Every edited video climbs the same ladder - dailies, selects, assembly, rough cut, fine cut, picture lock, conform, delivery - and the ladder matters because the useful feedback changes at every rung. Notes given at the wrong stage are wasted twice: once by the person who wrote them, and once by the editor who has to interpret them against footage that cannot do what they ask yet.
The ladder at a glance
| Stage | What exists | Who reviews | Useful feedback |
|---|---|---|---|
| Dailies | Raw footage from each shoot day | Director, DP, producer | Coverage gaps, performance, reshoot flags |
| Selects | Chosen takes, stringout reels | Editor, director | Best-take calls, missing material |
| Assembly | Everything in script order | Editor (mostly solo) | Structural: what is obviously missing or redundant |
| Rough cut | A watchable cut, still loose | Director, producers, first stakeholders | Story, structure, pacing, tone |
| Fine cut | Tightened cut, temp sound/graphics | All stakeholders, brand/legal | Specific notes at exact moments |
| Picture lock | Timing frozen | Decision owner: explicit approval | Approval - not new structural notes |
| Conform, color, mix | Finishing on the locked cut | Finishing team, final QA | Technical: exposure, mix, caption accuracy |
| Delivery | Masters, cutdowns, screeners | Client, platforms, festivals | Spec compliance, explicit sign-off |
Dailies and selects: before the edit
Dailies are the day's raw footage, reviewed the same night it is shot - reshoots are only cheap while the crew is still standing in the location. The review-side mechanics (secure links from set, watermarked access) are in what is dailies review. Selects are what survives: the takes worth building from, cut into stringout reels. A selects reel is a working tool, not a deliverable - the disambiguation from other reels is in what is a selects reel.
Assembly: the spine
The assembly cut puts every scripted (or planned) beat in order, in full, with no regard for pace - if the script says the scene exists, the assembly contains it, even when it is ugly, placeholder-graphicked, and twice as long as it will ever be. The assembly is where structural problems announce themselves: a missing beat, a redundant section, an argument that never lands. It is reviewed mostly by the editor and the closest collaborators, because feedback at this stage should be structural, and structural feedback from ten people is a committee.
Rough cut: the first real audience
The rough cut is the first version you show people who did not make it. It is watchable end to end, but still loose: temp music, temp graphics, uneven audio, sequences that breathe too long. Feedback here is story-level - structure, pacing, tone, whether the argument lands - and it is the cheapest place for stakeholders to give big notes, because nothing is polished yet to fall in love with and nothing is locked to protect. The definition and the edit-ladder context are in what is a rough cut.
Fine cut: precision notes begin
The fine cut is where the ladder turns from structural to surgical: sequences hold, timing tightens, temp elements become finals. This is the stage where frame-accurate comments and drawn annotations start paying for themselves - "tighten this transition" at the rough cut becomes "the cut lands one frame late at 0:42" at the fine cut, and the second note is fixable in seconds. It is also where brand, legal, and compliance reviewers should take their pass: they are checking facts, claims, rights, and representation on a cut that is close to final, early enough that changes are still cheap. Frame-accuracy and timecode-rough commenting are compared in timecoded comments vs frame-accurate annotations.
Picture lock: approval, not opinion
Picture lock freezes the timing. After lock, note-giving changes character entirely: the conversation moves from "should this beat exist" to finishing work that depends on the locked timing - conform, color, sound mix, graphics, captions. New structural notes after lock are not feedback; they are budget events. That is why lock is the explicit-approval moment: someone with authority marks the version approved, in the tool, where the record lives. The mechanics and the copyable policy are in what is picture lock.
Conform, finishing, delivery
After lock, the finishing team rebuilds the cut at full quality (the conform - the handoff between the offline edit and finishing is described in what is an edit decision list), then color, sound, and graphics finish the frame. Feedback at this stage is technical and factual: exposure, mix balance, caption accuracy, spec compliance. Delivery is the last rung - masters, cutdowns, and for films the controlled pre-release copies that go to festivals and press; the security model is in what is a screener.
Match the feedback to the rung
The ladder's practical lesson for anyone running client review: ask for the kind of feedback the stage can act on. Rough-cut rounds collect story notes; fine-cut rounds collect moment-precise and compliance notes; lock collects an approval. Teams that mix the kinds - execs giving structure notes at the fine cut, legal arriving after lock - pay for it in blown schedules and relitigated rounds. The round policy that keeps stakeholders in their rungs is in video review best practices, and the approval chain itself in the approval workflow guide.
FAQ
What comes after a rough cut?
The fine cut: the same structure, tightened - sequences trimmed to length, temp elements replaced or upgraded, timing frame-accurate. After the fine cut comes picture lock, then conform/color/mix, then delivery. The full ladder is the table at the top of this guide.
Is the fine cut the same as the final cut?
No - the fine cut is still pre-lock: timing can still change if a note lands. The "final cut" is everything after lock finished to delivery quality. Teams that treat a fine cut as final skip the lock step and lose the clean boundary between "notes welcome" and "finishing in progress."
How long does each stage take?
It scales with footage and stakes more than runtime: a 60-second ad can compress the whole ladder into a week, while a documentary spends months between dailies and rough cut alone. The durable rule is sequencing, not duration - structural notes before fine cut, precision notes before lock, approvals at lock.
Who should review at each stage?
Narrow early, widen late: dailies and selects belong to the core team, the rough cut to the first real stakeholders, the fine cut to everyone with a sign-off interest (brand, legal included), and lock to the single decision owner. Reviewer lanes keep those audiences from drowning each other - the mechanics are in the approval workflow guide.