How to build a paper edit from documentary rushes

The first documentary I cut properly, I did what I think everyone does. I dropped three hours of interview onto a timeline and started chopping.

Two days later I had a sequence that was technically fine and said nothing. Every good line was in it. There was no story.

The thing I was missing has a name, and it is older than any of the software. A paper edit is where you work out what the film is before you touch a timeline, using the words rather than the pictures.

What a paper edit actually is

A written list of the lines you are going to use, in the order you are going to use them, with a timecode against each one.

That is it. No transitions, no music, no thinking about pictures. Just the spine of what gets said and in what order.

It is called a paper edit because it used to be done on paper, with a printed transcript and a highlighter. Plenty of very good editors still do it exactly that way.

Why it works, and why skipping it does not

Cutting straight to timeline feels faster because you are producing something within minutes. The problem is that you are making structural decisions while you are also making frame-level ones, and the frame-level work is seductive. You spend forty minutes getting one transition right in a sequence that is going to be thrown away.

The other thing that happens is that you lose the overview. Once footage is on a timeline you can only see about ninety seconds of it at a time. On paper you can see the whole film on two sides.

And it is far cheaper to be wrong. Reordering a paper edit takes seconds. Reordering a timeline takes an afternoon and usually breaks the audio.

The process

One. Read the transcript before you watch anything.

Not skim. Read. You are looking for what the person actually said, separate from how well they said it, and picture gets in the way of that. A line delivered flatly can carry the film. A beautifully delivered line can be empty.

I read it once through without marking anything. The second pass is where I start pulling.

Two. Mark selects, and grade them.

I use three grades. A is a line the film cannot do without. B is a line I would use if it fits. C is useful information delivered badly, the sort of thing you might cover with pictures or replace with a caption.

Grading matters more than marking. Everyone can spot a good line. What is hard later is deciding which of forty good lines survive when you have eight minutes. If you graded as you went, that decision is already half made.

Be harder on A than feels comfortable. If a quarter of your selects are A grade, you have not really graded anything.

Three. Find the themes before you find the order.

Group the selects by what they are about, not by when they were said. Contributors almost never answer in a usable order. They circle back, they answer the question before last, they say the best version of something twenty minutes after you moved on.

This is the step that most obviously cannot be done on a timeline, and it is where the film usually appears.

Four. Order it.

Now build the sequence, on paper. What does the viewer need to know first. Where does the question get asked that the rest of the film answers. What is the last thing said.

I write it as a numbered list with the timecode and the first few words of each line. That is enough to find it again and short enough to see the whole shape.

Five. Read it out loud.

Sounds daft. It is the single most useful check I know. Structural problems that are invisible on a page are obvious the moment you hear them in sequence, because you can hear where the interest drops.

How long it takes

Roughly an hour of work per hour of rushes, once you are used to it. Less if the interview was well conducted, considerably more if it was not.

That feels like a lot until you compare it to the alternative, which is two days on a timeline and then starting again.

What to hand an editor

If someone else is cutting it, the paper edit is the handover. Sent properly it should contain:

The ordered list of selects, with timecode and the transcript text for each.

The grades, so the editor knows what is load-bearing and what is optional.

The lines you rejected and why, at least for the near misses. This saves the editor re-watching three hours to check whether you missed something.

Anything you know about the rushes that is not in the picture. Which take is the good one, where the contributor got upset, which section is unusable because of a passing aeroplane.

And the timecode has to be right, which is where part one comes back in. A paper edit with times against it is only useful if those times point at real frames. If your timecode was drifting, everything in the list is a guess.

The bit worth saying out loud

A paper edit is a plan, not a contract. The film changes when you see it moving. Lines that read beautifully die on screen, and a line you graded C turns out to hold the whole middle section together because of the way she pauses before it.

The point is not to lock the film in a document. It is to arrive at the timeline already knowing what you are trying to make, so the time you spend there is spent on how it plays rather than on what it is.

Common questions

What is a paper edit in documentary?

A written list of the interview lines you intend to use, in order, with timecode against each one. It is built from the transcript before any editing happens, and it defines the structure of the film in words rather than pictures.

How long does a paper edit take?

Roughly one hour of work per hour of rushes once you have a routine, though a rambling or badly conducted interview can take considerably longer.

Do you still need a paper edit if you are cutting it yourself?

Yes, and arguably more so. Editing your own rushes means you remember the shoot rather than seeing the material, and the paper edit is what forces you to judge what is actually there.

What is the difference between a paper edit and a transcript?

A transcript is everything that was said. A paper edit is a selection from it, graded and reordered into the shape of the film.

Should selects be graded?

It helps enormously. Grading as you go means the later decision about what survives a duration cut is already mostly made, rather than being made under time pressure at the end.

Earwig.io has its own paper edit that lets you select frame accurate words for your editor, so they can pull it straight into their editing timeline. Try it here: https://earwig.io/sync