TokiToki Internal
Atlas + S3
All runs
CompletedComfyUI

f45b221b-3ec9-46a4-8a0a-9791928cd760

Queued
Aug 29, 2026, 06:06 PM
Completed
Aug 29, 2026, 06:07 PM
Execution time
11.4s
Outputs
1 · 1.3 MB
Inputs
1
Cached nodes
1

Outputs

1

Click an image to view it full size.

character_ref_00790_.png

node 3 · 1.3 MB

Text prompts

2
Promptnode 1 · PrimitiveStringMultiline · valueCharacter Description (your input)

Use the uploaded image as the **identity reference for Basanthi**. Create a new version of the same woman, making her look **noticeably but subtly younger — approximately 3 years younger** than the uploaded reference. This is a **controlled age-regression edit**, not a recreation. ### KEEP HER IDENTITY EXACTLY THE SAME Her identity must remain unmistakable. Keep the same: * facial bone structure * face shape * eye shape and spacing * eyebrows * nose * lips * jaw and chin * natural facial proportions * skin tone * hairstyle * hairline * red flowers * bindi * jewelry * sari and blouse * overall character design ### MAKE THE AGE DIFFERENCE VISIBLE The result should genuinely look like a photograph of **Basanthi approximately 3 years earlier**. Compared with the reference, make these changes subtly: * skin should have **slightly more youthful elasticity** * reduce fine lines around the eyes by approximately 15–20% * slightly reduce the depth of nasolabial folds * slightly reduce subtle lines around the mouth * make the under-eye area **a little fuller and fresher** * very slightly soften age-related skin texture * give the cheeks a **small amount of natural youthful fullness** * make the overall complexion look subtly fresher * retain natural pores and realistic skin texture These changes should be **small but clearly perceptible when the two images are placed side by side**. ### IMPORTANT — DO NOT OVER-REJUVENATE She should NOT look dramatically younger. Do NOT make her look 10 years younger. Do NOT change her facial structure. Do NOT enlarge her eyes. Do NOT make her face thinner. Do NOT make her look like a younger actress or a different person. The target age difference is **ONLY ABOUT 3 YEARS**. Think of the transformation as: **Reference: Basanthi at age X → Output: Basanthi at approximately X−3 years.** ### NATURAL SKIN — NO AI BEAUTIFICATION Keep authentic skin. Visible natural pores, fine texture, tiny imperfections, subtle asymmetry and realistic skin variation must remain. **Do not smooth the skin excessively.** No: * beauty filter * airbrushing * plastic skin * porcelain skin * glossy skin * artificial glow * glamour retouching * excessive sharpening * perfect skin * face reshaping The output should look like a **real photograph of the same woman taken three years earlier**, not an AI beauty portrait. ### KEEP THE PHOTOGRAPH OTHERWISE CONSISTENT Keep the same: * straight-on eye-level camera * direct eye contact * neutral expression * head position * framing * gray background * lighting * hairstyle * clothing * jewelry **ONLY make the subtle 3-year age regression.** ### FINAL PRIORITY **Identity must remain 100% recognizable. Age must visibly decrease by approximately 3 years. Everything else should remain unchanged.** Do not interpret "younger" as "more beautiful." Do not beautify her. The desired result is simply: **THE SAME BASANTHI, ABOUT 3 YEARS YOUNGER, WITH NATURAL UNRETOUCHED SKIN.**

Systemnode 2 · GeminiImageDirect · system_promptCharacter Image Generator (Nano Banana)

Generate a single, hyper-realistic character reference image (strictly no collages, text, watermarks, or borders). The image must feature a cinematic, three-quarter portrait of the described character, maintaining a sharp, tack-focused emphasis on the face with natural skin textures, pore details, and realistic micro-expressions. The lighting should be cinematic and directional (e.g., subtle split or Rembrandt lighting) to give depth, blending seamlessly with the environment specified in the character description. The character must have a natural, living presence, looking slightly off-camera or directly at the lens with a subtle, context-aware expression. Ensure the full head, neck, and upper shoulders are completely unobstructed. This image serves as a high-fidelity identity blueprint for temporal-consistent video generation.

Input media

1

Source images and video this run consumed.

character_ref_00788_.png

node 4 · 1.3 MB