How to Build a First Assembly (Dialogue-First First Cut)

A first cut is the first complete attempt to turn selected material into a coherent sequence.

It doesn't need to be polished.

It doesn't need perfect sound.

It doesn't need final B-roll.

It needs to answer one question:

Does the basic edit work?

That means the first cut should be complete enough to reveal structural problems.

The biggest time-saving decision often happens before opening the sequence.

Before cutting, know:

• what the story is about
• which interviews matter
• what the major themes are
• which quotes are strongest
• what the audience needs to understand
• where the major story beats might occur

You don't need to know the final edit.

You need enough understanding to avoid building blindly.

For interview-driven projects, dialogue is often the structural backbone.

Review the interviews first.

Identify:

• essential information
• strong character moments
• emotional beats
• contradictions
• turning points
• memorable soundbites

Transcripts can dramatically accelerate this process because they make long conversations searchable.

But don't choose material based only on keywords.

Context matters.

Don't organize your selections only by interview.

Organize them by what they do in the story.

For example:

• setup
• context
• conflict
• escalation
• revelation
• consequence
• resolution

This makes the first assembly much faster because you're no longer searching through every interview for every section.

You're working from a map.

Before assembling individual cuts, establish the broad progression.

For example:

The exact structure will vary.

The important part is knowing what the audience should experience as the story progresses.

Interview subjects often repeat themselves.

Don't automatically include every useful answer.

Compare versions based on:

• clarity
• emotion
• specificity
• performance
• narrative purpose

The goal isn't to collect the most material.

It's to select the material that carries the most story.

For an interview-heavy documentary, dialogue often provides the initial structure.

Build the core sequence around the strongest statements.

Don't worry about making it visually beautiful yet.

You want to know:

Does the conversation work?

If the dialogue doesn't work without heavy visual support, adding B-roll won't necessarily solve the underlying problem.

Once the dialogue structure is working, introduce visuals.

Use B-roll to:

• establish context
• reveal character
• illustrate information
• create contrast
• cover necessary cuts
• change pacing

Don't use every beautiful shot simply because you have it.

The first cut benefits more from meaningful visuals than from visual density.

This is one of the most important time-saving rules.

Finish the first cut.

Don't spend an hour perfecting the opening before you've built the ending.

You need to see the entire story before you can accurately judge the beginning.

The ending can change the meaning of the opening.

The middle can change what the audience needs to know earlier.

A complete first cut reveals those relationships.

Avoid spending significant time on:

• perfect transitions
• detailed sound design
• final color correction
• polished graphics
• tiny timing adjustments
• perfect B-roll coverage

Those are later-stage decisions.

The first cut should prioritize structure and clarity.

The most valuable thing about a first cut isn't that it looks good.

It's that it exposes problems.

You can finally see:

• where the story slows down
• where information repeats
• where context is missing
• where an interview isn't contributing
• where the emotional arc breaks
• where the ending feels weak

You can't diagnose those problems properly while looking at isolated clips.

You need the whole sequence.

The first cut should optimize for learning, not polish.

Every decision should help answer:

"What does this story need?"

Not:

"How can I make this section look finished?"

That's what makes the first cut useful.

The first cut becomes expensive when every decision is made at once.

Story structure.

Dialogue.

B-roll.

Music.

Sound.

Transitions.

Graphics.

Trying to solve everything simultaneously creates constant rework.

A better workflow separates the decisions into layers.

Start with the narrative backbone.

Ask:

• What starts the story?
• What changes?
• Where does the conflict escalate?
• What is the turning point?
• What does the audience need to understand at the end?

If the structure doesn't work, no amount of polish will fix it.

Solve the biggest problem first.

For interview-driven projects, begin with the strongest dialogue.

Don't worry about making every transition seamless yet.

Focus on whether the sequence communicates the intended idea.

At this stage, you're testing:

• information
• character
• conflict
• progression
• emotion

If the dialogue works, you have a foundation.

Repetition is one of the biggest sources of unnecessary runtime.

The same idea may appear:

• several times in one interview
• across multiple interviews
• through both dialogue and B-roll

During the first cut, identify which version does the job best.

Keep it.

Remove the rest.

Don't try to fully cover the sequence immediately.

Start with B-roll that solves a specific problem.

For example:

• establishing where we are
• showing what the interviewee is describing
• creating a transition
• covering an unavoidable jump
• changing the emotional rhythm

This keeps the first cut flexible.

You can always add more later.

If you don't have the perfect shot yet, keep moving.

Use a temporary clip.

Leave a marker.

Write a note.

Move on.

Stopping the entire edit to find the perfect B-roll shot can destroy momentum.

The first cut is about understanding the whole sequence.

Don't spend most of the day perfecting the first five minutes.

Build the entire piece.

Even if later sections are rough.

A complete first cut gives you information that an unfinished one can't.

You can evaluate:

• overall pacing
• story progression
• emotional balance
• runtime
• repetition
• ending strength

Once the first cut is complete, watch it from beginning to end.

Don't start trimming while watching.

Take notes.

Separate problems into categories:

Story

• Is the narrative clear?

Structure

• Does each section appear in the right place?

Dialogue

• Are the strongest statements being used?

Pacing

• Does the story move?

Visuals

• Does the image support the dialogue?

This turns a vague feeling of "something isn't working" into actionable revisions.

Not every problem deserves immediate attention.

