Creating videos with Seedance 2.0 is straightforward. Upload images, videos, and audio files that show exactly what you want, then let the AI handle the rest.
What you need:
- Reference images (up to 9 files)
- Optional video clips (up to 3 files, 15 seconds total)
- Optional audio files (up to 3 files, 15 seconds total)
- A clear idea of your desired output
The process takes 5-10 minutes from start to finished video.
Setting Up Your Workspace
Accessing Seedance 2.0
Seedance 2.0 is available through Morphic. Create an account to access the video generation interface. You'll see upload areas for different file types and a text field for your prompt.
Understanding the @ mention system
The @ mention system tells Seedance 2.0 what each file should control. Uploaded files are labeled as @Image1, @Image2, @Video1, @Audio1, etc.
Example usage:
- @Image1 defines character appearance
- @Video1 demonstrates camera movement
- @Audio1 establishes rhythm and pacing
Reference these labels in your prompt to assign roles to each file.
Step 1: Gather Your Reference Materials
Collect the visual and audio elements you want to reference:
Character videos:
- Clear photos showing character appearance
- Environment or background references
Motion-focused videos:
- Video clips demonstrating camera movement
- Examples of pacing or editing style
Audio-driven content:
- Music tracks for mood or rhythm
- Sound effects or ambient audio
Tip: Name files descriptively (e.g., "hero-character.jpg", "camera-pan.mp4")
Step 2: Upload Your Files
Upload your references in the designated areas:
| File Type | Formats | Maximum | Size Limit |
|---|---|---|---|
| Images | JPG, JPEG, PNG, WEBP, BMP, TIFF, GIF | 9 images | 30 MB each |
| Videos | MP4, MOV | 3 videos (15s total) | 50 MB each |
| Audio | MP3, WAV | 3 files (15s total) | 15 MB each |
Important:
- Use high-resolution files for best results
- Wait for upload confirmation before proceeding
Step 3: Write Your Prompt
Use the @ mention system to combine your references:
@Image1 shows the main character wearing a red jacket. @Image2 provides the urban background setting. @Video1 demonstrates a slow dolly-in camera movement. @Audio1 sets a contemplative, atmospheric mood. Create a scene where the character stands on a rooftop at sunset, looking out over the city.
Prompt writing checklist:
- Be specific about what each reference contributes
- Describe the action or narrative clearly
- Mention important details not in references
- Use clear, direct language
Step 4: Configure Output Settings
Choose your aspect ratio:
| Ratio | Best For |
|---|---|
| 16:9 | YouTube, desktop viewing, widescreen |
| 9:16 | TikTok, Instagram Reels, mobile-first |
| 1:1 | Instagram feed posts, Facebook |
| 4:3 | Traditional media, standard format |
| 3:4 | Vertical content, Pinterest |
| 21:9 | Ultra-wide cinematic content |
Additional settings:
- Duration: 4-15 seconds (shorter = faster generation)
- Resolution: 1080p default (2K available in Pro mode)
Important: Review settings before generating—changes require starting over.
Step 5: Generate and Review
Generation process:
- Click generate button
- Wait 2-5 minutes for processing
- Preview your video when ready
- Download or adjust and regenerate
Note: Most videos complete within 2-5 minutes. Complex 15-second videos may take slightly longer.
Tips for Better Results
1. Use high-quality references
Clear, well-lit images produce better results than dark or blurry photos. If your reference material is ambiguous, the generated video will be too.
2. Be specific in prompts
- Vague: "Urban environment"
- Specific: "Rainy city street at night with neon signs"
3. Test with shorter durations
Generate 5-second tests before committing to longer videos. This saves time and lets you refine your approach quickly.
4. Build a reference library
Save successful images, videos, and audio for reuse across multiple projects. This speeds up future generations and ensures consistency.
5. Match references to output
Visual examples work better than text descriptions. Want a specific camera movement? Show it with a video reference.
Troubleshooting Common Issues
| Problem | Solution |
|---|---|
| Character doesn't match reference | Use clearer, high-resolution images. Add explicit instructions: "@Image1 for exact facial features" |
| Motion looks unnatural | Provide video references showing desired motion. Simplify complex movements into shorter segments |
| Video doesn't match prompt | Check @ mentions reference correct files. Make prompts more specific and detailed |
Getting Started
Ready to create your first video? Head to Morphic and set up your account. Start with something simple—a character portrait or product showcase. As you get comfortable with the @ mention system, you can tackle more complex projects.
The key is experimentation. Try different reference combinations and prompt styles to see what works for your goals.
Frequently Asked Questions
Most videos generate in 2-5 minutes once you click the generate button. Complex 15-second videos at higher resolutions may take slightly longer. The preparation time (gathering references, writing prompts) varies based on your project but typically takes 5-10 minutes.
Seedance 2.0 generates complete videos, but you can't edit them frame-by-frame within the platform. If you need changes, adjust your references or prompt and regenerate. For post-generation editing, download the video and use standard video editing software.
Good reference images are clear, high-resolution, and show exactly what you want the AI to understand. Use neutral lighting, uncluttered backgrounds, and direct angles. Avoid blurry, dark, or ambiguous images that could confuse the generation process.
This depends on your account type and credit balance. Each video generation consumes credits based on duration and resolution. Check your account dashboard on Morphic to see your current credit balance and usage.
Seedance 2.0 supports six aspect ratios: 16:9 (YouTube, desktop), 9:16 (TikTok, Instagram Reels), 1:1 (Instagram feeds), 4:3 (traditional media), 3:4 (vertical content), and 21:9 (ultra-wide cinematic). Choose based on your publishing platform. If you're unsure, 16:9 is the most versatile option.
Use the exact same reference images across all generations in your series. Save your character references, style guides, and successful prompts. Reusing identical references ensures visual consistency from video to video.
