Four of us made a twelve minute film over three weekends. The budget was whatever we spent on petrol, food and one extension cable. It is not a great film but it is a finished one, and the process taught me things that working as a runner on properly funded shoots had not.
The crew size determines the script
This is the part people get backwards. They write the script and then work out how to shoot it with what they have.
With four people, two of whom are also acting, you have one person on camera and one person doing everything else. That means no complicated lighting setups, no shots requiring a focus puller, no scenes with more than three people in them, and no locations requiring permission or crowd control.
We wrote to that. Two main characters, two locations, one supporting part played by a friend who turned up for an afternoon. Every scene was designed to be shootable by one person operating and one person managing sound and everything else.
The writing got better because of it. Constraints on what you can stage push you toward what people say to each other, which is where most of a short film lives anyway.
Sound was the only thing we spent real effort on
We borrowed a shotgun microphone and a recorder, and one person's entire job during any dialogue scene was holding a boom and watching levels.
Everything about the picture was compromised. We shot on a camera that was several years old, with available light plus one panel, in rooms we could not control. The image looks like what it is.
The sound is clean. And in the finished thing, people notice the sound quality and do not particularly notice the image quality, which is a lesson I have applied to everything since.
A slightly underexposed shot with clean dialogue reads as a stylistic choice. A well-lit shot with echoing, distant dialogue reads as amateur. I do not fully understand why the asymmetry is that strong but it is consistent.
Locations are the whole logistics problem
We used a flat belonging to one of us and a stretch of canal path.
The flat was straightforward except for noise. We discovered on the first day that the road outside was audible in every take, and that traffic was quieter between roughly seven and nine in the morning. So we shot the flat scenes early, which nobody enjoyed but which saved the audio.
The canal path was harder. Public space means people walking through, dogs, cyclists, and a boat going past at exactly the wrong moment on the best take. We learned to just keep rolling and go again, and to shoot the same dialogue several times so the edit had options.
The lesson is that a free location has costs, and the costs are in time and in takes rather than in money. Three weekends for twelve minutes is slow, and a lot of that slowness was waiting for a location to be usable.
Continuity falls apart without someone watching it
Nobody was assigned to continuity because there was nobody spare. We paid for that in the edit.
A glass that moves between shots. A jacket that is on and then off. A scene shot across two weekends where the weather is visibly different through a window.
Most of it we edited around, which cost us the shots we wanted to use. One sequence I had to restructure entirely because the coverage did not cut together.
If I did it again, someone would take photographs at the end of every setup. That is a five second job that would have saved several hours later.
The performances are the thing you can actually control
With no money, no gear and no crew, the one variable that is entirely available is how good the performances are.
We rehearsed. Not extensively, but we read the thing through several times before shooting and talked about what each scene was doing. Two of us were not actors in any real sense and the rehearsal is the only reason the scenes work at all.
Time spent rehearsing is free. Time spent shooting is expensive in every sense — daylight, patience, goodwill, the availability of the person whose flat you are in. Shifting effort from the shoot to the preparation was the single best decision we made.
What actually went wrong
The ending. We wrote an ending, shot it, and it did not work, and by the time we knew that we could not reassemble everyone.
The version in the finished film is built from a reaction shot and a wide we had for other reasons. It is not what was written and most people who have seen it think it was intentional, which is either encouraging or depressing depending on your mood.
The general lesson is to shoot more than you think you need of anything that carries weight. Extra takes cost nothing on a shoot with no crew rate. Not having the shot costs everything.
Would I recommend it
Yes, with one caveat. Finish it. There are an enormous number of half-shot short films sitting on hard drives, and the value of the exercise is almost entirely in the completion.
The edit taught me more than the shoot. Watching your own coverage and discovering what you failed to get is an education you cannot receive any other way, and you only get it if you sit down and actually cut the thing.
Ours is twelve minutes long and slightly too slow in the middle and I am glad it exists.