I Love The Smiths AI: Listener and Speaker photo guide
Use two separate photos, each showing one consenting adult. A group photo is not a substitute for the two role slots.
Listener: slot 1
Choose a sharp front-facing or slightly turned portrait with eyes, cheeks and jaw visible. You do not need to wear headphones in the input photo. Avoid accessories that hide the face.
Speaker: slot 2
Use even light and a clear face. Include the head and shoulders; avoid extreme angles, heavy beauty filters and motion blur. Confirm this is the person who should say the line.
File checklist
Use JPEG, PNG, WebP or HEIC (iPhone), up to 10 MB per photo. Images below 512 pixels on either side are rejected locally. Browser face checks run where supported; server photo checks follow login and consent. Confirm which person belongs in each slot.
Consent and privacy
Only use adults who consent to this AI recreation. Uploaded photos are eligible for cleanup after 30 days, with active jobs allowed to finish, subject to the Privacy Policy. Do not upload identity documents or photos of minors.