AI Video
Witness Horror Cinematic Base Prompt AI Photo Prompt
Photorealistic background-horror style for Gemini Omni: quiet observer camera, one subtle anomaly the character never notices.
The Prompt
witness-horror-cinematic-gemini.md
Witness Horror style. Photorealistic live-action cinematography, naturalistic motivated lighting (practical sources only - window, lamp, streetlight), realistic environment with ordinary clutter, subtle film grain, cinematic depth of field. Camera behaves like a quiet cinematic observer: slow push-ins, long static holds, slow pans that reveal background information. Maintain strict character, wardrobe, environment, and lighting continuity across every shot using the attached reference image(s) when provided - same face, same clothing, same room, same time of day in every frame unless the story specifies a scene change. Location: {{location}} Time / weather: {{time}} Character action: {{character_action}} Exactly one subtle background anomaly: {{anomaly}} Audio: {{audio}} {{negative_prompt}} No narration, no voiceover. Optimized for Google Gemini Omni video generation and matching still frames.
Negative prompt
Included when you copy the prompt
red/blue horror lighting, VHS filter, heavy glitch overlay, jump scare, gore, exaggerated acting, giant monster reveal, evil glowing eyes, demonic face close-up, excessive fog, floating ghost, blood everywhere, random glitching, extreme fisheye, screaming character, generic dark horror cinematic cliché
Fill in your details below - this prompt updates as you type, then copy it.
Fill in your details
0 of 6 filled. Preview updates as you type - then copy the finished prompt (negative prompt included when present).
witness-horror-cinematic-gemini.md
Prompt preview
Witness Horror style. Photorealistic live-action cinematography, naturalistic motivated lighting (practical sources only - window, lamp, streetlight), realistic environment with ordinary clutter, subtle film grain, cinematic depth of field. Camera behaves like a quiet cinematic observer: slow push-ins, long static holds, slow pans that reveal background information. Maintain strict character, wardrobe, environment, and lighting continuity across every shot using the attached reference image(s) when provided - same face, same clothing, same room, same time of day in every frame unless the story specifies a scene change. Location: {{location}} Time / weather: {{time}} Character action: {{character_action}} Exactly one subtle background anomaly: {{anomaly}} Audio: {{audio}} {{negative_prompt}} No narration, no voiceover. Optimized for Google Gemini Omni video generation and matching still frames.
Examples
Example input
Example 1 - Behind Her location: Ordinary apartment interior, kitchen doorway visible in deep background time: Evening, practical lamp and kitchen light only character_action: A woman speaks on the phone and walks toward the kitchen, never looking behind her anomaly: In the deep background, a figure slowly appears in the doorway, stays small in frame, then disappears when she turns audio: Soft phone conversation ambience, room tone, no music sting negative_prompt: red/blue horror lighting, VHS filter, heavy glitch overlay, jump scare, gore, exaggerated acting, giant monster reveal, evil glowing eyes, demonic face close-up, excessive fog, floating ghost, blood everywhere, random glitching, extreme fisheye, screaming character, generic dark horror cinematic cliché Example 2 - The Reflection location: Sterile but realistic bathroom time: Night, fluorescent/practical overhead light character_action: A man washes his hands, looks up at the mirror briefly, then looks away anomaly: His reflection keeps looking down for one extra second, then behaves normally; realistic mirror perspective otherwise audio: Water running, soft bathroom room tone, no score negative_prompt: red/blue horror lighting, VHS filter, heavy glitch overlay, jump scare, gore, exaggerated acting, giant monster reveal, evil glowing eyes, demonic face close-up, excessive fog, floating ghost, blood everywhere, random glitching, extreme fisheye, screaming character, generic dark horror cinematic cliché
What you should get
Example 1: Photoreal apartment clip - viewer sees the doorway figure; she never does; figure gone on her turn. Example 2: Bathroom mirror lag of one second only. One anomaly, motivated lighting, quiet observer camera - Gemini Omni ready.
How to use this AI photo prompt
- Upload a clear reference photo of your face (well-lit, front-facing works best).
- Fill any placeholders on this page, then tap Copy Prompt.
- Paste into Gemini (also works with Runway, Firefly) and generate.
- Keep your face locked - only wardrobe, pose, and scene should change.
Privacy: If you attach a face reference, do it in the AI tool. Codextroop Prompt Library never receives or stores your photos.
Works best with Gemini / Runway / Firefly
Identity-preserving image-to-image prompts need a strong face reference. Avoid heavy beauty filters on the upload and use the negative prompt when your tool supports it.
Tips for better ai video results
- Match the example aspect ratio when your tool allows it (often 4:5 for portraits).
- If the face drifts, re-upload a sharper photo and keep the same prompt text.
- Copy related ai video prompts below to try alternate looks with the same identity lock.
When to use this prompt
Use it when
- You need help with psychological horror shorts.
- You need help with gemini omni video.
- You need help with youtube series.
Skip it when
- You only need a one-line ask with no constraints, format, or context.
- You cannot supply the fill-in details the template asks for.
- You want a modern digital look or heavy filter effect instead of the era/style this prompt targets.
How to get a good result
Done when
Example 1: Photoreal apartment clip - viewer sees the doorway figure; she never does; figure gone on her turn. Example 2: Bathroom mirror lag of one second only. One anomaly, motivated lighting, quiet observer camera - Gemini Omni ready.
Common mistakes
- Leaving location, time, character_action vague - the model then invents placeholders.
- Asking Gemini, Runway, or Firefly for a shorter answer after a long structured prompt - that throws away the format rules.
- Stopping when the first draft appears instead of checking it against the success criteria below.
FAQ
- What do I fill in before I copy this prompt?
- Replace these placeholders with your details: location, time, character_action, anomaly, audio, negative_prompt. Required fields are marked on the page. The prompt text updates as you type.
- Which AI tools is this prompt written for?
- It works best with Gemini, Runway, Firefly. Copy the filled prompt and paste it into the tool you already use.
- What should a good result look like?
- Example 1: Photoreal apartment clip - viewer sees the doorway figure; she never does; figure gone on her turn. Example 2: Bathroom mirror lag of one second only. One anomaly, motivated lighting, quiet observer camera - Gemini Omni ready.
- What are common mistakes with this prompt?
- Leaving location, time, character_action vague - the model then invents placeholders. Asking Gemini, Runway, or Firefly for a shorter answer after a long structured prompt - that throws away the format rules. Stopping when the first draft appears instead of checking it against the success criteria below.
Best for
- Psychological horror shorts
- Gemini Omni video
- YouTube series
Related Prompts
More prompts like this — same topic and style.