DEBRIS.COMgood for a laugh, or possibly an aneurysm

Over the weekend I finally put some cables onto those microphones and recorded about a million takes of the dulcimer track for Ode to Soup. I was about halfway through when, during a playback with the guitar track enabled, I tapped the “mono” switch on my deck… and was surprised to hear the dulcimer virtually disappear from the mix.

I know a little about phase problems: any time two mics are used to record one sound source, care must be taken that the mics are in phase with each other. If two mics are perfectly out of phase, the signal recorded by one will zero out the signal recorded by the other. Although that never happens (except maybe in lab recordings of pure sine waves), most stereo recordings are probably partially out of phase, especially when the engineer is an amateur.

Listening to playback in mono is, as far as I know, the only easy way to check for phase coherence: if the recorded sound changes tone or volume dramatically when the playback is toggled from stereo to mono, there is probably a phase problem, at least for some segment of the frequency spectrum.

My mic placement was giving me a nice stereo spread, which of course goes away in mono. So I had heard the sound change when I listened back to an early scratch recording in mono, but I attributed it to the fact that the stereo field had compressed to a single point.

With the guitar in the mix, though, I heard a bigger problem: some chunk of the dulcimer’s frequency range must be getting mashed, because the relative volumes of the two instruments shifted in mono.

Focus on the dulcimer in the clip provided below. To my ear, the dulcimer all but drops out in the mono section. The dulcimer sounds forward in the mix at first, but at the repeat the guitar comes out in front.

Phase problem (2 bars stereo; 2 bars mono)

Tonight I tried six new mic positions, in an attempt to find a more phase-coherent setup. Making the phase problem go away is easy; it’s a matter of putting both mic capsules next to one another above the middle of the instrument. But this approach has its own problems: I don’t like the tone captured above the instrument, and the stereo field is too narrow. I much prefer the sound with the mics spread apart as pictured previously.

A bandmate suggested inverting the phase on one of the channels on the out-of-phase recording. Another possibility: nudge or delay one channel by a few milliseconds. I’m hesitant to try either trick, because I’d much rather learn how to record great tracks, than learn how to record deficient and flawed tracks that have to be surgically corrected later.