M/S processing vs. M/S recording

[LaceSensor]

Hello,

I am reading a book now which describes a M/S mic technique which uses a cardiod dynamic mic (on-axis) and a figure 8 mic (sideways 90°/270°). I am also familiar with plugins which have M/S processing. Therefore I am slightly confused.

1) does a M/S mic setup require additional M/S processing?
2) should M/S plugin processing be activated when using a M/S track?

I've seen one article which described reversing the polarity on the (S-) as compared to the (S+) and (M). However with the above mic technique there would only be one track for the (S) as compared to having individual/separate tracks for (S-) and (S+).

I am just trying to learn how I could start to use M/S in my recording and mixing. I think my confusion is stemming from the fact that some of these plugins with M/S processing are actually looking for a (L/R) stereo signal as the input. On top of this I believe there may be some required processing for a track which is recorded using a M/S mic technique.

 

[Farview]

The mic technique needs to be decoded. The manual way is to put:
Track 1 is the mid mic,
track 2 is the side mic panned to one side,
track 3 is the side mic with the polarity reversed panned to the other side.
The 'width' of the stereo field is adjusted by the volume of the mid mic. More mid=less stereo width, less mid=more stereo width

There are MS decoder plugins, which require a stereo track consisting of the mid mic on the left and the side mic on the right. It is doing the same thing as above, but inside the plugin.

There is also MS processing, which is an entirely separate thing from the micing technique. MS processing allows you to process the mono signal (L+R) separately from the stereo signal (L-R). For mastering purposes, it might be necessary to EQ or compress the center of a mix separately from the stuff on the sides. That is the sort of thing you would use that for.

 

[LaceSensor]

The mic technique needs to be decoded. The manual way is to put:
Track 1 is the mid mic,
track 2 is the side mic panned to one side,
track 3 is the side mic with the polarity reversed panned to the other side.
The 'width' of the stereo field is adjusted by the volume of the mid mic. More mid=less stereo width, less mid=more stereo width

There are MS decoder plugins, which require a stereo track consisting of the mid mic on the left and the side mic on the right. It is doing the same thing as above, but inside the plugin.

There is also MS processing, which is an entirely separate thing from the micing technique. MS processing allows you to process the mono signal (L+R) separately from the stereo signal (L-R). For mastering purposes, it might be necessary to EQ or compress the center of a mix separately from the stuff on the sides. That is the sort of thing you would use that for.

Wow, what a great reply. Thank you for this information.

 

[bouldersoundguy]

I can vouch for what Farview said.

Mid-side processing starts with a stereo file and reverses the decoding process, applies effects independently to the mid and side channels then re-decodes it back to stereo.

 

[rob aylestone]

I too can vouch for Farview’s input.

M/S is a really good technique, but what puts people off is that it’s awkward to monitor without the decoding. What I mean is that if you go to a location to record something, we all like to plug in headphones and have a listen. You can’t do that with M/S. The recorders designed for M/S have the little matrix circuit inside, so you can monitor. The old recorders I used didn’t do it, so M/S involved courage, so for some recordings, I used X/Y simply because I could satisfy my monitoring check. If you can monitor that is the key to giving it a whirl.

The real benefit of it though, is that you can adjust width back in the studio. You can even pick the mid mic to suit the location. It could be almost any mic. If your mic position is a bit too far away, swap the cardioid for a super, or a hyper or even a shotgun. You pick the mic as if it was a mono recording, and then the side mic adds the space left and right. Many stereo techniques go wrong, especially with choirs and amateur orchestras. Their problem, unlike the pros, is balance. Maybe they haven’t got enough violins, but too many cellos or the choir has far too few fellas. You do your recording and the problems are fixed in it. With M/S your centre mic becomes the key. Everything visually makes sense. The centre might not be the real centre, so you might look at the lack of some components and make decisions. Move the mic nearer to the few fellas, to capture more of them, or shift towards the few string players and get away from the four trombones you know will over power everything. With X/Y this shifting of centre is less successful because you are using a prescribed angle between the two mics and opening it creates a hole right in the middle or a vastly more mono result.

M/S can be quite exciting if back in your studio, you bring up the mid fader and hear a nice well balanced result, and then you either bring up the two side faders or raise the side plug-in and it sounds amazing.

as an aside, we often talk about stereo microphone matching, but with M/S the side mic is never matched with the mid mic, and naturally sound different.

once you have a figure 8 mic, you can even use M/S on acoustic guitars and it sounds nicer and less ‘created’ than the usual neck/hole systems. I suppose the fact we never had the fig 8 version of a solid cheap mic like an SM57 held it almost as a secret.

