Seeking advice and input for M/S recording

[Trevor B.]

Here's a short excerpt from my arrangement of Jock O'Hazeldean. I recorded this with an unmatched spaced pair; both mics 14" inches from my guitar. A Gefell M300 was pointed at the 12th fret and a Sennheiser MKH800 P48 pointed mid-way between the sound-hole and the bridge. So the signal chain was a Crafter GLXE 3000sk (steel string)>Gefell M300 & Sennheiser MKH800 P48>Apogee Quartet>2010 White MacBook>Logic Pro X. I used the Direction Mixer plugin to decode the M/S. I added a high pass filter and some reverb with the Space Designer. My recording space is as of yet acoustically untreated, however; baffles, bass traps and clouds are under construction as I type this and should be installed and fully functional by this time next week. What I've posted here represents my initial baby steps in the firmament of digital home recording. There's a lot to learn but I'm loving the journey so far.

Any advice and input is very welcome.
Thanks,
Trevor

[Doug Young]

Hi Trevor, you're off to a good start. The sound is a bit distant and a little harsh. I suspect both are at least partly your untreated room. The reverb may also not be really helping.

I'm a little confused about how you say the mics are positioned. Usually, you place MS mics very close together. The idea - in theory - is that the mics occupy the exact same spot, so in practice, you just try to have them really close. I place my MS mics right dead center on the guitar, above the sound hole and as close to each other as I can - remember that photo?. By separating them like I think you're doing, you're probably introducing phasing issues that are also contributing to the distant and/or harsh sound. The setup you describe sounds like spaced pairs, not MS, so that'd be fine if you set both mics to cardiod, but then you wouldnt use an MS decoder.

[Trevor B.]

Originally Posted by Doug Young

Hi Trevor, you're off to a good start. The sound is a bit distant and a little harsh. I suspect both are at least partly your untreated room. The reverb may also not be really helping.

I'm a little confused about how you say the mics are positioned. Usually, you place MS mics very close together. The idea - in theory - is that the mics occupy the exact same spot, so in practice, you just try to have them really close. I place my MS mics right dead center on the guitar, above the sound hole and as close to each other as I can - remember that photo?. By separating them like I think you're doing, you're probably introducing phasing issues that are also contributing to the distant and/or harsh sound. The setup you describe sounds like spaced pairs, not MS, so that'd be fine if you set both mics to cardiod, but then you wouldnt use an MS decoder.

Thanks for responding, Doug. My approach here was clearly wrong, vis-a-vis the spaced pair vs coincident placement. An old guy's oversight. I won't get to do another run at this for a few days but will try again and repost a new clip then.

[Wengr]

Hi Trevor. Imho, it's usually best for evaluation purposes to skip reverb.
Also, it would be really cool if you could record a dry track, and then rerecord the same dry track via the same method again, after your room treatment goes up for comparison purposes.

[ChuckS]

Originally Posted by Trevor B.

Thanks for responding, Doug. My approach here was clearly wrong, vis-a-vis the spaced pair vs coincident placement. An old guy's oversight. I won't get to do another run at this for a few days but will try again and repost a new clip then.

It might be helpful to look at some of the sites that describe Mid-Side recording: some of them have visualizations of the microphone orientation and their pickup patterns. For me it was helpful to review those, and then do the Mid-Side decoding manually with my DAW: it seemed to help to get the 'theory' to sink in. You need to use a figure 8 for the Side mic.

The Mid mic has center, left, and right information. The Side mic has left information and right information (since left and right come into the mic from opposite sides their electrical polarity are 180 out of phase to each other). When you duplicate the Mid-Side track you invert its polarity. So you then have 3 tracks:

Mid witch contains ...................... +Left +Center +Right
Side (original) which contains ...... -Left ............ +Right (assumes the + side of the figure 8 mic is pointed to the right side of the source)
Side (duplicated) which contains .. +Left ............ -Right


For your stereo output you:

Combine Mid and original Side for stereo right channel:
Mid witch contains ...................... +Left +Center +Right
Side (original) which contains ...... -Left ............. +Right

Combine Mid and duplicated Side for stereo left channel:
Mid witch contains ...................... +Left +Center +Right
Side (duplicated) which contains ... +Left ............ -Right

In your DAW, if you link your original Side and the duplicated Side (i.e. make them a stereo track) then you can increase/decrease their level by the same amount. The more Side channel amplitude you add to Mid, the more left and right separation you get in the stereo output. For example: in the right stereo track the addition of the +Left from the Mid mic and the -Left from the original Side mic reduce the amount of Left signal; the addition of the +Right from the Mid mic and the +Right from the original Side mic increase the amount of Right signal.

Clear as mud?

[Trevor B.]

Thank you, Wengr, for your comments and advice re: posting a dry track. Seems there's a consensus on that amongst the folks who know more than I.
I'd also like to thank Chuck for elaborating the process. Strange as it may seem I actually did read several online articles (SOS, uaudio, Logic Pro Help, etc) and followed all the steps and was successful at achieving the mid/side stereo image but couldn't abide some of the legato holes in the take. So I fixed those and decided to change my approach to record the follow-up. Instead of following all the steps I recorded a stereo track and used the Direction Mixer plugin (utility?) to decode the M/S. Who know why but in what can only be described as a seniors moment gone berserk I spaced my mics rather than placing them one right above the other both pointing at the upper waist of my guitar. Here's a screenshot that shows the whole experiment to that point.
http://s1362.photobucket.com/user/Gu...4bae3.png.html

[Doug Young]

Your screen snap shows another important point. You need to have the mic levels set to that if the side mic wasn't turned sideways - for example, set to cardiod mode and aimed as a normal cardiod mic - that the levels are the same between mics. Then when you are in figure-8 mode and turned sideways, the level will drop a lot. When you record mid-side, the side signal is quite weak compared to the mid - that's expected, the mid carries the bulk of the sound, the sides are ambient information. You have the opposite here, the mid signal is quieter than the sides.

[Trevor B.]

Originally Posted by Doug Young

Your screen snap shows another important point. You need to have the mic levels set to that if the side mic wasn't turned sideways - for example, set to cardiod mode and aimed as a normal cardiod mic - that the levels are the same between mics. Then when you are in figure-8 mode and turned sideways, the level will drop a lot. When you record mid-side, the side signal is quite weak compared to the mid - that's expected, the mid carries the bulk of the sound, the sides are ambient information. You have the opposite here, the mid signal is quieter than the sides.

I really struggled trying to get the Gefell mic volume louder than the Sennheiser without almost turning the Sennheiser off. I'll try your suggestion of balancing both mics straight on and set to cardiod before turning the side mic to a 90 degree angle and switching it to figure 8.

[Doug Young]

Here's a snap I took at some point of a stereo track MS signal. Mid on top, side on the bottom. Gives you an idea of the expected relative levels:

[Trevor B.]

Originally Posted by Doug Young

Here's a snap I took at some point of a stereo track MS signal. Mid on top, side on the bottom. Gives you an idea of the expected relative levels:

That's certainly a significant discrepancy. And an image to winm for during the calibration stage.

I'm going to try and post the screenshot right into this post.