Turning Existing Images into Stills [Prompt Included]

Good evening, everyone.
Are you having fun brainstorming with AI today?

(The prompt for turning images into stills is below)

Today's bootcamp is a combination of "mirror reflection," "desolate scenery," and "ripples on the water's surface."

The scenery suitable for this environment is...

That's right, let's go to the Salar de Uyuni.

It's not as easy to get to as Kyoto, but with AI, I should be able to generate an image that looks like the Salar de Uyuni!

The character's clothes have a similar color scheme to the scenery, so I was wondering how it would turn out, but it looks pretty good.

The fact that both the character and the clouds are reflected on the water's surface makes me think, "That's AI for you."

Placing an AI character in a place you'd like to visit yourself makes you feel like you're traveling just a little bit, doesn't it?

Prompt for turning existing images into stills

Recently, more people have been introducing bootcamp articles!

It's nice to have my character appear in note articles!

I thought, "Since I'm at it, why not create a prompt that can process those images to look like game event scenes?" and I made one right away.
(The prompt is at the bottom of the article)

I'm going to process an image generated by Tadapitoyo-san, who always writes articles with delicious bread recipes and photos.

A group photo of Suzushiro, the personality of my AI radio, Hibiki-chan from Itsuki-san's place, and Lunaria and the slime from Tadapiyo-san's place has been turned into an event-style image!

If you tweak the aspect ratio settings, you could probably set it as a smartphone wallpaper. (The default is a vertical 9:16)

Next, I'll process an image generated by Nyankichi-san, who wrote a support collaboration article, into an event style.

AI radio personalities Auko and Suzushiro are visiting Nyankichi-san's article.
Please listen to Suzushiro's "Road to Kindle: AI-damono, Metcha Boke-te ko!"!!

I can't sleep with my feet pointed toward Nyankichi-san.
……I don't know where they live, so I have no choice but to sleep standing up!!
Though I'd be doomed if they lived on the other side of the world in Brazil.

Anyway, the good thing about this prompt is that it automatically adds a frame and gives the image a pose and atmosphere that looks like an event CG, matching the mood of the original image.

It's outrageous that Kiyoshiro is touching Mirai-chan while taking advantage of the chaos.

Since the original image was landscape, they might have just zoomed in to fit it into a portrait aspect ratio.

If it's raining, it makes it a rainy scene, and no matter how many people are in the original image, it turns it into an event-style image, but the gacha element is quite strong.

Since I'm instructing it to "make it look like a game event scene/still," I feel like it does a better job with shots of individual characters.

The last image is the now-customary male version of Auko.

I brought him back because it was funny when I previously instructed the AI to "zoom in on the original image" and it generated a close-up of a very handsome guy.

Event Image (Still) Generation Prompt

How to use:Attach the image you want to turn into a still and run the following prompt.
Features:
・Inherits the taste of the original image (style such as watercolor, thick painting, cel shading, etc.) and the character's appearance.
・Reads the character's expression and atmosphere and changes them to an appropriate pose.
・Can handle 1 to multiple people (up to about 5 seems good).
・Aspect ratio is 9:16; please change it as necessary.
・I had Windows Copilot generate the prompt.

Analyze the attached character reference image and accurately preserve:

- the original art style (line quality, coloring method, shading style, texture, lighting, overall aesthetic),
- each character’s facial features, emotional nuance, personality, and relationship dynamics,
- the atmosphere and tone conveyed by the original image.

Do NOT reuse or replicate the original pose exactly.

Instead, reinterpret the characters with new poses and body language that naturally express their personalities and emotional connection as seen in the reference. Adjust their posture, gestures, and spacing so the scene feels alive and intentionally composed.

Create a fully redrawn 9:16 vertical event-CG-style illustration, as seen in visual novels or otome games. The composition must feel like a polished story moment: refined framing, elegant decorative borders or subtle mask processing, clean depth layering, and a cinematic sense of focus.

If multiple characters are present in the reference, preserve their interpersonal distance, eye contact, and emotional tension, but re-stage them in a way that enhances narrative impact. Their poses should reflect the mood implied by the reference (calm, playful, intimate, tense, etc.).

The background, lighting, and overall scene should be redesigned to match the characters’ vibe and elevate the moment into a dramatic still image. Maintain the original art style while transforming the scene into a cohesive, expressive, high-quality event CG.

Feel free to share your thoughts,
like "Look at the event image I got!"
!!

Well then, see you in the next optimal solution.

Previous Bootcamp Article