Rooms, Reflections, Phase and Stereo

Read first: SoundStudio theory section 12, Stereo, phase and mono, plus the room tone row in the table in section 8, Noise. The tutorial covers panning, phase cancellation and the mono check. It does not cover rooms or microphone placement, because SoundStudio works on files that are already recorded. This lesson covers the part that happens before the file exists, which is where most of these problems are cheapest to fix.

A room is part of every recording

A microphone hears a voice plus every wall, window, tabletop and ceiling that voice bounced off on the way. The first bounces arrive a few milliseconds after the direct sound. Then thousands more arrive, weaker and blurred together, until the sound dies away. That tail is reverberation.

The standard measure is reverberation time, or RT60: how long a sound takes to fade by 60 dB after the source stops. Clap once in an empty gym and count; that is a long RT60. Clap in a closet full of coats and the clap just stops. Acoustics firms publish typical targets, and NTi Audio's table is a useful reference: about 0.3 seconds for a small recording studio, 0.4 to 0.6 seconds for a classroom, 1.4 to 2.0 seconds for a concert hall, and anywhere from 2 to 10 seconds for a church.

For spoken word, shorter is better. A bare room with hard walls and a wood floor can easily run near a second, and a voice recorded there sounds like it was taped in a stairwell. Noise reduction will not save it. AI tools can reduce reverb some, but they leave artifacts and they never fully restore the dry voice. So handle the room before you press record:

• Get closer. Halving the mic distance makes the direct voice much louder compared with the room. A speaking voice 6 to 8 inches from a mic beats the same voice at 3 feet in any room.
• Add soft stuff. Couches, rugs, curtains, a closet full of clothes. Bookshelves full of books break up reflections well.
• Stay off the walls. Do not record facing a bare wall from a foot away. The reflection comes straight back into the mic.
• Kill the noise makers. Fridge, furnace fan, air handler. Every one is rumble or hum you will spend time removing.

Comb filtering: when a sound meets itself

When a sound reaches a microphone twice, once directly and once a moment later off a hard surface, the two copies add together. At some frequencies the waves line up and get louder. At others, one wave is going up while the other is going down, and they cancel. The cancellations fall at regular spacing across the spectrum, which on a graph looks like the teeth of a comb. Hence comb filtering. It sounds hollow, thin, phasey, like the voice is coming through a tube.

You can predict the notches. Sound travels about 343 meters per second (about 1,125 feet). If the bounce travels 1 meter farther than the direct sound, it arrives about 2.9 milliseconds late. The first notch lands where that delay is half a cycle, about 171 Hz, then repeats every 343 Hz: about 514, 857, 1,200 Hz and up. Right through the middle of the voice.

Two microphones and the 3:1 rule

Two microphones open on the same voice do the same thing as a reflection. The voice arrives at the near mic first and the far mic a little later, and when you mix the two channels you get a comb filter. This is common in panels, interviews with two lavaliers, and bands with two vocal mics.

The fix has a name. Shure, the microphone maker, calls it the basic formula for miking any group: the 3:1 rule. Keep the distance between microphones at least three times the distance from each mic to its source. If a singer is 1 foot from her mic, the next mic should be at least 3 feet away. At that ratio the voice reaches the far mic quiet enough, compared with its own mic, that the comb filtering stays mild. Treat it as a rule of thumb. Closer miking always makes it easier to follow.

Mute or pull down any mic that is not in use, because an open mic on an empty chair is just a second path for everyone else's voice. When you have separate tracks, line them up or simply use the one close mic for each voice instead of blending both.

Stereo and mono

The tutorial covers this well. The short version: a sound equally loud in both channels seems to come from the center. If the left channel carries an exact upside-down copy of the right, adding them gives silence. That is the extreme case of a phase problem, and milder versions are common: a stereo widener plugin, two mics on a guitar at different distances, a wire swapped somewhere.

Phones, smart speakers and a lot of radios play mono. If your mix falls apart there, a large share of your audience hears the broken version. Spoken word is simple: record voices mono, keep them centered, and a lot of these problems never happen. Music and stereo ambience are where you need to check.