How it works
- 01
Prepare two separate portraits
Choose two sharp, consenting adult photos with visible faces and similar lighting, up to 5 MB each. If both people start in one picture, use Crop first so each reference shows one person cleanly.
- 02
Generate one two-person interaction frame
Open the two-person group photo tool (Nano Banana 2.1, 6 credits). Place Reference 1 on the left with over-ear headphones and Reference 2 on the right in an elevator-style scene, then inspect both faces before continuing.
- 03
Animate the combined scene in Image to Video
Save the combined PNG and upload it as the single first frame on the Image to Video page. Describe one brief movement, such as turning toward each other or lifting one hand toward the headphones, and review the 40-paid-credit quote.
Why the trend uses two steps here
External "I love the Smiths" meme templates map two separate portraits directly onto a 14-to-24-second elevator dialogue sequence with movie audio. PromptToVisual does not accept two separate portraits inside its video form, so you first combine both people into one still scene with two-reference composition and then animate that single image.
Realistic video duration, ratio and audio
The current Seedance 2.5 video form outputs a four-second MP4 with generated audio at 720p in either square (1:1) or widescreen (16:9). It does not add copyrighted music or recreate spoken dialogue from a film, and trial image credits cannot fund video requests.
Check faces and permissions before sharing
Both image composition and video animation redraw pixels, so expressions, hairstyles and proportions can shift. Use photos of adults who consent to the edit, keep the original portraits to compare likeness, and label AI-generated clips clearly.
Compare your options
| Option | Best for | Keep in mind |
|---|---|---|
| Two-person group photo | Combining two separate adult portraits into one shared scene as Reference 1 and Reference 2. | Still PNG only (6 credits); expressions and likeness may shift. |
| Image to video | Turning the combined two-person image into a short four-second clip. | One first-frame image only; 4s, 720p, 1:1 or 16:9, 40 paid credits; no licensed song. |
| AI Image Combiner | Placing one subject from Reference 1 into a separate elevator background from Reference 2. | Still PNG only (6 credits); two references total. |
| Crop | Splitting a group shot into single portraits or matching framing before upload. | Local browser crop only; does not combine or animate photos. |
Questions, answered
Can I upload two photos directly into Image to Video?
No. The Image to Video form accepts one first-frame photo. Combine your two portraits into one PNG first using Two-person group photo, then animate that image.
Does PromptToVisual include the Smiths song or movie audio?
No. Seedance 2.5 generates a four-second 720p clip with general AI audio and does not supply copyrighted music or film dialogue.
What does the two-step workflow cost?
Preparing a prompt is free. Generating the combined still costs 6 credits, and animating it costs 40 paid credits (trial credits do not cover video).