A couple of questions about mid-side recording

[miroslav]

One side faces left and one side faces right.....when you do the Mid + (+ Side) and Mid + (-Side)....you end up with a left and right stereo image.

It's about the TWO MICS working together in combination.....you keep looking at the side mic as only a single mono mic.

 

[bouldersoundguy]

That's part of what I'm talking about. You've got two mics stacked vertically, which all the m/s setups I've seen are. One is aimed center, one is taking info from the sides but in a mono fashion. Where is the true Left / Rightness coming from?

If a sound originates from the left side of the mic then it's louder in the decoded left channel. If it comes from the front it's equal in both channels. If it comes from the right side of the mic it's louder in the right channel.

The mid mic captures the overall sound. The side mic captures the angle from center in the strength of its signal, and it distinguishes left and right with the polarity of the signal relative to the mid mic.

Copying/inverting/panning the side signal converts from one format (sum-difference) to a more convenient format (left-right). The information is the same.

 

[Farview]

Anything on the left will hit the side Mic and the mid Mic in a positive direction. Anything from the right will hit the side Mic in a negative direction and the mid Mic in a positive direction. So, when you add those two mics together, anything coming from the right will cancels out and anything coming from the left will stay. That is the left channel signal.

Now, copy the mid Mic signal, inverted the polarity and add it to the mid Mic. Now anything coming from the right will be positive and will add to the mid Mic and anything coming from the left will be negative and get cancelled by the mid Mic. That is the right channel signal.

This is the same way fm radio and old school television is broadcast, as a mid signal and a side signal. There are technical reasons that I can't remember as to why it is a better way to broadcast, but I can't recall the specifics right now.

 

[brassplyer]

There are technical reasons that I can't remember as to why it is a better way to broadcast, but I can't recall the specifics right now.

As I understand it when summed to mono the side mic cancels out bec of the inverted l/r wave forms leaving the mid signal making it totally mono compatible.

 

[Farview]

As I understand it when summed to mono the side mic cancels out bec of the inverted l/r wave forms leaving the mid signal making it totally mono compatible.

But that would happen even if it was broadcast in left and right. When a left right signal is encoded into a mid side signal, it is still only as mono compatible as the right left signals it was made from. The mid side mic technique is mono compatible for that reason, but encoding a left right mix that is not mono compatible to mid side doesn't magically make it mono compatible. You end up with the same thing you started with in both cases.

 

[Steenamaroo]

That's part of what I'm talking about. You've got two mics stacked vertically, which all the m/s setups I've seen are. One is aimed center, one is taking info from the sides but in a mono fashion. Where is the true Left / Rightness coming from?

Yes, the side mic is mono; The key point is that it picks up sound from both sides equally.

Without any decoding/duplicating/inverting, a mid side setup just plays back audio coming from the front left of the setup and cancels audio from the front right.(or the opposite)
When you dup/flip the opposite becomes true.

If you keep the three tracks centred, the side mic will cancel out and leave the centre mic remaining.
If you pan the side tracks out you reveal your stereo image.

All the information is there. It's just a question of how it's processed.

 

[ecc83]

"This is the same way fm radio and old school television is broadcast, as a mid signal and a side signal. There are technical reasons that I can't remember as to why it is a better way to broadcast, but I can't recall the specifics right now."

Back in The Day, all radio signals had to go down telephone lines rented from the phone company (BT here). Those lines needed to be matched for phase and response equality otherwise they screwed the stereo image (imagine using a cap mic for left and a dymo for right!) . The broadcasters did their best with complex equalizers but the best way is to split the signal into L+R and L-R . The matching of the lines is now much less critical since anomalies only result in increased crosstalk and as has been said, if you have the signal in its algebraic parts you can, to some extent, correct for that.
This is why, for slightly similar reasons, UK eschewed "Never Twice Same Colour" for CTV !
Dave.

 

[brassplyer]

As I understand it when summed to mono the side mic cancels out bec of the inverted l/r wave forms leaving the mid signal making it totally mono compatible.

But that would happen even if it was broadcast in left and right. When a left right signal is encoded into a mid side signal, it is still only as mono compatible as the right left signals it was made from.

Isn't it correct that when played in mono the side mic portion cancels altogether leaving *only* the mid mic portion? If you have a L/R only signal that isn't specifically formulated for mono compatibility, doesn't that leave the possibility of unintended partial cancellations? That's what I came away with from various videos on the m/s method.

 

[bouldersoundguy]

This is the same way fm radio and old school television is broadcast, as a mid signal and a side signal. There are technical reasons that I can't remember as to why it is a better way to broadcast, but I can't recall the specifics right now.

FM was originally mono. Stereo was added later by placing a difference signal (L - R) on a subcarrier. That way the existing and any future mono receivers would function normally but stereo receivers could decode the difference signal and, just like decoding M-S, create L and R from sum and difference. The subcarrier is more prone to interference so receivers can degrade gracefully to mono by ignoring the difference signal.

 

[John Willett]

...To a certain extent. The "true stereoness" depends on the interaction between the two mics. If you change the relationship by too much you end up with either mono, or the "fake stereo" which collapses to silence.

Yes

 

[Farview]

Isn't it correct that when played in mono the side mic portion cancels altogether leaving *only* the mid mic portion? If you have a L/R only signal that isn't specifically formulated for mono compatibility, doesn't that leave the possibility of unintended partial cancellations? That's what I came away with from various videos on the m/s method.

yes, but we are starting to confuse mid side micing with mid side encoding. with mid side micing, what you are saying is true. with mid side encoding of an existing left right signal, the mono part will only be as mono compatible as the original left right signal was in the first place.

Im sorry, I should not have brought up the encoding of normal left right signals, it just confuses the issue.