Which photo works for this Say It Right dance?
Start with a photo where the face is clear at normal viewing size. A waist-up portrait can work; a full-body image gives the model more evidence about clothing and shoes. The photo supplies the lead’s appearance, while the preset supplies the performance.
| Your photo | What to expect | Before uploading |
|---|---|---|
| Full body, one person | The strongest guide for the whole outfit and proportions. | Keep hands separate from the torso and include the shoes. |
| Waist-up selfie | Usable, but lower-body clothing and shoes must be invented. | Use an unobstructed face and a simple, visible outfit. |
| Face-only crop | Less evidence for the body, outfit and side profile. | Choose a wider portrait if you have one. |
| Group photo | The template has one replacement slot; it cannot assign several faces. | Upload a separate image of the intended lead. |
| Pet portrait | The prompt supports a pet lead, with generated upright dance movement. | Show the face, coat markings and body clearly; expect a playful AI result. |
JPG, PNG or WebP; 300–6000 pixels on each side and up to 20 MB. Use your own photo or one you have permission to use. Avoid sunglasses, heavy filters and another face overlapping the subject.
What stays in the Say It Right choreography?
This is a dedicated preset, so you do not need to find a dance video or describe each step. The model follows one continuous phone take: a lead dancer, side-to-side footwork, shoulder rolls, turns and a brief forward bend in a crowded night scene.
| Element | Comes from | What that means |
|---|---|---|
| Lead’s face and outfit | Your uploaded photo | Identity and clothing are guided by the photo, but fine details may drift. |
| Dance, timing and camera | The built-in Rio take | The phone moves with the scene; this is not a fixed studio camera. |
| Crowd, street lamp and tents | The preset scene | Your photo’s room or background is replaced. Other dancers are not extra upload slots. |
| Sound | Audio from the same take | The final MP4 includes it. A different remix may need a new starting point and timing. |
Choose 10, 15 or 20 seconds
All lengths start at the same moment. Choosing a longer clip continues the performance; it does not slow a short clip down or repeat it.
| Length | What is included | Models |
|---|---|---|
| 10 seconds | The close-up opening, stepping back, shoulder rolls and low hand movements. | All five models |
| 15 seconds | The opening plus a side-profile turn and the return toward the lens. | All five models |
| 20 seconds | The full preset, including the forward bend, brief camera jolt and return upright. | Seedance 2.5 only |
The initial selection is Seedance 2.5, 15 seconds, 480p. Mini at 10 seconds and 480p costs the fewest credits. Higher resolution adds pixels, not a guarantee of better anatomy or a closer likeness.