The choice between one camera and several is usually discussed as a question of production value. It is more usefully understood as a question of authorship: a cut decides where the viewer looks, and the stage director had already decided that.
What a locked wide preserves
A single fixed shot containing the whole stage keeps the composition the production made. Spatial relationships survive: who is upstage of whom, how far apart two people are standing, what the chorus is doing during the aria. The viewer scans the frame themselves, which is what they would have done in the theatre.
It also keeps every option open afterwards. A wide shot can be cropped later. A close shot cannot be widened. For any recording whose eventual use is not yet settled, that asymmetry is the strongest argument in the subject.
The costs are real. Faces are small, and on a phone they are very small. Text delivered quietly does not read. Two hours of an unchanging frame asks more of a remote viewer than of someone in a seat, because the seat came with a body, a room and an interval.
What a cut buys and what it spends
Cutting delivers faces, detail and rhythm, and it can rescue a production that reads poorly at a distance. It is also an interpretation laid over an interpretation. Every cut asserts that this is the important thing now, and it removes the audience's ability to disagree.
Where the staging depends on simultaneity, that removal is destructive. A reaction happening upstage during a speech is the point of the scene; a recording that is on the speaker's face for its duration has recorded the words and lost the play.
A multicamera shoot is a second production
It needs operators, someone cutting, communications, and a shot script marked against the text or the score in the same way a calling script is. It needs rehearsal with the show, because an operator finding a move for the first time during a performance will miss it. And it needs a technical rehearsal of its own, which competes for the stage time every other department also wants.
Productions that treat this as something that can be added to an existing schedule produce the recognisable result: correct shots arriving two seconds late, all evening.
What a small crew can actually hold
The realistic middle path is to stop cutting live. Record every camera in full for the whole performance, and decide afterwards. This removes the single hardest requirement, which is making irreversible decisions in real time with no rehearsal, and it converts the problem into editing time, which is cheaper and can happen when nobody is waiting.
Two or three unattended positions work well: a locked wide that never stops running, a locked mid on the main acting area, and if there is a third, something with a longer lens on the area where faces matter. None of them needs an operator. All of them need to start before the house opens and keep running through the interval, because a camera stopped to save storage is a camera that missed something.
The two things that ruin a small shoot
Synchronisation. Separate recorders drift, and aligning three hours of material by eye is miserable work. Feeding common timecode solves it properly; where that is impossible, a single sharp visual and audible event at the top of each act, in view of every camera, does most of the job for nothing.
Matching. Cameras set up independently will differ in exposure, colour and gamma, and cutting between them advertises the difference more visibly than any framing error. Setting all of them from the same reference, under the show's own lighting rather than under worklight, is ten minutes that saves a grade.
Where the cameras are allowed to be
Every position has to be agreed with the people who own the sightlines. A camera in the stalls blocks seats that were sold; a camera in an aisle is a means of escape obstructed; a tripod at the front of the circle is in the eyeline of the row behind it. The house manager has a legitimate veto and generally exercises it late if consulted late.
Operators are visible too. A person moving in a lit auditorium is more distracting than the equipment, which is why unattended locked positions solve a social problem as well as a staffing one, and why the performances chosen for capture are often the ones where a few seats can be held back from sale.
The safety shot
Whatever else is happening, one camera should be wide, locked, and never touched. It is the shot that covers a missed cue, a fluffed cut, a card change and an operator's cough. Every experienced crew has one, and every crew that does not have one has a story about the night they needed it.
What we cannot verify
Nothing in this article is measurable, and none of it is standardised. Practice varies enormously between opera houses, dance companies, studio theatres and schools, and what is described here is common practice as reported in the field rather than a rule. Claims about how many cameras a given crew size can manage depend on the show, the rehearsal available and the people, and we have no basis for a number.
The short version
- A cut decides where the viewer looks, and the stage director had already decided.
- A wide shot can be cropped later; a close shot cannot be widened.
- Cutting away from simultaneous action records the words and loses the play.
- Multicamera is a second production with its own script and rehearsal.
- Record every camera in full and cut afterwards; edit time is cheaper than live nerve.
- Common timecode and matched cameras prevent the two failures that ruin small shoots.