Skip to content
Get in touch

Make a video

Explore generated shots, animated stills, existing clips, narrated still sequences, and the finishing work that turns them into a usable video.

Go to My experiment

Explore

What are the ways to do this?

Two mixed workflows from one owned illustration

Create a 12-second invitation with the same source image and three four-second beats: bring one object, explore what it needs, learn together.

Notice: How does a cut between held views feel different from a continuous zoom?

An edited still sequenceA silent 12-second invitation uses three held views of one illustration, with a detail cut at four seconds and return at eight seconds. Turn on captions or read the transcript.

Credit: Renaissance Solutions; source illustration created with OpenAI image generation in Codex. Video assembled locally with FFmpeg. Credit details

An animated stillA silent 12-second invitation uses a gradual centered zoom into the same illustration. The source characters do not move independently.

Credit: Renaissance Solutions; source illustration created with OpenAI image generation in Codex. Video assembled locally with FFmpeg. Credit details

Read or download the transcript and description.

What this shows: Two complete H.264 exports with the same image, duration, frame size, and caption cues.

What it cannot show: Generated-motion quality, lip sync, live-action continuity, sound design, or audience comprehension.

  • Both examples were assembled locally with FFmpeg; no generative video model was tested.
  • The source characters do not move independently.
  • Human audience comprehension and preference remain untested.

Approach map

Open any approach to inspect its starting material, human work, outputs, and sequence.

Generate a short shotCreate a bounded piece of footage from text or an image, then treat it as source material for editing.
Starts with
A shot purpose · A prompt or permitted image · A duration and framing target
Main actions
Generate short candidates · Inspect every frame · Select or repair · Move the shot into a finishing timeline
Outputs
Short clips · Prompt, settings, and source records
Human work
Plan the shot · Reject continuity or identity failures · Finish the larger video
Useful when
A hard-to-film visual is useful · A short atmospheric shot can stand alone
  1. Plan one shotDescribe action, camera, duration, and the job the shot performs.
  2. Generate and inspectReview hands, faces, tools, text, edges, and motion frame by frame.
  3. Finish elsewhereEdit, caption, mix, and export; generation is only acquisition.
Animate a stillAdd camera movement or limited motion to an approved image.
Starts with
An owned or permitted still · A motion purpose · Protected original assets
Main actions
Choose camera or subject motion · Create a short animation · Check distortions · Trim and caption
Outputs
An animated clip · The reusable still
Human work
Choose meaningful motion · Protect identities and objects · Control timing in the edit
Useful when
The image already communicates well · A restrained move supports attention
  1. Approve the still firstFix composition and rights before adding motion.
  2. Choose motionKeep the smallest movement that helps.
    • Camera movePan, zoom, or crop: Preserve the still and animate the frame in an editor.
    • Generated subject motionInspect each frame: Watch for identity, hand, tool, and background drift.
  3. Assemble and captionTreat the clip as one piece of a finished video.
Assemble existing clipsRecord or license concrete shots and edit them into a sequence.
Starts with
A beat list · Owned recordings or licensed clips · Consent and release records
Main actions
Plan coverage · Capture or select clips · Edit for message and continuity · Add sound, captions, and export
Outputs
A timeline and master export · Reusable raw clips
Human work
Direct or source footage · Manage consent and bystanders · Edit the sequence
Useful when
Real actions matter · Reusable documentary footage is valuable
  1. Plan coverageGive each clip a message job.
  2. Capture or licenseRecord releases, locations, and stock terms.
  3. Build the sequenceChoose timing, sound, captions, and transitions.
Create a narrated still sequenceTime approved images, slides, or graphic cards to voice and captions.
Starts with
A short script or beat list · Owned images or editable slides · Narration and caption plan
Main actions
Assign one purpose to each beat · Set exact timing · Add narration and captions · Export and test
Outputs
A finished video · Reusable images, narration, captions, and layout
Human work
Write and sequence · Design readable cards · Check audio and accessibility
Useful when
Exact information matters · Controlled, editable assets beat cinematic motion · A simple method supports the audience
  1. Write three beatsKeep one idea per beat.
  2. Prepare reusable assetsKeep images, text, voice, and captions separate.
  3. Set timing in an editorMake words readable and voice understandable.
  4. Export and testCheck captions, device playback, and platform compression.
Combine pathsUse generated, recorded, animated, and graphic elements where each fits best.
Starts with
A complete beat plan · A source and permission record for each asset
Main actions
Assign an approach to each beat · Keep asset boundaries visible · Assemble in one timeline · Credit and archive sources
Outputs
A mixed-workflow master · Independent reusable assets and provenance
Human work
Choose the simplest useful method per beat · Make the pieces feel continuous · Explain what was generated, recorded, or edited
Useful when
No single method covers the message well · The project benefits from controlled text plus expressive imagery
  1. Separate the jobsDistinguish acquisition, narration, graphics, editing, captions, and export.
  2. Choose per beatUse the method that gives the needed control.
    • Controlled informationUse editable graphics or stills: Keep exact facts directly correctable.
    • Expressive momentUse recorded or generated footage: Inspect, credit, and finish it.
  3. Finish one masterUnify timing, color, sound, captions, and export.

Start with your intention

Use My experiment to note your audience, material, limits, equipment, and sharing restrictions. Every field is optional. “Try with our example” fills only an empty goal and constraints field.

