What Is an Assembly Edit? The First Full Cut, Explained

It is deliberately rough. That is the whole point of it. An editor who expects an assembly to look finished is grading it on the wrong curve, and a producer who panics at its runtime is reacting to something that is supposed to happen at this stage. Here is what an assembly actually contains, where it sits in the process, how long it should run, and where editors waste time trying to make it look better than it needs to.

What it is

A complete pass, not a polished one

An assembly takes every clip an editor has marked as usable and lays it down in the order the story currently suggests, start to finish. Cuts are straight, transitions are none, and music is either absent or a placeholder dropped in to keep the pacing honest. Nothing here is meant to survive untouched. The job of an assembly is to let the whole story be watched for the first time, so a slow open, a sagging middle, or an ending that doesn't land shows up on this watch, not three weeks later after a scene has been color-corrected and scored.

Treat it as a proof of concept, not a draft. A draft implies the shape is roughly right and just needs refining. A proof of concept is testing whether the shape is right at all. Watching an assembly back should feel closer to reading a first outline on paper than watching a nearly-finished video, and most of the frustration editors feel at this stage comes from judging it against a standard it was never trying to meet.

Here's what that first rough pass actually looks like in the room, before any of the polish that gets layered on afterward:

Where it sits

Between selects and a rough cut, not interchangeable with either

Assembly sits right after selects and right before a rough cut. Selects is deciding which material is usable, judged clip by clip. Assembly is placing that material in sequence, judged as a whole story. A rough cut is where trimming, pacing, and scene-level decisions start, judged scene by scene. Skip straight from selects to a rough cut and you're making trim decisions on material whose place in the overall story hasn't been tested yet, which is exactly how editors end up polishing a scene that gets cut entirely in the next pass.

Each stage answers a different question, which is why it needs its own tool and its own pass. Collapse all three into one and you're answering "is this usable," "does the story work," and "does each scene earn its runtime" at the same time, which is how a first cut ends up both too long and structurally unclear simultaneously. Editors who work from a transcript rather than scrubbing raw footage often do their selects and assembly in the same sitting, since a transcript-based editor turns both steps into text decisions before anyone opens an NLE timeline.

How long

Longer than the final cut, and that's expected

An assembly commonly runs around 140 percent of the intended final length, give or take, depending on the format. A 60-minute documentary might assemble at 84 minutes. A piece built from a single hour-long interview can easily run two hours once every usable moment is placed, because nothing has been cut for pace yet, only for basic usability. Treat that extra length as a sign the stage is working, not a problem to panic over.

Format changes the math. A 30-minute broadcast slot with fixed ad breaks leaves little room for an oversized assembly, since the final runtime is essentially locked by the schedule it airs in. A feature documentary or a long-form YouTube video has far more flexibility, and an assembly that blows well past 140 percent of a loose target is common when the footage genuinely supports a longer story than anyone planned going in. Use the 140 percent figure as a sanity check, not a hard ceiling to enforce while you're still building the thing.

What to watch for

Three structural questions, answered on one watch

Watch an assembly for exactly three things: does the opening earn attention in the first minute, where does the middle lose momentum, and does the ending land with the weight it should. These are story-level questions, not sentence-level ones, and an assembly is the only point in the process where you can answer them honestly, because every usable moment is still on the screen to judge against.

Note problems in general terms on this first watch instead of reaching for the timeline to fix them immediately. "The middle third sags around the second interview" is more useful right now than trimming that section on the spot, because the actual fix might be pulling material from elsewhere in the assembly rather than tightening what's already there. Diagnose before you treat. That keeps the watch focused on structure instead of sliding back into scene-level trimming before the bigger questions get answered.

