Six things a cut has to satisfy, in order · Editing Video

A ranking, not a checklist

Walter Murch, who edited Apocalypse Now and The Conversation, wrote a short book called In the Blink of an Eye in which he set out six criteria a cut can satisfy, and — more usefully — ranked them by how much they matter. The ranking is the valuable part, because cuts that satisfy all six are rare and the question in practice is always which ones to give up.

His weighting, roughly:

Emotion — does the cut serve what the audience should feel? About half the value.
Story — does it advance the narrative? A quarter.
Rhythm — does it come at an interesting and right moment? A fifth or so.
Eye-trace — does it respect where the audience is looking? A few per cent.
Two-dimensional plane of screen — does it respect screen direction and the line? A few per cent.
Three-dimensional space — is the physical geography continuous? Barely one per cent.

The numbers are Murch's own rough weighting and are not measurements. What they encode is a claim you can test against your own reactions: audiences forgive a spatial discontinuity readily and forgive an emotional wrong note almost never.

Why the bottom of the list is at the bottom

This is the part that surprises people who have been taught continuity as a set of rules.

If a cut is emotionally right but the actor's glass has moved between shots, essentially nobody notices, because the audience is busy feeling something. If a cut is technically flawless — perfect eyeline, perfect action match, perfect geography — but lands on the wrong beat emotionally, the scene dies and nobody can say why.

This is not permission to be sloppy. Continuity errors cost nothing individually and accumulate: enough of them and the piece reads as careless, which is itself an emotional signal. But when you have to choose — and in unscripted work you choose constantly — the ranking tells you which way.

Using it as a diagnostic

The practical value is in reverse. When a cut feels wrong and you cannot name why, walk the list from the bottom up, because the lower ones are the ones you can see:

Space: has the geography flipped? Is somebody suddenly on the wrong side of the room?
Screen plane: did the subject jump across the frame? Did screen direction reverse?
Eye-trace: where was my eye at the last frame of the outgoing shot, and is there something there in the first frame of the incoming one?
Rhythm: is this cut arriving at the same interval as the last four? Is it landing on the speech or between phrases?
Story: does the incoming shot tell me something new?
Emotion: am I being asked to feel the right thing at this moment, or am I being moved on before I have finished?

Four out of five times the answer is in the top two and you were looking at the bottom two.

Eye-trace, in practice

This is the one most worth learning to see, because it is invisible until somebody names it and then obvious forever.

At the last frame before a cut, your eye is somewhere specific — on a face, a hand, a moving object. At the first frame after, it has to find the new subject. If the new subject is roughly where the old one was, the transition is free. If it is on the other side of the frame, you spend a fraction of a second searching.

So: note where the attention is at the out point, and choose an in point that puts something worth looking at near the same place. This is what editors mean by "matching the frame", and it is why a cut between two wide shots of the same room often works better than a cut between a close-up on the left and a close-up on the right.

You can also use it deliberately. Deliberately putting the new subject somewhere else produces a small jolt, which is exactly what you want at a scene change or a reveal.

Rhythm is the one you can measure

Emotion and story are judgements. Rhythm has a number attached, and the next lesson is about it.

The honest limits of the framework

Murch was describing narrative fiction cut for a cinema audience watching in the dark for two hours. Three adjustments for the work most people are actually doing:

Short-form compresses everything. A fifteen-second vertical video has no room for emotional build, so story and rhythm move up the list and emotion is carried almost entirely by the performance rather than by the cutting.
Documentary and interview work often cannot satisfy the bottom four at all. One camera, one angle, no coverage. The whole reason jump cuts became acceptable is that the bottom of the list had to be abandoned wholesale.
Tutorials and explainers invert it. Comprehension is the emotion. A cut that arrives before the viewer has understood the previous point is the equivalent of a wrong emotional beat.

The ranking is a tool for making a choice, not a law. What is worth taking from it permanently is the habit of asking what the cut is for before asking whether it is clean.

Today

Take a cut in your own work that you know is not right. Walk the six, bottom to top, and write down where it fails. The naming is most of the fix.