Creating a Fantasy Suspense Short Film Using AI Tools: A Step-by-Step Guide

In today’s world, artificial intelligence (AI) has revolutionized the creative process, making it easier than ever for anyone to produce a high-quality short film. From scriptwriting to visual creation and final editing, AI tools can streamline the entire production process.

The Lantern on Maple Street

In this blog, we’ll walk you through the complete process of creating a short fantasy suspense film using various AI tools. We’ll explore how to:

1. Write a Story Script with ChatGPT
2. Convert the Story into a Film Script with ChatGPT
3. Generate Scene Images with MidJourney
4. Create Character Prompts with ChatGPT
5. Produce Scenes with LumaLabs AI
6. Edit the Film with CapCut

Let’s dive in!

Step 1: Writing the Story Script with ChatGPT

The first step in creating a short film is crafting a compelling story. In this case, we used ChatGPT to write a fantasy suspense story based on a simple prompt. The story centers around a mysterious, flickering lantern on a deserted street, creating an eerie atmosphere.

Prompt Used:

We started with a prompt that described a fantasy urban-tale-style adventure. Here’s an example of the prompt used to generate the story:

Please write a fantasy suspense story 3-4 minutes long, some mysterious adventure urban-tale-like story, from a narrator's perspective. The story should be simple and anecdotal, with a conversational tone. Keep the narrative easy to follow with a deep sense of involvement. Use conversations between the characters to show their feelings and move the story forward. The story should teach us something about life but don't tell us the lesson right away, let us figure it out as we read.

Based on this, ChatGPT generated a story about a narrator who stumbles upon an old lantern that flickers mysteriously, leading him to a strange conversation with an older man. The man shares a haunting tale about the lantern's power, which compels those who see it flicker to confront their life choices.

Story Summary:

The narrator encounters a broken lantern on Maple Street, which seems to flicker at random. He meets a mysterious older man who reveals that the lantern only lights up for those with unresolved decisions in their lives. The narrator learns that if you don’t act in time, the lantern will make the decision for you, often with dire consequences.

Step 2: Converting the Story into a Film Script with ChatGPT

Once we had the story, we needed to turn it into a screenplay format, suitable for film production. We used ChatGPT to convert the narrative into a film-like script, with detailed scene descriptions, character actions, and dialogue.

Film Script Example:

For instance, in one scene, the narrator stands in front of the flickering lantern, unsure whether to believe its supernatural power. ChatGPT transformed this moment into a screenplay format like this:

EXT. MAPLE STREET - NIGHT

The old lantern flickers faintly between two decaying buildings. The shadows stretch long as the sun sets behind the brick structures.

NARRATOR
(quietly, to himself)
"Must be a trick of the light."

He steps closer to the lantern, his eyes narrowing. The orange glow pulses weakly.

NARRATOR
(aloud)
"That's odd."

MYSTERIOUS MAN
(voice from behind)
"Everything’s odd around here."

Why This Is Important:

The screenplay format makes it easier to visualize the story's structure, character movements, and key moments, which are crucial for the next steps in the filmmaking process.

Step 3: Generating Scene Images with MidJourney

With the script in hand, we needed to visualize the settings, characters, and mood of each scene. To do this, we used MidJourney, an AI-based image generator, to create stunning scene images based on detailed prompts.

MidJourney Image Prompts:

For example, in the scene where the narrator walks down the narrow, empty street, we provided a detailed prompt:

Create an Anime Ghibli Style wide shot of a narrow, desolate street with an old, crooked lantern hanging between two decaying, weathered buildings. The lantern's glass panes are cracked, and the surrounding area is gloomy. The light is fading, with the evening sun casting long shadows. The street feels abandoned. The point of view is from a distance, and the lighting is dim and cold, creating a sense of neglect. Ensure the background shows crumbling brickwork and cracked pavement.

MidJourney generated an image that perfectly captured the eerie, abandoned street. These images helped bring the world of the story to life visually, allowing us to immerse ourselves in the setting and build the atmosphere for each scene.

Step 4: Creating Character Prompts with ChatGPT

Next, we needed to design the characters based on the story’s descriptions. We used ChatGPT to create detailed character profiles for the narrator and the mysterious man.

Character Example:

Here’s how we described the narrator and the mysterious man:

• Narrator: A young man in his early 30s, with short dark hair, wearing casual urban clothes (a hoodie, jeans, and sneakers). He has a thoughtful expression.
• Mysterious Man: An older gentleman, in his mid-50s, wiry with a thin, angular face, gray hair, and worn-out clothes. He has a weathered, lived-in look.

Character Visuals:

Using these descriptions, we generated consistent visuals for the characters throughout the film, ensuring they remained true to their appearances as the story progressed.

Step 5: Producing Scenes with LumaLabs AI

Once the visual aspects were set, it was time to bring the scenes to life with LumaLabs AI, a tool that generates dynamic video clips based on written descriptions.

Creating Video Clips:

For a pivotal scene where the lantern flickers and then goes dark, we used LumaLabs to generate a video clip based on a detailed prompt:

Generate a video clip of a flickering lantern, with a faint orange glow pulsing weakly inside cracked glass. The lantern dims, and then goes completely dark. The surrounding street is empty, and the atmosphere is eerie and tense.

LumaLabs AI generated a realistic video that captured the flickering effect of the lantern, adding motion to the static images we had previously created with MidJourney.

Step 6: Editing the Film with CapCut

Finally, after generating the scenes and visuals, it was time to assemble everything into a cohesive film. We used CapCut, a free AI-powered video editor, to combine the generated videos, images, and audio.

Editing Process:

• Importing Assets: We uploaded the AI-generated images and video clips from MidJourney and LumaLabs into CapCut.
• Assembling the Film: Using CapCut’s intuitive interface, we arranged the clips according to the screenplay, adjusted the timing, added transitions, and included sound effects to enhance the atmosphere.
• AI Features: CapCut’s AI tools helped us generate voiceovers, add automated subtitles, and fine-tune the audio and visuals, making the final edit smooth and professional.

Conclusion: Bringing the Story to Life with AI

By combining the power of AI tools like ChatGPT, MidJourney, LumaLabs, and CapCut, we were able to transform a simple story idea into a full-fledged short film. The process involved creating a script, designing characters, generating scenes, and editing the footage into a coherent narrative, all with the help of AI.

The result is a seamless, high-quality short film created with minimal resources and maximum creativity. Whether you’re a seasoned filmmaker or just getting started, AI tools can help you bring your stories to life with ease and efficiency.

Call to Action

Have you tried creating a film using AI tools? Share your thoughts, experiences, or questions in the comments below! We’d love to see what you create.

This blog post outlines the entire process, showcasing how various AI tools work together to streamline filmmaking. Let me know if you'd like to make any adjustments or if there's anything else you'd like to add!