Documentary practice

Why documentary assemblies run especially long, and get rebuilt more than once

Scripted work usually enters the edit with a locked shooting script, so the assembly mostly confirms what was already planned. Documentary work is different. The story often gets discovered inside the footage itself, so assemblies run longer relative to the finished piece and get rebuilt as the narrative becomes clear, sometimes more than once. Documentary editor Dilesh Korya, whose credits include the BBC and Open University co-production Empire with David Olusoga, has been cutting people-based documentaries for the BBC, Channel 4 and ITV since 1998. In a RedShark News interview, he put his approach to this stage plainly: "Keep moving. The only goal here is flow." Korya works sequentially, assembly, then rough cut, then fine cut, on duplicated timelines rather than overwriting the assembly as he goes, so the original structural pass stays sitting there if he needs to return to it.

Walter Murch, who won the Academy Award for Best Film Editing on The English Patient at the 1997 ceremony, wrote about exactly this problem in In the Blink of an Eye. He describes how, right after assembling and watching a first cut back, an editor can still "see himself" in each cut, remembering why every single one was made, which makes it almost impossible to watch the thing as an audience would. His counter-trick is to imagine the screen very large and himself very small, as if sitting in a theater, until the self-awareness fades enough to see the film for what it actually is rather than the sum of individual decisions he already knows by heart.

That gap between what an editor sees and what a stranger will eventually see is exactly why bringing in a second set of eyes on the assembly, a producer, a director, or another editor, is standard practice on documentary projects specifically. Someone who didn't sit through every take of every interview reacts the way a first-time audience eventually will, which is a genuinely different signal than an editor watching their own structural choices play back.

Common mistakes

Where editors burn hours they don't get back

Polishing dialogue timing, adding transitions, or color-correcting a scene during the assembly is the most common mistake, since that work usually gets redone or thrown out once the scene's place in the story shifts in the next pass. Skipping the assembly entirely and jumping straight to a rough cut is the second: it feels faster, but it usually means re-litigating structural problems mid-trim instead of catching them in one clean story-level watch. Treating a long assembly as a problem to fix immediately, rather than expected raw material to shape over the next passes, is the third, and it's the one that causes editors to cut material they'll want back later just to hit an arbitrary runtime early. On tight deadlines there's a fourth, quieter version of this mistake: skipping the full watch and trimming from memory of the footage instead, which reintroduces the exact blind spots the assembly stage exists to catch in the first place.

Getting there faster

Building an assembly from a transcript instead of scrubbing footage

The slowest part of building an assembly by hand usually isn't placing clips, it's finding them: scrubbing hours of raw footage to relocate a moment you remember but can't immediately place. Marking selects on a transcript with word-level timecode removes that search step entirely, since each marked line already carries the exact in and out point needed to drop the clip into a sequence. That turns assembly-building into an ordering exercise instead of a repeated hunt through raw footage, which is most of where the time savings in a transcript studio workflow actually come from at this specific stage.

It also changes who can build the first pass. Ordering marked transcript selections into a rough story sequence doesn't require the same NLE fluency that trimming and finishing does, which is part of why producers and directors, not just editors, often build a first version of this pass in a paper edit tool before handing an approved order to the editor. The editor's hours then go toward craft decisions the assembly can't answer on its own, instead of mechanically relocating footage someone already chose. A tool like ScriptCut works this way: build selects and arrange them against a transcript with locked timecode, watch any selection back to check tone, then hand off a frame-accurate timeline to DaVinci Resolve, Premiere Pro, Final Cut Pro, or Avid for the trim-and-finish work that follows.

The takeaway

Rough, complete, and honest about its own weaknesses

An assembly is supposed to be rough. Its entire value is being complete enough to judge honestly, before anyone spends real time polishing material that structural changes might remove anyway. Build it fast. Expect it to run long. And use the first full watch to find the real problems before the careful work even starts.

Sources

• RedShark News: Cutting With Curiosity, Inside a Documentary Editor's Workflow
• Walter Murch, In the Blink of an Eye, ASC/AC Store
• Academy Awards: 1997 ceremony records (The English Patient)
• PBS: Empire with David Olusoga