Mid-Side Recording Demo

[Doug Young]

Originally Posted by min7b5

Question for you Doug;
In the tutorial you say that when using the Voxengo decoder plugin and setting the gain at the preamp, that you want to match them with the figure of eight microphone temporarily turned to face forward. And then after matched to the side microphone you rotate it back to be being perpendicular. That whole setup has worked great for me. But I am now playing around with using a manually setup mid/side matrix in Logic so I can playing around with dual mono reverbs on just the side microphone tracks, among other fun experiments.. And I'm wondering if there's a reason to use the same gain setup process, or is that unique to using a decoder plugin?

Hey Eric, nothing to do with the plugin. It's related to the mid-side physics. If you had matched mics, you wouldn't need to do this, you just set the gain to the same levels for both mics - the value on the dial. When you do that, and one mic is turned to the side position, the side channel should be about 6db down if you look at the recorded waveform. Then when you do the mid-side to stereo transformation, whether with the plugin or using three channels and doing it the manual way, the side signal gets added back in twice, bringing it partly back up (should come up by 3db, if I'm recalling right). The remaining 3db "loss" from having the mic turned is because the side info is inherently less than the mid component in any stereo signal.

You could take an XY-recorded track, feed it thru the Voxengo in "encode" mode, and you'd get the expected MS input, with the side signal down by around 6db from the mid.

The reason to do the calibration step of setting the mic levels before turning the side mic is that not all mics have the same output levels, unless you're working with relatively matched pairs. I don't bother to do this is I'm using my Schoeps, for example, because the output is the same, and I have no way to do it if I use the AT 4050 mid-side mic, but I know both capsules are the same on that mic. So I just set my gain on both channels to the same thing, say 40db, and it's good, even tho the VU meters will show the side mic as being weak on the un-decoded input.

If you do this calibration (or have matched mics), then setting the mid and side levels in the plugin to 0 results in an equivalent of XY stereo. If you raise the sides a few DB, it gets close to ORTF. And there's nothing to keep you from picking different values, anything from a narrow image (by turning down the side level) to ultra-wide, by turning up the side. If you do the mid-side decoding manually using 3 tracks with phase reversal, you'd set the playback volume of all tracks to the same to achieve the same nominal "XY"-ish sound.

Hope that's not too confusing....

[min7b5]

Originally Posted by Doug Young

Hey Eric, nothing to do with the plugin. It's related to the mid-side physics..

Wowza. Thanks boss. You da man