Use Seedance Reference to Video to keep characters, products, and visual details recognizable while generating fresh scenes, natural motion, and cohesive stories from one or more reference images.
Choose a person, product, object, or visual style as your reference. Seedance carries the defining facial features, colors, clothing, and design details into the generated video, even as motion and framing change.
Place a referenced character or object into a completely new environment without rebuilding its appearance from scratch. Combine people, props, and locations in one controlled generation while preserving the visual identity that matters.
Prompt
A recurring heroine enters a neon-lit street market, turns toward the camera, and smiles while the crowd moves naturally behind her.
Move the same character or product between different settings, actions, and camera angles. Reference to video keeps the core subject stable, making it easier to build campaigns, story sequences, and repeatable creative variations.
Upload separate character, product, or scene references and guide them into a unified shot. Seedance helps connect the visual elements with believable composition and motion instead of producing a disconnected collage.
Prompt
A lifestyle creator picks up the referenced bottle on a sunlit café table, shows the label to camera, and walks into the city street.
Seedance analyzes the subjects and relationships in your reference images, then builds motion that feels connected across the full scene. Characters, objects, backgrounds, and camera movement work together instead of shifting independently.
Keep the assets your audience recognizes while adapting them for editing, marketing, animation, and social publishing.
Video Editors
Build recurring characters and compatible B-roll without manually matching every shot.
Brand Marketers
Reuse the same product and brand details across fresh settings, formats, and campaign concepts.
Animators & Game Teams
Develop storyboards, animated sequences, and game concepts while keeping character traits recognizable.
Social Creators
Bring the same persona or visual world into reels, shorts, and recurring social stories.
Turn one or more visual references into a coherent AI video without a complicated production workflow.
Step 1
Add clear images of the characters, objects, products, or environments you want Seedance to preserve.
Step 2
Write a prompt covering the action, setting, camera behavior, mood, and any important subject interactions.
Step 3
Create the video, check identity and scene continuity, then adjust the prompt or references for a stronger result.
Common questions about creating consistent characters, products, and scenes from reference images.
A reference to video generator uses one or more uploaded images as visual anchors while creating a new video. The images guide the identity, appearance, or style of characters and objects without requiring the generated clip to begin from the exact uploaded frame.
Image to video usually animates a single source image as the opening frame. Reference to video uses images as identity and design guidance, so the subject can appear in new poses, camera angles, actions, and environments.
Yes. Clear references can guide people, fictional characters, products, props, clothing, environments, and overall visual styles. Results are strongest when the defining details are clearly visible.
One clear image can work, while multiple complementary angles often provide stronger identity coverage. Use references that agree on the subject's appearance and avoid images with conflicting details.
Use sharp, well-lit images with an unobstructed subject and enough resolution to show important facial, clothing, shape, and texture details. Simple compositions make the visual signal easier for the model to follow.
Reference-driven workflows are useful for product demos, campaign variations, recurring brand characters, storyboards, and social series. Make sure you have the rights and permissions required for every uploaded reference and intended use.
Upload your references, describe the next scene, and turn familiar characters or products into a coherent AI video.
