Week 8 Workshop: Take selection

Generating and diagnosing several takes of one performance beat, then choosing on the record

19 April 2027

Production walkthrough: See how The Last One carried out this stage in a real production.

Before the workshop#

Bring one performance beat from your film that hasn’t worked cleanly yet.

In the workshop#

Activity 1 — Recap and beat selection (10 min)#

Instructions: State the lecture’s three areas to investigate — note, reference and model/workflow — and distinguish a visible mismatch from its possible cause. Choose the beat you’re bringing today and state its objective and tactic before generating anything.

Expected output: One chosen beat, with its objective and tactic written down.

Activity 2 — Generate three takes (50 min)#

Instructions: Generate at least three takes of your beat. Vary the note between takes only where you have a specific reason to — don’t treat this as three unrelated attempts.

Expected output: At least three takes of the same beat.

Activity 3 — Diagnose the rejected takes (20 min)#

Instructions: For each take you’re rejecting, write a one-line diagnosis stating the observed mismatch, a possible note/reference/model cause with uncertainty, and the next check. An unresolved cause is allowed. Select your remaining take and write a justification against the beat’s stated objective and tactic, not against how the takes compare to each other.

Expected output: A one-line diagnosis per rejected take and a selection justified against the beat’s original intent.

Activity 4 — Group dailies-review (30 min)#

Instructions: In a small group, screen each take before revealing your diagnosis. First ask the group to describe only the visible behaviour. Then show the submitted direction, actual reference and relevant tool settings. Only then discuss possible causes and compare them with your diagnosis. Agree on one next check and state what it would establish. Repeat for every member’s beat; do not force a single causal label when evidence is insufficient.

Expected output: A group-checked observation and provisional diagnosis for at least one rejected take, with uncertainty and a next verification recorded.

Activity 5 — Completion check (10 min)#

Instructions: Walk the two spec bullets together. Do you have at least three takes and a one-line diagnosis for each rejected one? Is your selection justified against the beat’s objective and tactic, not against which take you preferred?

Expected output: Both spec bullets verified, or a named gap and what would close it.

Afterwards#

The diagnosis log you wrote today is what a director’s shot log looks like — keep the habit for every beat from here on.