Question What is the smallest proof of concept that proves a short is shootable?

[Miranda_]

I’m developing a visually ambitious dystopian short for a small team. Before committing to the full approach, what would you test first: one difficult shot, a complete dialogue scene, the main location, or a short sequence combining image, sound and editing?

I’m looking for the smallest test that exposes weaknesses—not just one attractive frame. What did your proof of concept reveal that storyboards or planning did not?

 

[Cinema Doktor]

I’m developing a visually ambitious dystopian short for a small team. Before committing to the full approach, what would you test first: one difficult shot, a complete dialogue scene, the main location, or a short sequence combining image, sound and editing?

I’m looking for the smallest test that exposes weaknesses—not just one attractive frame. What did your proof of concept reveal that storyboards or planning did not?

That's a great question. In my opinion, the most valuable proof of concept isn't necessarily the one with the most impressive shot it's the one that reveals the biggest practical challenges before production begins.

If I had to choose, I'd probably test a short sequence that combines cinematography, sound, editing, and performance. It's often the interaction between those elements that reveals issues you simply can't predict from storyboards or planning alone.

I'm curious to hear what experienced directors and cinematographers here think. Have you ever had a proof of concept completely change your approach before production?

 

[Miranda_]

That makes sense. A short sequence seems like the best stress test because weak transitions show up quickly. I’d keep it to 30–60 seconds and include one performance beat, one difficult lighting setup, and finished sound. What would you cut first if that test already felt too expensive?

 

[Adam Films]

That makes sense. A short sequence seems like the best stress test because weak transitions show up quickly. I’d keep it to 30–60 seconds and include one performance beat, one difficult lighting setup, and finished sound. What would you cut first if that test already felt too expensive?

I'd probably cut complexity before ambition. Keep the one element that's hardest to fake whether that's a performance, a lighting challenge, or a practical effect and simplify everything else. Fewer locations, fewer setups, and fewer variables make it much easier to identify what's actually causing problems.

One thing I've learned is that if a proof of concept only works under perfect conditions, it isn't really proving the production is ready. The best test is one your team could realistically repeat on the actual shoot.

 

[Miranda_]

That repeatability test is the useful distinction. I’ll treat the proof as a rehearsal: same crew, same time window, same gear, and one deliberately difficult element. If it only works with extra favors or extra hours, it has exposed the problem rather than proved the approach.

 

[Max]

That repeatability test is the useful distinction. I’ll treat the proof as a rehearsal: same crew, same time window, same gear, and one deliberately difficult element. If it only works with extra favors or extra hours, it has exposed the problem rather than proved the approach.

I like your idea of treating the proof of concept as a rehearsal. One thing I'd add is to test how quickly the scene can be reset.

I'd shoot the same 20–30 seconds several times in a row, just as you would on the actual production. That's often when the hidden problems appear. Continuity starts drifting, lighting becomes harder to reproduce, props take longer to reset, and even performances can become less consistent.

In my experience, a proof of concept shouldn't prove that you can create one great take. It should prove that you can achieve the same result repeatedly, within the time and resources you'll actually have on the shooting day.

If you can only make it work once, you've proved the shot not the production.

 

[Lucas]

That repeatability test is the useful distinction. I’ll treat the proof as a rehearsal: same crew, same time window, same gear, and one deliberately difficult element. If it only works with extra favors or extra hours, it has exposed the problem rather than proved the approach.

One thing I’d add is an audience test. Show the proof of concept to a few people who know nothing about the project, without explaining the story or what they should notice.

With a dystopian film, the visuals can work beautifully while the world or mood you’re trying to create doesn’t communicate as clearly as you expected.

So perhaps the proof should test two things: can we shoot it, and does the audience receive what we think we’re communicating?

I’d be curious whether anyone here has changed a scene after a proof of concept worked technically, but didn’t work for the audience.