Prompt · version 1

Investigate ways to make a short video

When you have a message but no settled combination of shots, stills, narration, editing, and captions.

Separate making a shot from finishing a useful video and map feasible workflows.

Example: A repair café needs a 10–15 second invitation with three beats and readable captions.

Show the full prompt

The published prompt stays unchanged.

Source checked ·

Tools to investigate

A practical starting shortlist, chosen for different workflows rather than ranked quality. Features were checked against official sources on this date; we did not run a comparative test of these services. Plans and interfaces can change.

This shortlist reflects the date above. Confirm current features, terms, pricing, privacy, exports, and accessibility before choosing.

A high-level workflow

  1. Plan a short beginning, middle and ending before generating anything.
  2. Choose generated shots, recorded footage, or still images with simple movement.
  3. Assemble the shots in an editor and add narration, sound and captions.
  4. Watch the full export, check continuity and readability, and carry credits into the finished video.

Runway

Purpose and fit
Trying a short generated shot from a written idea or starting image.
How to explore it
Plan one shot, describe the action or supply a starting image, generate a short clip, then select and edit it into your story.
Watch for
A convincing first frame does not guarantee stable motion or continuity. Review the full clip and check the current model, plan and export limits.

Sources checked

Adobe Firefly

Purpose and fit
Adding movement to an image you already have permission to use.
How to explore it
Start with a clear image, describe the desired motion, generate a short shot, then assemble it with titles, audio and captions in an editor.
Watch for
Check the model and output terms, and compare the ending frame with the starting image for unwanted changes.

Sources checked

DaVinci Resolve

Purpose and fit
Building a controlled video from stills, recorded footage and sound.
How to explore it
Arrange your shots or stills on a timeline, set their duration, add narration and captions, then export and watch the whole result.
Watch for
This is an editor, not a one-prompt video generator. Begin with a few clips; some advanced tools require the Studio edition.

Sources checked

Decide

What matters for your project?

Use a few criteria and a separate must-have list. An unknown must-have remains unresolved.

Must-have

Message clarity

A first-time viewer understands what the invitation asks.

Acceptable: After one viewing, the viewer can state the purpose and next action.

Must-have

Captions and timing

Every essential line is readable for long enough.

Acceptable: Captions match the intended text and cover all three beats.

Must-have

Rights, consent, and privacy

Every face, voice, clip, image, font, and upload is permitted.

Acceptable: Sources, releases, service terms, storage, and disclosure are recorded.

Preference

Continuity

People, objects, and space remain understandable across the sequence.

Acceptable: No change distracts from the three-beat message.

Preference

Correction effort

The cost of fixing one detail is visible.

Acceptable: A maker can change one caption or timing detail and record the work required.

Must-have

Reusable assets and export

The project leaves useful source material and a playable master.

Acceptable: Images, captions, audio, project settings, and a common video format can be retained.

AI reported

What an AI response says. Follow the sources before relying on it.

Source checked

What an opened source states. This does not prove product performance.

We tested

What RS observed in a dated, bounded trial.

Still uncertain

What is unknown, unverified, or outside the trial.

Prompt · version 1.0

Shape video criteria one question at a time

Before choosing a production path or when one impressive shot is distracting from the full message.

Turn a moving-image idea into criteria for a complete, shareable result.

Example: The invitation needs three clear beats, captions, continuity, and a reusable poster frame.

Show the full prompt

The published prompt stays unchanged.

Prompt · version 1.0

Compare two video workflows, including finishing work

After producing the same short invitation as an animated still and an edited still sequence, or another feasible pair.

Compare complete short-video paths on message, continuity, effort, and usable outputs.

Example: Compare two 10–15 second repair-café invitations built from the same three beats.

Show the full prompt

The published prompt stays unchanged.

Experiment

Build the same 12-second invitation in two feasible workflows

Compare a controlled still sequence with an animated still or another available path, including captions and export.

Ready-to-use brief

Make a 10–15 second fictional repair-café invitation in three beats: bring one object, explore what it needs, learn together. Use owned assets, keep text editable during finishing, and provide captions.

Input

One approved repair-café illustration, a three-beat timing plan, caption text, and an optional permitted voice or sound bed.

Limit

Aim for 60–90 minutes and no purchase. Choose the destination, dimensions, and spending limit before starting.

Path 1

Assemble three held views or stills in an editor with direct timing control.

Path 2

Animate one approved still with a simple pan or zoom, then finish it in the same editor.

Only one candidate? Build one workflow, then correct one caption or timing detail and record how much work the change requires.

Define success before you begin

  1. Define the three beats and success before making footage.
  2. Preserve source images, caption text, and exact settings.
  3. Build two feasible paths with the same duration and message.
  4. Add or attach captions and export both files.
  5. Review frame changes, timing, continuity, accessibility, rights, effort, and cost.
  6. Choose the next step without claiming a general product winner.

Observation questions

  • Can a viewer state the message after one viewing?
  • Do people, objects, and space remain continuous?
  • Are captions readable and timed?
  • Which source assets remain reusable?
  • How much correction, editing, time, and cost were required?

What would you do next? Continue, change an approach, check a claim, try a simpler method, or stop.

Recorded on 2026-09-17

We ran the landscape prompt

Separates footage generation from finishing and proposes a paired generated-motion versus controlled-asset 12-second experiment.

Read the prompt, response, sources, evidence labels, limits, and next question.