Couples
Smalltown Boy AI Couple: A Two-Photo Dance Guide
For a Smalltown Boy AI couple video, prepare one clear photo of each adult, agree who takes the first reaction shot, and choose a two-subject dance template. Review both faces after the swap, then add the chosen audio. You can prepare the photo pair in this studio without an account.
Use one portrait per partner
A shared selfie can be useful as a reference for the idea, but a tool with two identity slots needs two clearly separated subjects. Crop each person from the original or use two individual photos. Avoid leaving a second face at the edge of either crop.
Match the angle and clarity where possible. If one portrait is a sharp daylight image and the other is a dark side view, the inputs give the video model very different information. Start with a comparable pair before experimenting with dramatic styling.
Plan the joke and the role order
Choose who reacts first and who begins dancing in the background. A couple edit works best when that choice is intentional: one person can be the deadpan partner and the other the more theatrical performer. The role order also gives you a clear way to check the result.
Keep clothing instructions tied to a named subject. For example, “Subject A keeps the blue shirt; Subject B keeps the cream jacket.” Avoid vague instructions such as “swap them,” which could refer to either their positions or their identities.
Prepare your couple photo pair
- Select a portrait for Deck 01 and one for Deck 02 in our studio.
- Compare the crops and confirm both faces are unobstructed.
- Download the two-photo PNG if you want a visual plan. Keep the two original image files as well.
- For the animated result, choose a video provider that supports both identity references and the required dance format.
The PNG is a composition of your photos. Sign in to use video generation and view current service availability. Preparing the images does not create motion or a soundtrack.
Check faces, outfits and role changes
- Opening: the foreground face matches the intended partner.
- Exchange: each face remains recognizable when the performers move.
- Wide shot: outfits, body shapes and height relationships remain plausible.
- Ending: both partners are present, without one turning into a reference performer.
Compare the complete clip with both original photos before posting. A strong first frame alone does not show whether the identities survived the exchange.
Choose a rendering route
Mitte’s public Smalltown Boy page describes a two-person preset, while Media.io’s guide describes a reference-and-prompt approach. Check the current inputs and export terms for the service you choose. Our app guide explains how these routes differ.
Use photos you have permission to use, and agree with your partner before sharing a realistic edit. You can plan a playful duet without copying a specific creator’s faces or claiming a licensed example was generated from your uploads.