Vidu Q4 Reference to Video
Vidu Q4 Reference to Video: Use Your Character and Scene Images
Reference to video uses multiple images to guide a generated scene. Add character artwork, product photos or location images, then describe how the subjects interact and how the shot unfolds.
What is Vidu Q4 Reference to Video?
Reference to video uses a set of images to describe the people, objects and setting of a scene. A reference is visual guidance rather than a frame that must appear at the start. On Vidu Q4, you can organize up to nine images, explain the role of each one and describe the action that connects them. This is useful when your idea needs both a recognizable subject and a particular environment.
Key features of Vidu Q4 Reference to Video
From source material to shot direction, organize the idea into a scene you can revise.
Multiple image references
Combine character artwork, a product photograph and a location image in one scene brief. Choose only images that have a clear purpose; additional pictures can introduce conflicting details.
A clear role for each image
Tell the prompt which reference supplies the person, which supplies the object and which supplies the background. Explicit assignments make a complex idea easier to revise.
Subject and camera direction
References establish appearance; the prompt describes what happens. Add one main action and a camera movement, such as a slow push-in or a tracking shot, instead of requesting several unrelated scenes.
Flexible framing
Select the frame shape before arranging your composition. A vertical product reveal and a wide character scene need different subject placement, even when they use the same reference assets.
Optional scene audio
Enable audio when sound belongs to the scene and describe the ambience or a brief line in the prompt. This workflow accepts image references; it does not provide voice-file uploads.
Reusable reference assets
Keep reference images in your library and save the associated prompt and settings as a draft. You can return to the brief without rebuilding the image selection from memory.
How to use this video workflow
Start with your material, describe the shot, then choose the available settings.
- 01
Select reference images
Add up to nine images from your device or asset library.
- 02
Describe the scene
Explain the subjects, their interaction and the camera movement.
- 03
Choose settings and save
Set the duration, resolution, aspect ratio and audio, then save your scene to continue editing.
Why prepare your video with ViduAI?
Bring assets, prompts and settings together around one clear shot.
Start with your own assets
Existing character, product and scene images can supply the visual starting point, while the prompt describes action and camera direction. Identify the important details and choose the right input mode before building a variation.
Keep the brief and settings together
The prompt, duration, aspect ratio, resolution and audio choice belong to one scene brief. Save the draft so the next revision can change one factor while retaining the rest of the inputs and settings.
Plan clips for the final edit
Break a complex idea into focused short shots, then organize order, captions and timing in an editor. A clip can explore composition and movement; exact brand text and final delivery quality still require a careful review.
What can you plan with Vidu Q4 Reference to Video?
Character storytelling
Bring a character design into a new location and describe a single reaction or gesture. Start with restrained movement so you can assess the face, clothing and scene together.
Product concepts
Use a clean product image with a setting reference to plan an advertising shot. Review logos, packaging text and small details before using any generated material in a campaign.
Illustrated worlds
Combine a subject drawing and a landscape illustration, then describe weather, light and motion. Keep the visual styles compatible to avoid asking the scene to resolve incompatible art directions.
A series of shot briefs
Reuse the same subject references in separate drafts while changing the action or camera direction. Review each result individually; shared inputs do not guarantee identical details across clips.
Write the action, not just the atmosphere
Choose distinct references with compatible lighting and style. Avoid assigning two different faces to the same character or combining several contradictory room layouts. Reference input guides generation; it does not copy every pixel into the result.
Start with the subject and its relationship to the inputs, then add the main action, camera direction and any sound instructions. Change one factor at a time so revisions remain easy to assess. Adapt the example prompt to your assets and creative direction.
What to check before you start
Generated scenes can change faces, hands, lettering, geometry or contact between subjects. References guide creation and do not guarantee pixel-for-pixel preservation or identical details across clips. Upload material you have permission to use and obtain appropriate consent for real-person depictions.
Frequently asked questions
Inputs, output settings and access on Vidu Q4.
How many reference images can I add?
The current Vidu Q4 workspace accepts up to nine reference images. Official Vidu model limits are separate from the limits of this integration.
Can I use references with first and last frames?
No. Reference-image input and first/last-frame input are separate modes here. Choose the mode that matches how your images should guide the scene.
Can I upload reference video or voice audio?
Not in this release. The reference workflow currently accepts images. Generated audio can be enabled separately, with sound directions in the prompt.
What clip settings are available?
Choose an integer duration from 4 to 15 seconds, 480p or 720p, an aspect ratio and optional generated audio. These are the current workspace settings.
Do reference images become the first frame?
No. Use image to video when a specific picture should define the opening composition. Reference mode uses the pictures to guide subjects and setting, so the opening frame may differ from any single reference.
Can I combine reference images with a last frame?
The current workspace separates reference mode from first/last-frame mode. Choose the mode that fits your shot rather than combining their inputs in a single draft.
Bring your reference images together
Choose your assets, describe the shot and save your first video draft.
Open the video workspace