Multitrack Drum Recording: Phase Coherence & Mic Placement Guide
Recording an acoustic drum kit is the most complex physical challenge in audio engineering. With 8 to 16 open microphones capturing one acoustic source simultaneously, phase cancellation can destroy low-end punch and smear stereo imaging. This guide covers physical phase alignment, overhead geometry, polarity inversion, and cymbal bleed management.
1. The Physics of Phase Cancellation & Comb Filtering
Sound travels through air at approximately 343 meters per second (1.13 feet per millisecond). When a drumstick strikes the snare drum:
- The snare top microphone (2 inches away) captures the transient in 0.15 milliseconds.
- The stereo overhead microphones (4 feet away) capture the exact same strike 3.5 milliseconds later.
- When these two signals are summed together in the DAW, the 3.35 ms time delay creates severe comb filtering, cutting out specific fundamental bass frequencies (such as 150 Hz) and making the snare sound paper-thin.
2. Overhead Microphones: The True Anchor of the Kit
A fatal beginner mistake is treating overheads as "cymbal microphones". Professional engineers build the entire drum mix around the overheads:
Spaced Pair (A/B)
Two condenser mics placed 40 inches above the cymbals. The Golden Tape Rule: Measure with a mic cable from the center of the snare batter head to both capsule tips. Both overhead capsules must be exactly equidistant from the snare center.
The Recorderman Technique
One mic directly above the snare (32 inches) and a second mic over the drummer's right shoulder pointing at the snare. Both mics are positioned exactly equidistant from both the snare and kick beater using a drumstick guide.
The Glyn Johns Setup
One overhead 4 feet above the snare, a second mic placed near the floor tom pointing across the snare, plus close kick and snare spot mics for vintage organic punch.
3. Snare & Kick Polarity Inversion (The 180° Rule)
Dual-miking drum components requires immediate polarity checks:
- Snare Top vs. Snare Bottom: When the stick hits the top head, the head moves downward (pushing air toward the top mic). Simultaneously, the bottom resonant head bulges outward (pushing away from the bottom mic). The signals are 180° acoustically out of phase. Always engage the Polarity Reverse (Ø) button on the snare bottom preamp.
- Kick In vs. Kick Out: The inside dynamic mic (capturing beater click) and outside sub-kick (capturing low sub-bass) must be checked in solo. Flip the phase button on the outside mic: whichever setting yields fuller low-end bass is the correct alignment.
4. Close Mic Placement & The 3:1 Distance Rule
To eliminate unwanted cymbal spill into tom microphones:
Aim the cardioid microphone's rear rejection null point directly at the nearest crash or hi-hat cymbal. Apply the 3:1 Distance Rule: the distance between any two microphones should be at least three times the distance from each microphone to its intended drumhead.
Tested: the 3.35 ms snare delay, measured
We built the example from section 1 in FFmpeg 8.1.2 on 2026-09-23: one signal split into a close mic and an overhead that arrives 3.35 ms later (161 samples at 48 kHz), then summed as a DAW mix bus sums them. We measured the frequency response of the sum at four overhead levels. A flat response would read the same at every frequency.
| Overhead level in the sum | First notch | Level at the notch | Highest peak |
|---|---|---|---|
| Same level as the close mic | 149 Hz | -56.8 dB | +6.0 dB |
| 6 dB lower | 149 Hz | -6.0 dB | +3.5 dB |
| 9.5 dB lower (the 3:1 rule) | 149 Hz | -3.5 dB | +2.5 dB |
| 20 dB lower | 149 Hz | -0.9 dB | +0.8 dB |
| Same level, close mic delayed 3.35 ms to line up | none | +6.0 dB | +6.0 dB |
The first notch fell at 149 Hz every time, which is one divided by twice the delay, and the notches repeated every 298 Hz above it. The level decided how deep they went. With the overhead as loud as the close mic, 149 Hz dropped to -56.8 dB, close to silence. The 3:1 rule puts the second mic three times as far from the source, 9.5 dB quieter by the inverse square law, and that left ripple of -3.5 to +2.5 dB. At 20 dB down the ripple was under 1 dB.
Delaying the close mic by the same 3.35 ms removed the comb completely: the sum read +6.0 dB at every frequency from 20 Hz to 20 kHz. Flipping the overhead's polarity did not. It turned 149 Hz into a peak of +6.0 dB but put a notch of -50.8 dB at 298 Hz and cut 20 Hz by 7.6 dB. The polarity button in section 3 fixes two mics facing opposite ways. A difference in distance needs a time shift or a change in placement.
To shift a close mic later by 161 samples in FFmpeg: ffmpeg -i snare-top.wav -af adelay=161S snare-top-aligned.wav
Validated. Measured 2026-09-23 with FFmpeg 8.1.2 on macOS, Apple silicon. An impulse through a pure delay, so this shows the delay alone. A real overhead also hears the room and the rest of the kit, which blurs the notches.
Where This Fits
Phase between drum mics is a distance problem, and Audio Recording Fundamentals explains how distance shapes a single mic before you add the rest.
Working Vocabulary
Key acoustic drum tracking and phase engineering terms:
- Phase Cancellation. The acoustic phenomenon where two identical sound waves arriving at different times cancel out shared frequencies.
- Comb Filtering. A series of deep frequency notches and peaks created when a direct sound is combined with a time-delayed replica.
- Polarity Inversion (Ø). Reversing the positive and negative electrical terminals of an audio signal by 180 degrees.
- Transient. The initial high-amplitude, short-duration peak at the beginning of a percussive sound wave.
- Glyn Johns Method. A classic four-microphone drum tracking technique developed for The Rolling Stones and Led Zeppelin.
- Bleed. Unwanted acoustic leakage from adjacent instruments entering a close spot microphone.