Mid-side Recording Question

D

DaveLeeNC

New member
I would like to try MS recording on my classical guitar. I have a pair of Octava SDCs and a Blue Yeti Pro stereo (usb) mic. If I were to set the Yeti to fig. 8 for the Left-right data (left/right channels of a stereo track), and record one of the SDCs mono into another (mono) stereo track, all I need to do is to mix these two together. No phase reversals or other processing required. That right? I am assuming here that the Yeti will be putting the left side sound on the left channel, vice versa.

FWIW, I would probably use Audacity for the Yeti mic track, and Reaper (through a 2i2) for the Mid track.

dave
 
To do MS recording, record the figure 8 mic to two tracks at the same time (or on a single track and copy it to a second track). You then invert phase on one channel, and pan them 100% left and right. Now you can add in the SDC mono input, which is your center channel. By varying the amount of center channel relative to the L/R channels, you can adjust the apparent stereo spread.

When I record the figure 8, I combine the two figure 8 channels into one stereo track, then it's really easy to adjust the center/side levels without worrying about keeping the two L&R channels equal.

UA has a nice description of the process here: M/S recording

As for doing the recording on two different programs at the same time, the issue I see is making sure you have everything aligned properly. A clacker might be needed to provide a reference transient since there's no way to start both programs at precisely the same moment. Then you need to make sure the two programs will record at the same time, and that they track time consistently.
 
If the Yeti mic in Figure-of-8 mode sends two (stereo) tracks, then the conversion is (I assume) done in the mic's preamp. If it only sends a mono-signal, then the normal process in your DAW is to duplicate the mono track, and flip the polarity (aka "phase") of the duplicate. Hard pan the two tracks L-R. Your "M" track is mono, and whatever mic you use to record that should be set to cardioid mode, or use a cardioid capsule if the SDC has interchangeable ones.

Using two different recording apps, I would use something like a video clapboard, or make something with a couple pieces of wood and a hinge. Do the clap from the center of your "soundstage" at the start and end of recording, with both apps running. After putting all the tracks into your DAW, zoom in as far as you can while still keeping the peak of the clapboard obvious in the waveform, and assure the phase of that piece of information is fully in phase with the center and one of the S mics, while the other S mic is 180 out of phase, i.e., the sound from the center of the space will be nulled out of the two S tracks.

I would ask if you are recording in a large space that sounds really good, because, personally, I don't find M/S recording that advantageous in home recording scenarios. It really works best when you need a stereo image with excellent mono compatibility, or you really want to capture more of the space. Since you're recording a single source, I have to assume it's the latter, because a spaced pair usually gives me better results in smaller spaces when I (rarely) need a stereo recording of a single guitar. My 2¢.
 
The timing of the mid and side has to be effectively perfect for the proper phase relationship to be correct. That would normally mean recording the two inputs through converters that are clocked together. Even a little drift will make the image shift over time.

A possibly acceptable alternative would be real time sample rate conversion like using aggregate device on a Mac.
 
First question for M/S success. What does the room sound like? This in a guitar will be the main feature. In fairness, if the room is gorgeous sounding, or the room is totally dead, then M/S lets you set the width of the sound source afterwards, but in a guitar, what actual sound spread do you really have? Tone from the sound body end, then the hole area then the sratchy stuff from the fingerboard. Do you want to really add that murkey sound and the tinny sound to the central darker area? It would record well, but through speakers even in nearfield, the guitar would sound pretty big, so you'd lower the side mics making it more mono, so what would the point be? Real piano, harps, even drums, that have at least speaker width dimensions in real life work well and it's quite realistic, but M/S on smaller instruments is a sort of audio magnifying glass. Record big things and it's a great tool, but for a typical guitar - which isn't really and instrument with real width, just different tones from different parts, why not just two mics, add a tiny bit of pan and blend the two together. Most real stereo techniques go very strange when used close in. I saw somebody recording a violin the other day with a close X/Y pair. About a foot from the violin. When the player stood ready to start, one mic was pointing to his chest, the other pointing to a wall. Then he started, and he wasn't a really wave around player, but the left mic went from his boddy to the bridge area, the right from the wall to the bridge area. I smiled thinking that in the mix, both mics would have to be centred on the pans or the listeners would have been sick, especially on headphones. Guitars are not as bad because they don't move that much, so two close mics in X/Y just act as two mics with different aiming points - not as 'stereo' mics. On a forum people were discussing X/Y or ORTF as a close mic technique, missing the entire point that close in - they are just two mics. All that amazing stereo science going right out of the window. How many drum kits have you miked up where by accident, two mics are at 90 degrees to each other, with their ends closeish? Loads, I bet - they're not stereo mics are they?
 
I have read there are ways to circumvent the one-soundcard-at-a-time limitations of most computers - have you already figured out how to use the USB mic AND the 2i2 interface at the same time? Tracking into 2 DAWs at the same time?
Since you are talking about a nylon string guitar, the usual best sound comes from 12-24" out from the guitar - by that time, the sound from body, strings and neck has pretty much blended together, but you could try just tracking with the two SDCs at one time, one out form the neck/12th fret position, the other pointing towards the body behind the bridge.
 
Thanks for all the helpful responses. After doing more research I have decided that M/S recording of a classical guitar in my small/acoustically crappy room is unlikely to be a good idea. So far AB recording at the bridge and neck about 2 feet out seems to work the best, so I think that I will stick with that for now.

dave
 
Oddly, I never had any luck till an accidental recording I made of a guitar in my very dead office studio. I'd got a few mics set up for a video and one was an M/S that accidentally was at around the same distance as you just mentioned - if you have the kit available, try it and see, against something else and have a listen. For many years I have had a zoom H6 and when I tried out the M/S capsule, it sort of worked as long as you didn't really push the width fader when it just sounded wrong. Try it and discount it if you hate it. Always good to have a mental list of failures. My very first Decca Tree recording that worked was this year - after 40+ years. I had discounted it when I was much younger, because I used it in the wrong venue. When I found the right one, I was really impressed.
 
You must log in or register to reply here.
Back
Top