What's happening is that the M and the S are recording too close to the same signal. Perhaps you are too close to the source or didn't really have the S mic perpendicular to the M or had the whole M/S pair twisted closer to Blumlein. If you look at a closeup of your two waveforms I think you'll see that the M is very close to the L.

Could the side mic have been set to cardioid rather than figure-8?

bouldersound, post: 376082 wrote: Could the side mic have been set to cardioid rather than figure-8?

That is another possibility, but the time I had this same thing happen is with a Beyer 130/160 pair. I usually set it 2-3 feet from the top of the guitar for M/S, but I had a guy with no fingernails playing nylon string and I had to move in real close. The signals from both mics were too close to being in phase - even with the cardioid and the figure 8. Bottom line is that it was too close for real stereo information. (Mono signal was fine though.)

Thanks for all the great answers. The mic was set properly for figure 8 and perpendicular but believe I was too close to get proper stereo image. That makes sense since the guitar body(R side) is louder than the neck(L side) at that close position. I was thinking that since one side was out of phase that it was also cancelling the same side out on the Mid mic resulting in the volume difference. I'll keep working on it and I appreciate your time and input.

That doesn't sound quite the right explanation, as in your first post you said the L side became louder when you raised the level of the M channel. Was your guitarist left-handed?

The pickup pattern of a Fig-8 microphone has one lobe out of phase with the other lobe. At the M-S decoder, the phase inversion applied to the S input in the R channel to generate the width information automatically corrects this relative to the M channel at all microphone distances that the M and S diaphragms can be considered to be coincident. To make this work in practice in a situation where the sound source has a broad area, you should operate an M-S pair at a range such that the distance between the M and S diaphragms is much less than the distance from the microphones to the nearest point of the sound source.

I've used a pair of Rode NT2-A LDCs in M-S configuration very successfully on acoustic guitar with the mics positioned as close together as possible (a diaphragm centre spacing of around 2 inches) at a distance of around 15 inches from the guitar body. The sound field narrows but does not shift off-centre as you raise the level of the M channel in the decode.

Can you tell us the make and models of the microphones you used for this recording, and how you positioned them as an M-S pair?