Turn a written brief or reference image into a video with synchronized sound. Build a single shot up to 30 seconds long, from the opening moment to the final reveal.
Introduce the setting, bring the product into view, and finish on a shot that gives the viewer time to read the shape, material, or packaging. A reference still helps define the object while your prompt controls the action around it. Keep the product and the camera move simple for the first take.
Plan an entrance, a change in movement, and a closing pose instead of asking for constant action. Describe the stage, lighting, and sound as part of the performance. For a recurring character, select references that show the same appearance and wardrobe, then specify how the subject should move through the scene.
Test a tracking shot, a slow orbit, or a reveal through a doorway before building a larger sequence. Give the camera a clear starting position and destination, with a small number of visual beats between them. Once the movement works, adapt the framing and pace to the final brief.
Use text-to-video when you want to invent a scene without a starting image. Describe the subject, setting, action, camera movement, and sound. This is a useful starting point for exploring a new visual idea before preparing additional material.
Use image-to-video when you already have a frame that establishes the subject and composition. Write the motion you want to add: how the subject moves, how the camera responds, and what changes in the environment. A clear starting image and a focused movement instruction are easier to assess than a brief that changes the whole scene at once.
Choose reference mode when you need to combine material for appearance, movement, or sound. These inputs give you more ways to explain the intended shot; they do not replace the need to describe how the scene should unfold.
Give each reference a specific role. An image can establish a character, a product, or a location. A video can demonstrate movement or camera language. Audio can help communicate a sound direction. Identify the part you want to carry into the new shot, especially when a source contains details you do not want to reproduce.
For a product scene, begin with one clear view of the object and a short description of the action. Add another image when a second angle resolves an important detail. For a moving subject, choose a reference with an unobstructed view rather than several images that disagree about clothing or lighting.
Keep instructions explicit about what stays fixed and what can change. For example, preserve the subject and wardrobe while changing the background, or preserve the location while altering the camera path. Add another reference only when it addresses a specific weakness in the result.
Seedance 2.0 already supports multimodal references and joint audio-video generation. The main differences in the current Pixwit integration are the available shot length and reference capacity. Use the comparison below to decide whether those larger budgets matter for your scene.
Choose 2.5 when a continuous shot needs more time for setup, action, and an ending, or when several reference elements are necessary to describe the scene. If a short 2.0 workflow already produces the result you need, compare both versions with the same prompt and references before changing your production setup. More inputs or a longer generation do not automatically improve an individual shot.
| Setting | Seedance 2.0 | Seedance 2.5 |
|---|---|---|
| Output duration | 4–15 seconds | 4–30 seconds |
| Reference images | Up to 9 | Up to 30 |
| Reference video files | Up to 3 | Up to 10 |
| Reference audio files | Up to 3 | Up to 10 |
| Total reference video duration | Up to 15 seconds | Up to 30 seconds |
| Total reference audio duration | Up to 15 seconds | Up to 30 seconds |
Work through these five steps to turn a brief into a usable first take. Keep the subject, setting, and intended action consistent between iterations so you can tell which change improved the result.
Start with text for a new scene, an image for a defined opening frame, or reference mode for a more specific appearance or movement. Keep Seedance 2.5 selected as you change modes, and upload only the material that contributes to the shot.
Name the main subject, the location, and the visual details that must remain consistent. For an upload, check that the subject is readable and the framing suits the intended output. Avoid references with conflicting outfits, lighting, or backgrounds unless you explain which elements to use.
Write the scene in the order it should happen: the opening moment, the main action, and the ending. Give the camera one clear direction, such as a slow follow or an orbit. Specify dialogue, room tone, footsteps, or music where they belong instead of leaving the sound direction implicit.
Choose an aspect ratio for the destination, enough time for the planned action, and an available resolution. Review the credit amount displayed on the generation button before submitting. When the framing or movement is still undecided, make a shorter test before committing to the full scene.
Watch the opening, the main action, and the ending. Look for unwanted changes to the subject, abrupt camera movement, and audio that misses the intended timing. Revise the instruction or reference responsible for that problem while keeping the rest of the brief stable; then generate the next take.
Example prompt
16:9 | A singer walks from a dressing room through a backstage corridor, meets dancers, steps onto a lit stage — one continuous 20-second take, warm tungsten backstage then cool stage wash, footsteps, crowd swell, handheld follow
Start with one clear brief, preview the result, and refine the shot.
