Hedra Omnia

Prompt Guide

Gallery

Click any example to view details

Focus:

What is Omnia?

Omnia is Hedra's image-to-video model designed for creating realistic human motion and dynamic scenes. Unlike text-to-video models, Omnia transforms a single image into fluid, natural video with precise control over movement, camera, and background elements.

Natural Human Motion

Industry-leading lip sync and body movement that maintains identity consistency throughout the video

Camera Control

Precise camera movements including orbits, push-ins, tracking shots, and dynamic angles

Background Dynamics

Add dramatic elements like explosions, weather effects, or fantastical creatures to any scene

Video Duration

Omnia generates videos up to 8 seconds in length. This duration is optimal for social content while maintaining high quality throughout.

Best for: Social media clips, product showcases, talking head content, UGC-style videos
Audio: Keep your audio under 8 seconds for best results. Natural pacing with brief pauses works better than rushed delivery.

Prompt Structure

Omnia prompts work best when they're concise and focused. A good prompt combines camera motion, subject action, and background elements to create dynamic, engaging video.

Camera Motion

Describes how the virtual camera moves through the scene. This creates cinematic depth and visual interest.

slow push in orbit around tracking shot dolly zoom handheld static tripod

Subject Action

What the person or subject physically does. Omnia excels at natural human movements and expressions.

walking forward turns head adjusts glasses gestures while talking stands up dances

Background Elements

Dynamic elements that happen around the subject. Great for adding drama, atmosphere, or surreal effects.

explosion behind rain falling dragon flies overhead cars passing smoke fills room leaves blowing

Use Cases & Examples

Podcast & Interview

Perfect for talking head content where the subject speaks directly to camera.

Example: Stationary tripod shot with subtle push-in on podcast host gesturing while speaking, studio lighting.
Example: Medium shot of host laughing on leather sofa, soft room ambience, natural hand movements.

Social Media & UGC

Selfie-style content that feels authentic and relatable. Handheld camera motion adds energy.

Example: Handheld selfie shot, woman talking excitedly to camera, walking on city street, natural movement.
Example: POV selfie angle, subject winking and smiling, slight camera shake, urban background.

Cinematic & Film

Dramatic scenes with bold camera movements and atmospheric effects.

Example: Cinematic slow push-in on woman in rain, neon lights reflecting, dramatic atmosphere.
Example: Orbit around warrior, burning ruins in background, smoke filling the scene dramatically.

Product & Commercial

Showcase products with authentic presenter style and natural movements.

Example: Close-up of hands holding product, gentle handheld movement, soft lighting, product in focus.
Example: Handheld selfie shot of influencer holding skincare bottle, talking enthusiastically, bathroom setting.

Creative & Experimental

Push creative boundaries with dramatic motion and surreal elements.

Example: Woman walking confidently across zebra crossing, massive explosion erupts behind her.
Example: Dramatic zoom into subject's face, dragon flies overhead in the background.

Audio & Voice

Audio is essential for bringing Omnia videos to life. The model synchronizes lip movements with your audio, creating natural-looking speech.

Voice Options

AI Voices: Access a library of high-quality AI voices with different accents, ages, and styles. Choose voices that match your subject's appearance for the most natural results.

Upload Your Own: For maximum authenticity, upload your own voice recordings. This is ideal for branded content, personal projects, or when you need a specific voice that matches the subject exactly.

Duration

Maximum length: Audio must be 8 seconds or less. This is the maximum supported duration for Omnia videos.

Pacing: Natural speech patterns work best. Avoid rushing through dialogue; let the character breathe between sentences.

Voice Matching

Match the character: Select a voice that aligns with the subject's apparent age, gender, and personality.

Consider the context: Podcast hosts sound energetic. Narrators are calm and authoritative. Product reviewers are enthusiastic.

Regional accents: If the image suggests a specific region or culture, consider matching the voice accent for authenticity.

Best Practices

Do

  • Keep prompts concise—under 25 words works best
  • Use descriptive verbs: "walking," "gesturing," "adjusting"
  • Match camera style to content type (handheld for UGC, tripod for podcasts)
  • Include environmental context that supports the scene
  • Consider the subject's starting position in the source image

Don't

  • Use vague descriptions like "cool video" or "awesome scene"
  • Request actions that contradict the source image's pose
  • Overcomplicate background elements—one dramatic element is enough
  • Exceed 8 seconds of audio
  • Rush dialogue—natural pacing works better

Quick Reference Templates

Talking Head
Stationary [shot type] with subtle [camera motion] on [subject description], [setting details].

Example: Stationary medium shot with subtle push-in on podcast host gesturing, studio lighting with plants in background.

Social/UGC
Handheld selfie shot, [subject] [action] [emotion], [location], natural movement.

Example: Handheld selfie shot, woman talking excitedly holding coffee, rainy city street, natural movement.

Cinematic
[Camera motion] on [subject], [atmospheric element], [mood descriptor].

Example: Slow orbit around detective, rain falling heavily, noir atmosphere with neon reflections.

Dynamic Background
[Subject] [simple pose/action], [dramatic background element] behind.

Example: Stylish woman crosses street confidently, massive explosion erupts behind her.