll do that..

bear with me..

[geeek]

Ok.. Sample 1 is of the guitar with neumann m269 as the fig 8 (S) mic..
The screenshot was taken just before I printed that sample.

Sample 2 uses a smaller diaphragm ribbon as the fig 8 S mic.

I've included both because there were some people who thought it may be due to the mic picking up more sound from the L side of the fig 8 mic.. I don't know if this is possible.. so here's both anyway.

Thanks for your help with this, really appreciated.

[Corran]

Though figure 8 mics are not "stereo", they do pick up sounds from both sides. The left side pressure sends a + voltage and the right side sends a - voltage, which is why if you flip the phase it cancels.

The quickest way to fix this is to bounce to a straight L-R signal (as you have in the samples) and pan that to taste, or simply raise the volume of the right channel. I did that on the mixer here when listening to these demos and it sounded great!

By the way, the wikipedia is just giving the the maths so don't worry, it sounds like you are doing it right.

[d_fu]

What's the EQ doing in the right S channel, swapping phase?
When you mute M and pan both S channels to center (with phase flipped on S-R), do they cancel out?

[geeek]

heyy thanks...

so when you listened to the samples i've posted, and raised the right hand signal on your playback system, do you think the imaging was accurate?
I suppose I could just make a stereo channel for the summed M and S signal and pan that accordingly..

I'm just annoyed coz it should just work! lol

You like the sound generally tho?

Maybe we shouldn't get into the "should I use the ribbon or valve mic" topic..

Lol

It is annoying me tho!

[geeek]

Originally Posted by d_fu

What's the EQ doing in the right S channel, swapping phase?
When you mute M and pan both S channels to center (with phase flipped on S-R), do they cancel out?

Yeah the e.q on the R channel is just flipping phase..
I've also tried doing this in Audiosuide, so no latency is involved.. Same issue.

Yehh when I do that, they cancel out completely.


I find it all really strange, because the levels of L and R of the S mic are identical, but when summed with the M channel, the level shifts to the left.

Grrrr

[Corran]

It all sounds good with the level adjustment.

By the way, what ribbon are you using? Some newer ribbons are not centered in the magnet and have slightly different pickup patterns on each side. Royers do this.

[d_fu]

Originally Posted by binary

I find it all really strange, because the levels of L and R of the S mic are identical, but when summed with the M channel, the level shifts to the left.

Which proves that with MS you can actually achieve L/R localization...
It's obviously the signal itself which is tilted. An XY stereo would have given you the same result. Why don't you just pan the resulting stereo signal right by a dB or two (doesn't seem to be much more)? Or try to attenuate the L-S channel a bit...

[geeek]

Thanks Corran..

I've just tried shifting the panning of the summed M and S, and it does sound good yeh.. until I put it into mono.. coz the signal shifts over!

The ribbon is a Crowley and Tripp Naked Eye..

It's doing it on the ribbon sample and the valve sample tho..

The valve mic is a Neumann M269 which is basically a U67 with a different valve.

The same thing happens with both..

I could monitor in M/S I suppose and move the S mic round some... From what I understand, the Fig 8 S mic needs to be precisely 90deg to the Cardiod..

Obviously the cardiod is pointing at the sweet spot.. so the fig 8 mic will most likely be pickup up more + energy from the L (front) of the S mic, than the rear of the 8.

I spoke to a guy on the phone about it now, and he reckons that the fig 8 patterns aren't too tight on some mics.. but I really dunno.

Surely this is the only explanation for it.. that there is more energy from the front of the fig 8 due to it facing the body of the guitar.


Problem is, this is for an album, where I want a degree of uniformity in the guitar sound on the album, with some maybe wider than other.. hence using M/S to be able to adjust this in post..

I guess it's either scrap it and start fresh (only tracked 2 guitar parts so far) or keep going and risk it shifting a slight degree when played in mono.

Hmmmmmph :(

[d_fu]

Originally Posted by binary

I've just tried shifting the panning of the summed M and S, and it does sound good yeh.. until I put it into mono.. coz the signal shifts over!

If you adjust the volume/balance of the S channels, this shouldn't be happening... M will be M.

[geeek]

Originally Posted by d_fu

If you adjust the volume/balance of the S channels, this shouldn't be happening... M will be M.

You're right.. But.. If I adjust the balance solely of the S channel(s) and not the summed M and S, the imaging literally vanishes.. it just sounds mono..

I guess I get the flavour of the other mic I want by doing that tho, just not the width.

I do think it's to do with energy being stronger in the + (front of S mic) which is L when panned..

Maybe E.Q on one of the S channels would fix this.. I did try playing with it, but I don't want to do too much to the guitar, and make it sound unrealistically big/wide.



Hmmmmmmph :-/

[stuntbutt]

Originally Posted by binary

The ribbon is a Crowley and Tripp Naked Eye..

:(

From the Crowley and Tripp site:

What is True Dual Voicing?
True Dual Voicing means that each side of the Naked Eye has a very different voice, or tone. Crowley and Tripp have specifically tuned Naked Eye to have two distinct personalities, each with excellent characteristics for your recording needs: The front, emblem side has that classic ribbon sound - that thick and creamy tone guitars love. The back has a significantly brighter, rising curve that mimics the rise of high quality vocal condensers,

This could explain the image leaning to one side.