Log interviews and build a factual story

Turn transcripts, interviews, observation, B-roll, and verified facts into a paper edit and radio cut that preserve meaning and reveal change.

By the end of this module

  • Create searchable transcripts and selects.
  • Build a paper edit around claims and evidence.
  • Assemble a coherent interview-led radio cut before adding coverage.
Six-stage workflow diagram for Log interviews and build a factual story.
Use this workflow to keep technical operations connected to editorial or musical decisions.

Core knowledge

Principles before controls

Factual editing has an evidence duty

Compression and reordering must not change what a contributor means or imply unsupported causality.

A transcript is an index, not truth

Automated text must be checked against audio, names, terminology, accents, and incomplete phrases.

The radio cut exposes story

Without B-roll, graphics, or music, the sequence must still establish context, development, evidence, and resolution.

B-roll should prove or locate

Coverage can show action, environment, process, reaction, or transition. Decorative images should not imply evidence they do not contain.

Guided build / approximately 90-120 minutes

Build it in the application

  1. 1

    Watch every interview and observational sequence at least once before constructing the story.

  2. 2

    Generate or import transcripts, correct names and critical terms, and mark uncertain text rather than silently guessing.

  3. 3

    Create selects sequences by topic, claim, emotion, process, visual action, and potential transition.

  4. 4

    Write a one-page paper edit listing audience question, verified facts, contributor roles, strongest evidence, exclusions, and intended ending.

  5. 5

    Build a radio cut using only interview and essential natural sound, preserving complete thought and honest context.

  6. 6

    Remove repetition while keeping qualifiers that materially affect a claim.

  7. 7

    Create a fact-check marker and source-note convention for every externally verifiable statement.

  8. 8

    Screen the radio cut with transcript visible and test whether reordered excerpts alter implied chronology or causality.

  9. 9

    Save STORY_v01 and a separate cut log before visual coverage begins.

Independent challenge / 45-60 minutes

Prove the principle

Create two 90-second radio cuts from the same interviews: process-led and character-led. Identify factual risks introduced by each structure.

Portfolio checkpoint

Evidence to retain

Submit corrected transcripts, selects sequences, paper edit, radio cut, fact-source notes, and ethical edit log.

Store this evidence in 01_ADMIN/checkpoints/module-2 and include it in the final decision log.

Failure modes

Diagnose before retrying

Text-based edit creates unnatural joins

Return to source audio, restore breaths and context, and treat transcript edits as provisional selections.

The contributor appears to answer a different question

Check preceding prompt, complete answer, chronology, and whether the new context changes meaning.

The story has facts but no progression

Define what the audience understands at the start, midpoint, and end, then reorder evidence around that change.

Working vocabulary

Terms you must be able to explain

paper edit
A written sequence plan derived from transcripts, facts, and available evidence.
radio cut
A story assembly built primarily from speech and essential sound before visual coverage.
selects sequence
An organized timeline of candidate material grouped for editorial use.
fact-check marker
A visible timeline note connecting a claim to verification status and source.

Primary sources

Verify current behavior

Lesson reviewed 2026-07-13. Verify menus and shortcuts against the lab release.

Translate this page

Machine translation provided by Google Translate, on Google’s servers. We do not check these translations and they will get technical terms wrong. The English page is the authoritative one. Following a link sends this page’s address to Google. Your browser may also offer to translate this page itself, which keeps the request on your device.