Interactive tool
Horror Scene Studio: image and video prompt builder
A privacy-first creative tool for horror portraits. Grade a photo locally in your browser, then generate a structured image prompt and an image-to-video prompt for WaveSpeed, SeeAPI, fal, Replicate or Kie.
At a glance
Quick answer
Horror Scene Studio is a free, browser-based tool for making an original horror portrait. Choose a photo and adjust grain, vignette, fog, contrast and warmth; the image is processed locally and can be downloaded as a PNG. The same canvas can export a five-second WebM clip with a slow push-in. The prompt builder turns subject, setting, lighting, camera, motion and aspect ratio into a structured image prompt and an image-to-video prompt for providers such as WaveSpeedAI, SeeAPI, fal, Replicate and Kie. It does not upload your photo, call a model or charge per generation. The game's characters and cover art belong to catproblem9735, so use the tool for original scenes rather than reproducing Idimya. Provider links are informational and this site has no affiliate relationship; check current model names and prices before spending. It is privacy-first.
How it works
A prompt studio, not a hidden paid API
Most "AI horror portrait" pages are a file upload wired to a paid model. That is useful, but it hides the cost, the provider and the privacy trade-off. This page takes the opposite approach: the photo editor runs entirely in your browser, and the prompt builder gives you a clean, structured prompt to paste into the model you already use.
Local editor
Grade a photo in your browser
Local photo editor
Choose a photo and grade it in your browser. The image never leaves this page: there is no upload, account or model call.
Your edited preview appears here.
Choose a JPG, PNG or WebP to begin.
AI image and video prompt builder
Build a structured prompt for an original horror scene. Copy it into the provider you use; this page does not send the prompt anywhere.
Provider links open external sites. Check current model names and prices before you generate.
Providers
Where to take the prompt
The providers below are useful for different jobs. Model catalogs and prices change, so check the current model page before you spend anything. This site has no affiliate relationship with them.
| Provider | Models to try | Use it for |
|---|---|---|
| WaveSpeedAI | Wan 2.x image-to-video, Seedream edit, FLUX Kontext, Kling | A broad image plus video catalog and low-cost short clips |
| SeeAPI | Whatever image and video models are exposed in its catalog | One API surface for several model families; check the current list |
| fal | FLUX, Kling, Wan and other hosted models | Fast serverless calls when you already have a fal workflow |
| Replicate | FLUX, Kling, Wan and community models | Experimenting with many models from one account |
| Kie | Nano Banana, Kling and other image/video models | A simple API for image editing and image-to-video |
Provider links: WaveSpeedAI - SeeAPI - fal - Replicate - Kie.
Copy a recipe
Image prompt template
Image editing prompt
Use the uploaded photo as the primary reference. Preserve the subject's recognizable facial identity, facial proportions, skin tone and natural anatomy. Transform the subject into an original horror portrait built from geometric color blocks and a palette of umber, amber and teal. Change only the wardrobe, hairstyle and setting; do not change the face. Place the subject in a moonlit Victorian garden with soft fog and warm lanterns. Realistic editorial photography, believable fabric texture and cinematic soft lighting. Keep the same number of people as the uploaded photo. Avoid text, logos, watermarks, extra limbs and distorted hands.Animate it
Image-to-video template
Image-to-video prompt
Use the provided image as the exact visual reference and starting composition. Create a continuous five-second cinematic animation. Keep the subject's face, costume, hairstyle and body proportions consistent with the reference. The subject slowly turns their head toward the camera with a subtle mysterious expression. A gentle breeze moves loose hair strands and the edge of the clothing. Soft fog drifts through the garden. The camera performs a slow and smooth push-in. Maintain realistic motion, stable anatomy and consistent lighting. Do not change the person's identity, replace the costume, add characters or introduce sudden scene cuts. No text or watermarks.Workflow
A four-step horror portrait workflow
-
Start with a clean portrait
Use a photo with one person, even lighting and a visible face. The editor works best on a simple background.
-
Grade it for the mood
Add contrast and warmth first, then a vignette, then fog. Keep the grain low if you plan to animate; heavy grain can shimmer.
-
Generate the still
Paste the image prompt into your provider. If the face drifts, lower the change strength or use a dedicated identity-preserving edit model.
-
Animate the still
Use the image-to-video prompt with a short duration and subtle motion. Avoid fast camera moves and scene cuts on the first attempt.
Questions
Scene Studio FAQ
Does this page upload my photo?
No. The photo editor uses a canvas in your browser. The image is not sent to this site or to a model provider.
Does the prompt builder call an AI model?
No. It generates a structured text prompt. You copy the prompt into WaveSpeed, SeeAPI, fal, Replicate, Kie or another provider yourself.
Can I use this to make art of Idimya?
Do not reproduce the game's character or cover art. The game and its art belong to the creator. Use the visual language for original scenes instead.
Which provider is best?
It depends on the model you need. WaveSpeed and Kie are convenient for image-to-video; fal and Replicate are good for experimenting with several model families. Check current model names and prices before you generate.