 

[Lancep]

What everyone else said. I’m just finishing a song where the acoustic guitar is serving as the anchor chord element and it’s recorded in m/s with a 57 and a fathead( the only mics I currently have.) One thing I would add for m/s recording is mono compatibility. Because the right and left are created by flipping the polarity of a single mic, if you collapse it to mono the sides cancel each other out preventing phase issues.

 

[leddy]

Also, you don’t have to use a cardioid for the mid mic. You can use omni or another figure 8. A cardioid mid decodes to a stereo pair of hyper cardioids. Omni decodes to two cardioids at 180° from each other (but the directionality can change with the amount of side mic used), and a fig 8 mid is essentially blumlein but you can adjust the mic angle with side mic gain changes.

Each one has its use. Just don’t go overboard with widening things too much.

 

[rob aylestone]

An M/S with fig-8 mics is a poor one to try because it’s a skewed Blumlein. At a frequency where the two mics have identical gain, then it would be a Blumlein rotated 45 degrees. The snag is mic frequency responses are not linear, so you get rotation that changes as the frequency changes. Not good.

I’m not sure about the notion that cardioids ‘decode’ to hypers. Not in my experience. They remain cardioid. They don’t get narrower unless you drop the amount of centre channel, which would be unusual? I get that if you start with just the fig-8and gradually swap to the centre, the two halves have a bit of rear from their front and back, sort of cardioid, polar pattern, but that morphs to cardioid, or omni, depending on the centre mic, but I can’t see where hyper comes from?

 

[leddy]

At equal m/s gain, a cardioid mid still has a rear lobe. You’re right though that they become cardioid depending on gain.

Re Blumlein MS, it’s definitely not rotated. Traditional MS your capsules are 45° relative to the source. In Blumlein MS, you have the mid directly on axis just like any other MS pair. The two Blumlein methods are the same except for any capsule irregularities based on angle to the source, which I think you eleuded to. Better mics should be less problematic. I’ve got a pair of MKH30’s so I’m good.

 

[rob aylestone]

We'll just have to disagree - not to worry. They become different patterns based on the ratio of mid to side which changes with frequency. Your diagram shows the 45 degree rotation compared with the normal orientation of two fig-8 in Blumlein. The two mics have a resultant 'direction' at 0 degrees (with one 45 left and one 45 right) M/S have forward and backwards at 0/180, and the side 90/270, so the 'middle' is rotated.

 

[leddy]

We'll just have to disagree - not to worry. They become different patterns based on the ratio of mid to side which changes with frequency. Your diagram shows the 45 degree rotation compared with the normal orientation of two fig-8 in Blumlein. The two mics have a resultant 'direction' at 0 degrees (with one 45 left and one 45 right) M/S have forward and backwards at 0/180, and the side 90/270, so the 'middle' is rotated.

I’m genuinely curious here. When you’re talking about pattern change with ratio of mid side which changes with frequency, I need some more detail.

 

[leddy]

The resultant directions are all derived from sum and difference same as any other mid pattern. If these problems exist with a fig 8 mid, why don’t they exist with a cardioid?

 

[rob aylestone]

No, we're just talking about rotation when looking at the orientation of the array - Blumlein is oriented to different commas positions than an M/S array doing the same thing, that's all. It's not a problem, just an interesting fact. Blumlein, with it's main focus of direction from the two mics at +45 and -45 has a discernible front and rear, despite these only being formed by the sum of the left and right being equal in the forward direction - being formed by the polar pattern only just starting to decay. In the M/S, the forward direction is the forward mode of the fig-8. The sideways facing fig-8 provides the usual M/S side component. The result is the same, but the M/S and Blumlein arrays are in a different position for this same effect, 45 degrees rotated. That's kind of interesting. Two identical microphones that provide left to right stereo field by pointing in different directions.

There is a difference however in the rear. Blumlein has the rear field swapped around - front right is also rear left. In the M/S version, the have frontal and rear pickup, but the left and right components are front and rear left and front and rear right. In a typical church setting. The annoying person with a cough behind the microphone would be on the other side in Blumlein, compared to M/S. I don't know for certain, but I think this is what adds the special 'bigger' sound that Blumlein has compared to M/S? I've not read any research on the rear swapping impact, but it probably does add to the overall sound. Maybe this is why the use of fig-8 or omni for centre channel is not that popular? One of those diagrams also looks a bit out of proportion - the fig-8 side with cardioid. It's drawn so that the cardioid ONLY is forward of the side mic centre, when the polar pattern of the cardioid for the same pressure points would be drawn further back, so the cardioid centre is towards the back. I think the person who drew it distorted it to make a point.

 

[leddy]

I’ve always thought the rear lobes were also reversed in MS Blumlein. I’m going to have to look at that again, maybe experiment and see for myself.

I appreciate the discussion. Always trying to learn.