AI Dance Video Reference Image Guide: Prepare a Performer for Dance Mode

Updated June 9, 2026. Use this guide to choose a clear, rights-safe performer reference before you spend Dance Mode credits.

The Short Answer: Use One Clear Performer

Use a reference image where one performer or character is unmistakable. The body shape, outfit, and silhouette should be easy to read, with no competing foreground subjects. The image does not need to show a dance pose; it needs to make the intended dancer obvious and rights-safe.

VibeMV Dance Mode is a per-shot workflow. A better reference helps the system understand who should move before it generates a 4-10 second Dance segment.

Which guide should you read next? For the full workflow, read How to Make an AI Dance Music Video. For cost planning, read AI Dance Video Pricing and Credit Budget. For product facts, read AI Dance Video Generator.

Reference Image Checklist

Good Reference Patterns

Use images like:

• solo artist press photo with clear body framing
• original character art with one obvious performer
• stylized avatar with readable outfit and limbs
• full-body or waist-up image against a simple background

The reference can be realistic, stylized, illustrated, or avatar-like if the main subject is clear and the style is intended for the video.

Risky Reference Patterns

Avoid images like:

• crowded group photos
• two or more people with equal prominence
• tiny subject in a wide landscape
• face-only portrait when the movement needs body clarity
• heavy blur, darkness, or compression
• blocked arms or legs
• celebrity, influencer, or real-person likeness without approval

These may cause rejection, weak output, or rights problems after generation.

Single Performer Beats Group Shots

Dance Mode is built around one clear performer or character. Group choreography is a different job. If you need multiple dancers, start by deciding whether the release truly depends on that. If it does, filmed choreography or a broader video production workflow may be more appropriate.

For VibeMV Dance Mode, use the cleanest single-performer reference first. You can still assemble a richer MV by combining that Dance shot with normal scenes, lip-sync sections, and non-Dance visuals.

Real Person And Likeness Rights

Do not treat image upload as rights clearance.

Before using a real person's image, confirm:

• they approved the image use
• they approved AI generation for the intended context
• they approved commercial or public use if needed
• no photographer, label, brand, venue, or platform restriction blocks the use

VibeMV can help generate a video asset. It does not clear likeness rights, copyright, publicity rights, music rights, or platform policy requirements.

For broader release review, read the music video copyright guide.

Prompt Tips For Reference Images

The reference tells Dance Mode who should move. The prompt tells it what the shot should do.

Good prompt details:

• one performer centered
• simple background
• chorus hook or beat drop energy
• 9:16 vertical or 16:9 landscape
• no crowd, no extra dancers, no celebrity copy

Example:

one clear performer from the reference image, centered full-body 9:16 dance hook, confident chorus movement, simple stage lighting, clean background, readable silhouette, no crowd, no extra dancers, no celebrity likeness

Preflight Before Spending Credits

Before generating a Dance shot, ask:

• Is there exactly one intended dancer?
• Is the body visible enough for movement?
• Would a reviewer understand who should move?
• Is the image rights-safe?
• Does the prompt avoid real-person or celebrity imitation?
• Is Dance Mode worth 12 credits per generated second for this section?
• Would normal mode or lip-sync solve the section more cheaply?

Reference Image FAQ

What reference image works best for an AI dance video?

Use one clear performer or character with readable body shape, visible outfit, enough image quality, and no competing foreground subjects.

Can I use a group photo for Dance Mode?

Usually no. Dance Mode works best with one clear main performer. Group photos can create ambiguity when several people compete for attention.

Can I use a real person's photo for an AI dance video?

Only use real-person images when you have the right to use that likeness for the intended generation and publication. AI generation does not clear consent, likeness, copyright, or publicity rights.

Why might a Dance Mode reference image be rejected?

A reference may be rejected or perform poorly when there is no clear main subject, multiple competing people, an unclear body, unsafe content, heavy blur, blocked limbs, or rights-sensitive material.

Does the reference image need to show a dance pose?

No. A neutral or performance-ready reference can work. The reference should identify the performer or character; the Dance prompt and selected song moment guide the movement.

Next Step

After preparing a clean reference, use AI Dance Video Generator for the product workflow or How to Make an AI Dance Music Video for the full per-shot process.