DI or Microphone: How to Choose for Live Sound

The DI versus microphone question gets answered as a matter of taste, and it mostly is not. Live, the choice is driven by bleed, feedback, and consistency, and those are conditions you can read off the room rather than preferences you argue about. The studio answer and the live answer are frequently different for the same instrument, which is why advice imported from studio sources leads people wrong.

What a DI actually buys you live: isolation. A direct box takes the instrument signal electrically, before it becomes sound in the room, so it picks up nothing else. No drum bleed, no guitar amp, no crowd. That isolation is what gives you real control over that channel in the mix, and it is also what removes that channel as a feedback path, because there is no open microphone.

What a microphone buys you: the sound of the instrument in the air, including the body of an acoustic instrument and the character of an amplifier. A DI on an acoustic guitar with an undersaddle pickup captures the pickup, which is a thin, quacky representation of a guitar that a microphone would render much more naturally. The tradeoff is that the microphone also captures everything else on the stage.

The decision rule that works: the louder the stage, the more the DI wins. In a quiet room with a controlled stage, a microphone on an acoustic instrument sounds better and the bleed is tolerable. In a loud room with drums, the microphone is capturing so much of the rest of the band that the instrument you wanted is a minority of what you get, and the DI is the only version of that instrument you actually control.

Feedback is the other half of the rule. An acoustic guitar with a hollow body in front of a loud PA is a resonant system waiting to feed back, particularly in the low mids around 100 to 250 Hz. A DI removes that loop entirely. This is why acoustic acts at volume run pickups rather than mics, and why the same act in a listening room can use a beautiful microphone instead.

Blending both is standard, and it is where the phase problems start. The instrument arrives at the DI electrically, essentially instantly, and at the microphone acoustically, delayed by the distance from the instrument to the mic. Sound covers roughly one foot per millisecond, so a mic a foot away arrives about a millisecond late. Sum those two signals and the delay causes comb filtering, a series of cancellations that make the blend sound thinner and hollower than either source alone.

The fix is delay compensation, and every digital console has it. Delay the DI channel to match the microphone arrival, in the range of one millisecond per foot of mic distance. Start with the measured distance, then adjust by ear while listening to the two summed and mono. When you hit alignment the low end firms up and the hollowness disappears, and the change is much more obvious than people expect.

A faster way to hear it: sum both channels to mono, flip the polarity of one, and adjust the delay until the sound gets as thin and hollow as possible. That is maximum cancellation, which means maximum alignment. Flip the polarity back and you are aligned. Chasing the worst-sounding point is easier to hear precisely than chasing the best-sounding one.

When not to blend at all. If you do not have time to align, do not blend. One source alone always beats two misaligned sources summed, because comb filtering is worse than either signal on its own. Under a 15-minute soundcheck, pick the source that fits the room by the rule above and move on. Blending is a refinement, not a default.

Amplifiers follow the same logic with different weights. A guitar amp mic-ed close is standard because the amp is the instrument and a DI before it misses the point. But a loud amp in a small room may already be filling the space without reinforcement, in which case the mic exists to give you a little control rather than to carry the guitar, and it should sit much lower than instinct suggests.