A useful hierarchy is:

Always start at the top.

There's little value in perfecting a transition inside a section that may eventually be removed.

Editors frequently encounter predictable problems during the first cut.

The first cut isn't the final product.

Its purpose is to reveal whether the edit works.

Constantly stopping to fix problems makes it difficult to experience the film as a whole.

Watch first.

Revise second.

Good enough is often enough for the first cut.

Mark missing shots and keep moving.

The amount of effort required to find a clip doesn't determine its value.

If it doesn't serve the story, remove it.

If you're constantly moving the same section around, step back.

You may have a structural problem rather than a timing problem.

A first cut is ready when the audience can experience the complete story—even if many elements are still rough.

You should be able to identify:

• the central idea
• the major story beats
• the key interviews
• the emotional progression
• the ending

And most importantly:

You should be able to watch it from beginning to end without needing to explain what is supposed to happen.

That's the point at which feedback becomes useful.

The best first cut doesn't answer every question.

It answers the most important ones.

• Does the story work?
• What's missing?
• What's unnecessary?
• What's in the wrong place?
• What deserves more attention?

Once those questions are visible, the next revision becomes much more focused.

And that's where the real time savings come from.

The traditional first-cut process can take hours before the story is even visible.

• Watch the interviews.
• Log the material.
• Search transcripts.
• Compare answers.
• Build stringouts.
• Find B-roll.
• Assemble the sequence.

Then discover that half of it needs to change.

AI can shorten parts of that process by helping editors understand and organize the material before they commit to the timeline.

This sounds contradictory, but it's one of the most important ideas in modern editing.

The first cut doesn't really begin when you place the first clip on the timeline.

It begins when you start deciding:

• what matters
• what repeats
• what connects
• what changes
• what should be left out

AI can help surface those patterns across large amounts of interview material.

The editor can then make decisions with more context.

Imagine having 15 hours of interviews.

A traditional workflow might require manually searching each conversation for relevant material.

An AI-assisted workflow can help surface:

• relevant passages
• recurring themes
• similar answers
• contrasting perspectives
• emotional moments
• potential soundbites

The editor still chooses what matters.

But the search space becomes much smaller.

The workflow can become:

The important part is what happens before the timeline.

The more uncertainty you remove beforehand, the less likely the first cut is to become an expensive experiment.

One of the biggest sources of wasted time is rebuilding material that was assembled without enough context.

For example:

• You build a five-minute section.
• Then realize two interviews repeat the same information.
• You search for another quote.
• You restructure the section.
• Then discover that the same idea appears again later.

The problem wasn't the editing speed.

The problem was insufficient understanding of the material before assembly.

AI can help expose those relationships earlier.

There is an important limit.

AI can identify that several interviews discuss the same theme.

It can surface candidate quotes.

It can suggest relationships.

But it shouldn't automatically determine:

• which character matters most
• which perspective should dominate
• what the documentary is trying to say
• where a moment should emotionally land
• what the audience should believe or feel

Those decisions belong to the editor.

The value of AI is reducing the mechanical work surrounding those decisions.

The goal isn't to generate something that looks final.

It's to generate something that is structurally useful.

A good first cut should make it easier to answer:

• What's working?
• What's missing?
• What's repetitive?
• What's in the wrong place?
• What's emotionally weak?
• What deserves another pass?

That's much more valuable than simply producing a polished-looking timeline quickly.

This changes how editors can think about the first cut.

Instead of:

"Let's see what happens if we put these clips together."

the workflow becomes:

"We've identified the strongest material and a possible structure. Now let's test whether it works."

The timeline becomes an experiment with a hypothesis.

Not a search through thousands of clips.

Even in an AI-assisted workflow, editors remain responsible for the most important decisions.

They determine:

Meaning — What does this moment actually communicate?

Context — What does the audience need to know before hearing it?

Emotion — What should the audience feel here?

Rhythm — When should the moment arrive and when should it end?

Point of view — Whose perspective should shape the story?

AI can help organize the material.

The editor gives it meaning.

The most effective workflow isn't simply faster.

It's more iterative.

AI can accelerate the first two stages.

The editor uses the rough cut to validate the decisions.

Then the process repeats.

That's how time is actually saved.

Not by eliminating revision.

By making each revision more informed.

The first cut workflow that actually saves time isn't the one that gets a timeline together in record time.

It's the one that minimizes unnecessary rebuilding.

Professional editors save time by solving the biggest questions first:

• What is the story?
• Which material matters?
• Which interviews belong together?
• What repeats?
• Where does the narrative change?

AI can help answer some of those questions faster by analyzing transcripts, comparing interviews, surfacing themes, and identifying potential soundbites.

But the editor still decides what the documentary means.

That's the important distinction.

AI can reduce the search. The editor makes the story.

And when the first cut starts from better decisions, the entire editing process becomes more efficient—from the first assembly all the way to the fine cut.

Supacut helps editors reduce the work between raw interview footage and the first cut. It analyzes transcripts, connects related ideas across interviews, surfaces recurring themes and strong soundbites, and turns those discoveries into a story-first assembly that can be refined in Premiere Pro, DaVinci Resolve, or Final Cut Pro.

Instead of spending hours searching for what might work, editors can start with a clearer understanding of what the material contains.

Because the fastest first cut isn't the one you build fastest.

It's the one you don't have to rebuild.

Get Supacut