PROMPT OR TEMPLATE
Do you need an I Love The Smiths AI prompt or a template?
DIRECT ANSWER
The short answer
You do not need to write a prompt for this site. It uses a fixed two-role template: you choose a Listener photo and a Speaker photo while the scene structure and motion guidance are configured privately. A generic text-to-video tool may require a prompt, but it is a different workflow.
What is the difference between a prompt and a template?
A text prompt describes a desired video in words and leaves the model to interpret many details. A fixed template defines roles, staging and motion ahead of time, then asks the visitor for the variable inputs—in this case, two portraits.
Templates are useful for a recognisable format because visitors do not have to reconstruct camera direction, timing and role order in a blank prompt box. They are less flexible: you cannot assume you can change the setting, dialogue, duration or number of characters.
Why does this generator not show a prompt field?
The public workflow asks only for a Listener and a Speaker. The effect prompt and reference path are server-only configuration so visitors cannot accidentally expose or alter operational instructions. That separation also keeps provider details and cost controls out of browser code.
After you confirm permission and start a generation, the server securely combines the two uploaded portraits with the prepared template. You do not need to copy, reveal or tune the private prompt.
What if you use a generic image-to-video model?
A generic model may ask for a motion description, but copying a random prompt cannot guarantee the same framing, identities, dialogue or timing. Model capabilities and provider rules differ, and some services do not support two identity references in one shot.
Describe original staging rather than requesting a frame-for-frame copy of a copyrighted scene. Avoid claiming that generated people are the real actors or that an unofficial result is endorsed by the film, band or rights holders.
Editorial example: Create an original, clearly fictional elevator meet-cute with two consenting adult characters. Keep the first person as the headphone listener and the second as the conversation starter; preserve both identities and use restrained, natural movement.
Do you need to upload the original elevator clip?
No. Visitors do not need to find, download or upload a film clip. The studio uses an independently created or properly licensed reference asset stored privately by the operator. Its storage path is not exposed through public configuration.
The site is an independent fan-made tool and is not affiliated with The Smiths, the film, its studio, cast or rights holders. Inspiration does not grant permission to redistribute copyrighted footage or music.
What can a fixed template still get wrong?
A template narrows the instruction, but it cannot guarantee perfect faces, lip movement, gaze, hands or continuity. Two-person video is especially sensitive to identity blending and role reversal. Review the entire result before sharing it as an AI-generated interpretation.