still some M/S issues to resolve

If two signals of the same polarity are panned left and right then they wont cancel. If you pan them both into the centre they will add and you will end up with a mono signal that is 6 dB louder. 

If you invert the polarity of one of the sides and pan them both to the centre they will cancel and you should hear silence. Now with one of them inverted if you pan them L and R you will hear them both again except you will get out of phase sound which sounds sort of strange.

When they are both centre and you are hearing nothing then the sides are set-up correctly. Pan them both back to L and R and bring the M signal back in (centre). When you do that you will hear normal stereo again.

If the levels for both the Sides and M signals are the same, the sides and the M signal all need to be the same level to get the correct stereo back. ie all three mixer channels the same. A good thing to do is to group the two sides so you only have to move one of the sides and the other will follow.

Now you can play with the width. By increasing the M and reducing the sides, the signal will narrow down towards mono. By increasing the Sides and reducing the M signal the sound will get wider. You always need the M signal present otherwise you are just hearing the sides with that in phase and out of phase sound which is not very useful. ie the Sides don't sound good without M. Only M can be on its own really.

Hope that helps. I still suggest forget the silly 3 channel mixer decode concept and just feed the M S recording into a plugin to decode back to stereo. You can still vary M and S levels even with the decode plugin. It is far easier and you cannot go wrong. There is room for error with the 3 channel decoding concept.

If you are doing the recording using an M/S mic configuration you should not have any issue using 3 mono tracks sent to a stereo buss.  If you can figure out how to mix it this way I think it will help you to understand the concept.

As we discussed:

-Record your sound source with 2 mics.  A mid and a side mic.
-The mid mic should point at the sound source and the side mic (set to figure 8) should be 90 degrees to the source.
-The track for the mid-mic gets panned right down the middle.
-The track with the side-mic gets panned hard left.
-Copy the side-mic track into a third track, invert its polarity and pan it hard right.
-Make sure you label your tracks (stating the obvious I hope)
-Send the 3 mono tracks to a stereo buss.

Now do some listening experiments with these tracks ...

-Mute the side tracks and listen and listen to just the mid track.
-Mute the mid track and listen to just the side tracks.
-Pan the 2 side tracks to the middle (with the mid-track still muted) and if everything is set up correctly the 2 sides will cancel and you should hear silence (this simulates folding to mono).

If the M/S configuration is used during the recording process, the resulting stereo mix is completely mono compatible.  That is … the “side-mic” components in the resulting 2 channel stereo mix cancel each other completely leaving only the “mid-mic” component when the stereo mix is folded to mono. 


*As a side note (pun intended), go back to the listening experiment where you mute the mid-mic  and listen to the 2 side-mic  channels that are copies of each other with reversed polarity.  There is another name for this sort of processing if you were to use it NOT in the context of a M/S recording … it is called the “out of speaker trick” and is described on page 174 of Izhaki’s Mixing Audio.