Two modes, one model
Use text-to-video to generate a scene from a description, or image-to-video to animate a specific photo. For products and brand work, image-to-video keeps your exact item; for concepts and b-roll, text-to-video gives more freedom.
Pick duration and ratio up front
Choose the aspect ratio for your channel before generating: 9:16 for TikTok/Reels/Shorts, 16:9 for YouTube and websites, 1:1 for feed posts. Keep clips short (around 4–12 seconds) for the most stable, artifact-free motion.
Write motion-aware prompts
State the subject, one clear action, the camera movement, and the mood — for example, 'a perfume bottle on marble, slow 360 rotation, soft studio light, elegant'. One strong action beats several competing ones, which tend to produce jitter.
Quick settings reference
| Setting | Recommendation |
|---|---|
| Duration | 4–12s (short = more stable) |
| Aspect ratio | 9:16 social · 16:9 web |
| Mode | i2v for products · t2v for concepts |
| Source image | High-resolution, sharp |
Frequently asked questions
Does Seedance 1.5 Pro support image-to-video?
Yes — it supports both text-to-video and image-to-video in one model.
What clip length is best?
Around 4–12 seconds gives the most stable motion; stitch clips for longer sequences.
Which aspect ratio should I use?
Match the channel: 9:16 for short-form, 16:9 for YouTube/web, 1:1 for feeds.
How do I reduce jitter or artifacts?
Use one clear action, keep clips short, and provide a sharp high-resolution source for image-to-video.
Is it good for e-commerce?
Very — animate a product main image into a short clip for ads and detail pages.
Text-to-video or image-to-video for ads?
Image-to-video when the exact product matters; text-to-video for stylized concept ads.
Vidxo