Dark Cinematic Fantasy Portrait

Overview

This template transforms an existing mirror selfie into a dark, cinematic fantasy scene while preserving the subject exactly as captured. The person’s face, pose, outfit, lighting, background, and camera angle remain completely unchanged.

A shadowy demonic presence is introduced as a realistic smoke-based entity emerging from behind the subject, designed to feel physically present and integrated into the environment rather than overlaid. The result is a moody, high-contrast cinematic image that blends grounded realism with subtle supernatural tension.

This template is ideal for fantasy edits, cinematic profile visuals, or dramatic storytelling images where identity preservation and realism are non-negotiable.

Related Tags

#Cinematic#darksmoke

Image Prompt1 Recommended Attachment

Recommended Attachments

Mirror Selfie

The original image defining facial identity, pose, lighting, background, and camera angle.

Copy the Prompt

Click the "Copy" button above to copy the complete prompt text to your clipboard. This will copy all the text including any placeholders.

The prompt will be copied exactly as shown

Prepare Your Attachments

Important

Optional: "Mirror Selfie" (image). Have these files ready before proceeding to the AI tool.

Use high-quality images for better results

Ensure files are in supported formats (JPG, PNG, WebP for images)

Open Gemini

Go to gemini.google.com and sign in to your Google account. Make sure you have access to the image generation feature.

Use Google Chrome for the best experience

Ensure you're signed in with a Google account

Gemini Advanced (Pro) provides better image quality

Configure Settings

In Gemini, look for the "Create image" or image generation option. If available, switch to Gemini Pro or Advanced mode for higher quality outputs.

Gemini Pro/Advanced produces more detailed images

Some features may require a paid subscription

Check the model selection if multiple options are available

Upload Your Attachments

Important

Before or along with your prompt, upload the required files. Most AI tools have an attachment/upload button near the prompt input area.

Look for a paperclip or "+" icon to attach files

Wait for uploads to complete before sending

You can usually preview attached files before generating

Paste & Generate

Paste the copied prompt (Ctrl/Cmd + V) into the AI tool's text input field. Make sure your attachments are uploaded, then click the send/generate button to create your image.

Double-check the prompt for any unfilled placeholders

Verify all required attachments are attached

Generation may take a few seconds to a few minutes

Review & Refine

Review the generated image. If it doesn't match your expectations, try regenerating or adjusting the prompt with more specific details. Most AI tools allow multiple generation attempts.

Add more descriptive details if results are too generic

Specify style, mood, or composition for better control

Try different variations to find the best result

Save outputs you like immediately

Advanced Pro Tips

Save prompts that work well for you by adding them to your Wishlist.
Experiment with different AI tools — each produces radically different compositions and intrinsic visual styles.
Be extremely specific with lighting, camera angles, and color grading for high-fidelity yields.
Upload pristine, high-resolution original references. The AI heavily samples the texture quality of your inputs.

Attachments Needed

Mirror Selfieimage

The original image defining facial identity, pose, lighting, background, and camera angle.

Prompt

Use the uploaded mirror selfie as the locked base image. Identity & camera lock (absolute): Keep the person exactly the same No changes to face, facial features, skin, expression, or hair No changes to pose, body position, outfit, or accessories No changes to lighting direction, intensity, or color No changes to background or camera angle The original selfie must remain perfectly recognizable Fantasy entity: Add a shadowy demonic entity emerging from behind the person’s back Entity is made entirely of swirling black smoke Semi-transparent with layered volumetric depth Glowing white eyes, subtle but clearly visible No solid body parts, no exaggerated monster anatomy Feels ominous, restrained, and cinematic rather than cartoonish Smoke behavior: Smoke wraps naturally around the subject’s body without obscuring key features Flows downward to the floor and outward into the environment Interacts realistically with gravity, air, and space Smoke blends seamlessly with the existing background Lighting & realism: Dark, moody cinematic atmosphere High contrast while preserving detail Smoke reacts accurately to existing light sources Realistic shadows, reflections, and depth Entity must feel physically present in the scene, not pasted on Quality & finish: Ultra-realistic rendering 4K-level detail and clarity Natural texture preservation No distortion, no artifacts, no visual glitches No text, watermark, or UI elements

Create Image

Try on OpenArt Try on Higgsfield