mid-side (MS) recording method
I’ve been experimenting with the mid-side (MS) recording method.
So I understand how to do it. When I look at the phase relationship of the three signals on a spectral form (in Ozone 3), the signal is out of phase. When I hit the mono activator just the middle mic is remaining. Should the overall aural picture be out of phase? Also, it seems that the channel which is not out of phase is louder than the one which has its polarity switched. When I make the channel which has had its polarity switched louder it throws the overall aural picture out of phase. Is this the way it is supposed to be?
The side channels should basically always be raised and lowered in volume together, since it’s really just the same audio with the polarity reversed on one side. The MS technique really only works when there is some combination of the side channels added to the mid channel. If you drop the mid-channel out, and then use your mono button, yes the sides will cancel each other out totally.
Did you have your mics set up properly when recording? The mid mic should be a cardioid pattern aimed in the direction of the sound source, while the side mic needs to be a figure of eight pattern aimed 90 degrees (perpendicular) to the sound source (to the side, basically). The mics should be as close together as possible to avoid phase issues, which is why they usually have one right on top of the other. If you had your side mic pointed in the same direction as the mid mic, that would explain the problem you have with the in phase side being louder when added to the mid mic than the out of phase side.
There is a really good explanation of the whole technique, along with matrixing and decoding on the EM site here:
If you drop the mid-channel out, and then use your mono button, yes the sides will cancel each other out totally.
So then the side channels should be out of phase?
Also, according to the article you recommended (and thanks so much) I may have had my guitar not centered and that could account for the lopsided nature of the recording.
Yes, the side “channels” are really just one mono audio channel that you duplicate and pan hard left-right, with the polarity of the right channel being reversed. Even though a figure-of-eight pattern microphone picks up audio from either side, it is NOT a stereo microphone, and there is only one channel of audio produced. So, all you are really doing is copying that channel and reversing the polarity (phase) of the copy. The fact that the two resulting side channels are out of phase is what gives you that wide stereo sound.
Many simple stereo widening processors make things seem wider by flipping the phase of one the channels and mixing some of that back in (but reversing the L/R channels first)… but those effects are not mono compatible and will cause cancellation of audio when mixed to mono.
The M/S technique gives you mono compatibility because of the mid channel. If you sum to mono, the side channels cancel out, but the mid channel is not affected at all. The mid channel carries the most important part of the audio since it is pointed at the source, while the side channel is there to mostly pick up the ambiance of the room since it is pointed 90 degrees away from the source, and thus the source should be right in the biggest null spot of the figure-of-eight pattern. You then basically make the ambiance seem wide and in stereo by copying that channel, panning it hard L/R, and flipping the polarity of the right channel. Since the M/S signals are separate, you can control the blend of the “direct” audio (the mid channel) and the “ambiance” (the side channels) to achieve the desired effect.
Hopefully that makes more sense.
Generally, in a small home studio, I don’t think the M/S technique is all that useful unless you’ve got a nice big room with great sounding ambiance, and can position the microphones appropriately to capture it all correctly. I usually end up going with a straight ahead X-Y configuration, or a spaced stereo pair, if I want to record something in stereo in a smaller room (or up close to the source). But, I should probably play around with the M/S recording technique more than I do.
It’s also useful to note than you can use plug-in M/S decoders on stereo files that have not been recorded in M/S to adjust instruments panned in the center of a mix along with the stereo elements separately. This can be useful in mastering, allowing you to do some extra compression/expansion to just the center, or stereo, elements of a mix, as well as to expand or shrink the stereo image a bit as desired.
That is perfectly clear now. Thanks so much. After hearing how nice this sounds why isn’t everytging recorded in stereo? Why stick an SM57 in front of a guitar cab when you could have a nice stereo signal?
Stereo recording techniques, such as M/S, are more useful for classical music, and for other types of music where you are recording solo instruments primarily, or relatively small groups, where you are looking for a more natural stereo type sound. As mentioned, they work best when you have a great sounding room of proper size to generate some nice ambiance.
For most modern “popular” styles of music, though, where instruments are overdubbed and layered, and many instruments are electric/electronic, there simply isn’t room in the sound field for everything to be recorded in stereo. For example, if you are recording synth based music, and you record everything from the stereo outputs of the synth, and keep them all panned hard L/R the way they come out of the synth, you end up with what we call “big mono”. You just get a big mess of stereo signals all competing for the same space in the mix. In popular music, many things are recorded mono, and then carefully positioned in the stereo field to make room for all the various instruments. When you want a really big wide sound, such as huge rock guitars, in popular music, you will usually double-track the part and pan one take hard left, and the other hard right… that’s the classic big and wide guitar sound… it’s not a “natural” sound, but it’s more wide and big sounding than using something like an M/S recording technique on one guitar amp playing the part just once. Double-tracking is often used on acoustic guitars as well… you get that big wide sound out of the guitars, which are panned hard L/R, leaving the middle of the stereo field open for vocals, bass, drums, etc. When recording popular music, about the only time you regularly might use a stereo microphone technique would be for an acoustic drum kit, with the overhead mics, or room mics…. but, even then, you might not pan them all the way hard L/R, especially if you’ve got double-tracked guitars already hard L/R.