Multi-Shot Sequence Plan
Plan multi-shot video generation on the Astorie canvas to build a connected multi-shot AI video sequence. Organize shared reference images, generate each shot with a focused motion brief and review continuity across cuts before assembling the final edit.
A convincing multi-shot AI video sequence depends on what stays consistent from one cut to the next. Comparing isolated one-shot prompts with a planned canvas sequence shows why shared visual references and per-shot motion briefs prevent scenes from drifting apart.
Start with a short shot list and the key visual references that anchor your scene on the Astorie canvas.
Break your scene into individual shots, such as an establishing wide, a medium action beat and a close reaction. Prepare or generate reference images in Image nodes for the characters and locations that must stay recognizable.
Create a Video node for each planned beat and connect the relevant single Image node or multiple Image nodes for Multi-Image Reference where supported. Write a concise motion prompt that describes only the action and camera move for that specific shot.
When a follow-up shot continues the exact moment of a previous clip, use Frame Extraction to capture the current or final frame as a JPEG still. Use that captured frame as the visual starting point or reference check for the next shot.
Review the generated clips in sequence order to check screen direction, lighting and pacing. Regenerate any shot that breaks continuity, then export the approved takes for timeline cutting and audio sync in your NLE.
Check three continuity anchors across every take before locking your selects for the edit.
Compare facial features, hair, clothing colors and held props across wide and close-up shots. Keeping the same approved reference image connected on the canvas reduces drift between separate generation runs.
Check that key light direction, time of day, color temperature and background architecture match across cuts set in the same location. Even a small shift in shadow direction can make two shots feel like different scenes.
Inspect the first and last frames of adjacent clips to confirm that subject posture, gaze direction and movement speed connect smoothly across the cut.
Keep your shot plan, reference connections and editorial handoff aligned across the sequence.
| Input | A shot list, approved reference images in Image nodes and per-shot motion prompts connected to Video nodes on the canvas. | |
|---|---|---|
| Output | Individual reviewed video clips designed to cut together as a multi-shot sequence in your editing software. | |
| Review | Check character identity, wardrobe, lighting continuity, screen direction and clean in/out frames across adjacent shots. | |
| Plans and credits | Each shot generation and revision uses workspace credits according to the selected video model; see the pricing page for plan details. | |
| Rights and assembly | This workflow builds a sequence shot by shot on the canvas; final timeline editing, transitions and sound design happen in your NLE. |
Use these connected tools and guides on Astorie to anchor references, extract bridge frames and hand off your sequence.
Start individual video shots from an approved reference image to keep subject appearance grounded.
Capture the current or last frame of an approved clip to bridge into the next shot in your sequence.
Review Image to Video, Multi-Image Reference and Text to Video connections on the canvas.
Organize your approved multi-shot selects and version labels for finishing in an external editor.
Rather than relying on one prompt to guess an entire edited scene, this workflow helps you plan and generate a sequence shot by shot on the Astorie canvas. Building one shot per node gives you direct control over framing, reference images and retakes when a single beat needs revision.
Start from a clear character reference still in an Image node and reuse that verified reference across the Video nodes in your scene. Where your chosen model supports Multi-Image Reference, you can also combine character and environment references before generating.
Use Frame Extraction when the next shot needs to pick up directly from the ending pose, lighting or camera position of the previous clip. Capturing the final frame gives you a concrete visual reference to guide the next generation step.
Once you have reviewed and approved the individual clips on the canvas, export the selected takes and bring them into your nonlinear editor to trim cut points, add transitions, balance color and mix dialogue and music.
Lay out your reference frames, generate one focused shot per beat and review how the sequence cuts together.
