==================================================
The target is:
Approximately 3–4 natural spoken lines per 10-second dialogue clip.
Use natural conversational Telugu.
Allow tiny pauses.
Do not create extremely fast speech.
Do not cram too many words.
The dialogue must be:
Emotionally meaningful
Natural
Speakable
Approximately 10-second compatible
Relevant to the visual action
If 3–4 lines cannot naturally fit because the scene is highly visual:
Use fewer lines.
Never sacrifice realism just to force 4 lines.
DIALOGUE LENGTH CONTROL
==================================================
For each clip internally estimate:
Dialogue duration ≈ 8–10 seconds maximum.
Leave a small amount of time for:
Facial reaction
Breathing
Movement
Natural pause
Never write dialogue that requires more than 10 seconds.
If the dialogue is too long:
Automatically shorten it while preserving the meaning.
If the dialogue is too short:
Expand it naturally with emotionally relevant lines.
DIALOGUE STYLE
==================================================
Dialogue must sound like real Telugu people speaking.
Avoid:
Bookish Telugu
Robotic Telugu
Overly formal Telugu
Repeated words
Unnatural dramatic speeches
Excessively long sentences
Use conversational Telugu.
Characters should speak according to:
Age
Relationship
Education
Background
Location
Emotion
Personality
VIDEO PROMPT CONTENT
==================================================
Every 10-second Google Omni prompt must include:
Character
Uploaded Character Reference
Character Identity
Location
Background
Time of Day
Weather if relevant
Action
Emotion
Body Language
Exact Telugu Dialogue
Camera
Camera Movement
Lens
Framing
Lighting
Visual Style
Ambient Audio
Sound Effects
Lip Synchronization
Continuity
VIDEO PROMPT FORMAT
==================================================
For every clip provide:
CLIP [NUMBER]
Duration:
10 seconds
Dialogue:
[Character Name]:
"[Telugu dialogue — approximately 3–4 natural lines]"
Then:
GOOGLE OMNI VIDEO PROMPT
The actual prompt MUST be inside its own separate code block.
Example:
Generate exactly one continuous 10-second cinematic video clip using the uploaded character reference image as the exact visual identity reference for [CHARACTER NAME]. Preserve the same face, hairstyle, age, skin tone, body proportions, clothing and accessories. [Character] is in [LOCATION] during [TIME OF DAY]. [Specific action and emotional performance]. [Character Name] speaks naturally in Telugu: "[EXACT TELUGU DIALOGUE]." Accurate Telugu lip synchronization, realistic facial expressions, natural body movement, [CAMERA SHOT], [CAMERA MOVEMENT], [LENS], cinematic [LIGHTING], consistent [SELECTED VISUAL STYLE], detailed environment, ambient [SOUND], appropriate sound effects. Begin from the exact ending state of the previous clip and end in a natural position that continues directly into the next clip. Generate exactly one continuous 10-second clip.
==================================================
VIDEO NEGATIVE PROMPT
==================================================
Provide a separate copy-ready negative prompt.
No character identity change, no face morphing, no hairstyle change, no age change, no random clothing change, no body distortion, no extra fingers, no extra limbs, no duplicated characters, no merged characters, no unnatural movement, no random objects, no background inconsistency, no lighting inconsistency, no flickering, no visual glitches, no subtitles, no text, no watermark, no logo.
==================================================
CHARACTER REFERENCE USAGE
==================================================
The user will upload the generated character reference image to Google Omni.
Every prompt must say:
"Use the uploaded character reference image as the exact visual identity reference."
