Two philosophies of recording a live performance sit at opposite ends of the stage, and most arguments about the results are really arguments about which one was chosen. A pair of microphones placed back in the room captures the event as the hall assembled it. A microphone on each source captures the sources and leaves the assembly for later.
The main pair records the room's version
Placed at a distance, a stereo pair receives direct sound and the hall's reflections in roughly the proportion a listener in a good seat receives them. The blend between players, the balance the conductor shaped, the depth that comes from some instruments being further away: all of it arrives already done.
Established techniques differ in how they build the stereo image. Coincident arrangements place two directional capsules at the same point and derive direction from level differences alone. Near-coincident arrangements separate them slightly, adding small time differences. Spaced arrangements set them well apart and rely mostly on timing. Each has a characteristic width and centre stability, and each has been documented in the recording literature for decades.
What they share is that the balance is fixed at the moment of recording. If the second horn is buried, the second horn stays buried.
Close microphones record the players
A capsule near a source hears mostly that source, which is the whole point and the whole problem. It delivers a high ratio of direct to reverberant sound, isolation from neighbours, and complete freedom to change the balance afterwards.
It also delivers a perspective nobody in the building had. A violin heard from three hundred millimetres away is not the instrument the audience heard, and an orchestra assembled entirely from such perspectives sounds detailed, close and slightly unreal, because it is a place that does not exist.
The interference nobody planned
Once more than one microphone hears the same source, the two copies arrive at different times and sum with each other. The result is cancellation at some frequencies and reinforcement at others, heard as a hollow or phasey colouration rather than as an echo.
The long-standing guideline in recording practice is to keep the distance between two microphones at least three times the distance from either to its own source, which reduces the level of the spill enough that the interference stops being audible. It is a rule of thumb from the field rather than a standard, and it fails in the direction people expect: crowded stages.
The hybrid, and the delay that makes it work
Most serious live recording uses both. A main pair supplies the sound of the event, and spot microphones are added underneath at low level to recover detail and rescue anything the pair could not resolve.
The problem is arrival order, and it is the same mechanism as the precedence effect described in the sound section. A spot microphone is close to its source, so its signal reaches the recording before the main pair's does. Left alone, the ear localises to the spot and the ensemble collapses toward whichever instruments were spotted. Delaying each spot so that it arrives with the main pair, rather than ahead of it, keeps the image where the pair put it. The spots are then doing what they were meant to do, which is adding detail rather than taking over.
Placement is decided in a room the recording team may not control
Both approaches assume the microphones can go where the technique requires. In a concert hall that is usually negotiable. In a staged production it frequently is not: a main pair hung where it would hear the ensemble properly is in the audience's sightline, and spot microphones placed where an instrument needs one are visible from the stalls or in the way of the blocking.
The result is that live recording in a theatre is a series of compromises against the staging rather than a clean application of technique, and the recordings that succeed are usually the ones where the capture requirement was known early enough to influence where things were put. Arriving on the day with a preferred technique and no negotiating position produces the recording the room allows, not the one the technique promises.
The audience is part of the record, or it is not
Room microphones hear coughs, seats, applause and the particular silence of a full house. Close microphones hear far less of it. Which of those is a fault depends entirely on what is being made.
A document of an occasion that removes every trace of the occasion has thrown away its subject. A commercial release built from the same evening usually wants the applause moved, shortened or removed altogether. The distinction between those two products, and the different decisions each one justifies, has its own article in this section.
What we cannot verify
Microphone technique in the recording field is transmitted through practice, teaching and published technique literature rather than through standards, and the geometries named here are conventions with many documented variants. Performance claims for specific microphones come from their manufacturers and are measured in conditions unrelated to a hall. Nothing here substitutes for listening in the room and then listening to the recording, which is the only test that reveals the gap between them.
The short version
- A main pair captures the balance the hall already made; a spot captures a source.
- Coincident, near-coincident and spaced pairs build the image from level, or time, or both.
- A pair fixes the balance at the moment of recording.
- Close perspectives are detailed and describe a place nobody sat in.
- Two microphones on one source interfere; the three to one guideline limits it.
- Spots must be delayed to the pair, or the image collapses onto them.