Independent visual research / Public editionReviewed 04 Sep 2026
Motion fieldnotesOpenAI film study
About this study

Presenter demo / 01:30 / Published 2025-03-25

Detailed Directions with 4o Image Generation

A long list of requested objects is enlarged before a single illustrated grid, with presenter and filmed-laptop inserts around the result.

Watch official film ↗2 screen examplesAll 100 films
Representative frame from Detailed Directions with 4o Image Generation
OpenAI · representative sample 00:56.000 · Source moment ↗

Useful to borrow

Give a complex input a readable close-up, then hold one output long enough to compare it.

Do not infer

Do not infer that every constraint was satisfied without checking each one; the menus shown are historical interface states.

7 shot groups.

Read across from the observed content to its treatment and possible adaptation. Enlarge the frame for detail.

Approximate study intervals; horizontal scrolling reveals all columns on small screens.

Moment / sourceReference frameWhat is on screenFraming / editUseful adaptation
Shot group 01
≈ 00:00–00:12
Title / identityWatch from here ↗
Sample 00:08.000
Directions topic

A couch setup leads into the detailed-directions title.

Topic typography over an established human setting.State the constraint being explored.
Shot group 02
≈ 00:12–00:32
App / browser screenWatch from here ↗
Sample 00:20.000
Setup menu

Presenter and laptop inserts lead to a visible model menu.

Human context with a brief interface insert.Keep setup secondary to the actual example.
Shot group 03
≈ 00:32–00:48
App / browser screenWatch from here ↗
Sample 00:40.000
Constraint list

The composer displays a long numbered set of requested shapes, objects, and text.

Extreme input crop.Make the complexity of the request visible.
Shot group 04
≈ 00:48–00:56
People / environmentWatch from here ↗
Sample 00:52.000
Presenter bridge

The film returns to the seated presenter.

Human pause before the output.Treat this as an editorial bridge, not proof of processing time.
Shot group 05
≈ 00:56–01:00
Isolated outputWatch from here ↗
Sample 00:56.000
Illustrated grid

A grid presents many requested objects together.

Single isolated output.Allow a direct input-to-output comparison.
Shot group 06
≈ 01:00–01:28
People / environmentWatch from here ↗
Sample 01:08.000
Discussion and laptop

Presenter shots include a filmed laptop showing another result at smaller scale.

Human-led discussion with physical-screen context.Avoid counting unreadable background content as a detailed UI workflow.
Shot group 07
≈ 01:28–01:30
Title / identityWatch from here ↗
Sample 01:28.000
End mark

A simple brand ending closes the film.

White closing card.Keep the outro short.

Reference frame

Enlarged reference frame

Watch this moment on the official channel ↗