What a Smalltown Boy AI prompt can describe
In the Smalltown Boy AI Video Generator, separate appearance from performance when planning the edit. The two still images identify your cast: faces, hair or fur, distinctive markings, clothing and accessories where visible. The reference video provides the sequence of gestures, shot changes, camera movement and timing. A prompt connects each image to the correct moving character.
A description of a dance alone is not the reference file used by this template. If you experiment in a different editor, first check whether it accepts a video reference and two character images. The example below is intended for that kind of reference-guided edit; it is not a claim that every text-to-video tool accepts those inputs or will reproduce this choreography.
Name the characters by their opening roles
Call Image 1 Character A: the curly-haired dancer who originally wears purple and starts nearest the lens. Call Image 2 Character B: the mustached dancer originally in red-and-black plaid behind them. These visual descriptions identify targets in the supplied performance. They do not require your new characters to wear the source performers’ outfits.
Keep those names after the performers change positions. An instruction such as ‘replace the person in front’ is ambiguous once the other dancer comes forward. State that A remains A and B remains B through close-ups, rear views, side-by-side shots and partial occlusion. This is also why swapping the two upload slots before generation changes the intended casting.
| Reference | Controls | Keep separate from |
|---|---|---|
| Image 1 | Character A’s appearance | Who happens to be closest in a later shot |
| Image 2 | Character B’s appearance | A duplicate of Character A |
| Prepared video | Motion, shots, setting and timing | The replacement character photos |
A Smalltown Boy dance prompt example
Use the supplied dance video as the reference performance from its first frame. Image 1 defines Character A, who starts in the foreground; Image 2 defines Character B, who starts behind. Follow the same two characters when they exchange positions, appear side by side and reach toward the camera. Keep the assigned identities distinct throughout every shot.
Carry each character’s recognizable face, hair or fur, markings and visible outfit from its own image. If a reference shows a pet, retain that species and its distinctive features while adapting the dance to a natural character shape. Follow the reference’s head gestures, hand or paw movements, footwork, interactions, camera angles, wooden fence, plants, pavement and shot timing. Keep the existing opening and ending. Do not introduce extra characters, new scenes, text, logos or a different routine.
This is an explanatory prompt example, not a field you need to fill in on EasyVideoAI. The studio already assigns the two references and uses the prepared performance. Choose your photos and settings there; use a custom prompt only in a separate tool that actually supports editing one.
Describe continuity instead of adding more actions
The most useful movement instruction is to follow the reference sequence. Adding a new jump, costume change, extra camera cut or different setting creates another request for the model to satisfy and changes the scene you intended to recreate. Keep the original gestures and interactions as the motion guide when the goal is this specific dance.
For a Smalltown Boy cat dance prompt, identity and anatomy need attention together. A cat can have recognizable fur while a reaching paw or crossed limb still looks inconsistent. Ask for the same character through those moments, then inspect the actual result. Repeating stronger adjectives does not establish that the model will handle every overlap or close-up correctly.
Which audio belongs in the finished video?
In this EasyVideoAI workflow, the selected source video is sent without an audio track, with audio generation disabled. After generation, the application takes the generated picture and restores audio from the same selected source segment. The finished soundtrack therefore comes from the prepared performance, not from a new song or voice created from the uploaded portraits.
The public cat demonstration is a separate file with its own sound and framing. It does not supply the audio for your personalized result. Your choice of 12 or 21 seconds controls the generated segment and the source audio restored into it. A prompt requesting a longer story does not extend the actual duration selected in the studio.
Smalltown Boy AI prompt questions
Where do I paste the example? There is no free-text prompt field in the EasyVideoAI studio. The example explains the role instructions; your actionable controls are the two photo slots, length, resolution, frame and final confirmation.
Can a longer prompt fix the wrong opening character? First compare the upload thumbnails with Image 1 and Image 2. If the assignments are reversed, use the swap control before submitting. Changing text in another editor would not change a request already created here.
Will this prompt guarantee the same cats or exact gestures? No. Photos and the reference clip guide the result, and generated details can vary. Use the Smalltown Boy AI prompt principles as a review checklist: correct identities, consistent tracking, the selected performance and its source sound. For a finished result, inspect the video rather than relying on the wording alone.
Reviewed against the EasyVideoAI studio on October 19, 2026. Check the live studio for current availability and the final credit quote.