AI Spot – Entirely Generated with Artificial Intelligence
AI Spot – Entirely Generated with Artificial IntelligenceAI Spot – Entirely Generated with Artificial IntelligenceAI Spot – Entirely Generated with Artificial IntelligenceCustomer
BigFive
Activity
Video and Animations
Cinematic AI Spot
The creative idea
A micro thriller in 30 seconds: dry dialogues, suspicion, and a final twist which reveals the true theme of the video.
We created the characters to be consistent throughout the video, defined every shot, and forced the consistency of the voices to make it seem real until the very end.
Production made with the most advanced AI models for images, videos and audio available in 2026.
How did we do it?
The project started from a narrative idea and a precise script, built like a micro-thriller.
We defined the rhythm, dialogue and structure before generating any images.
Only after that did we move on to creating the characters and designing the shots.
AI was the tool, but the direction and creativity is completely human.
AI Actor Casting
The first step was not to create characters consistent with generations.
As if it were a casting, we designed the characters' faces, styling and features to be staged using Higgfields Cinema Studio 2 and Nano Banana Pro.
Only after having validated the visual integrity of the protagonists We reworked the characters with Photoshop to reduce imperfections and eliminate inconsistencies, continuing to build the story with these character as a reference in the generation of images and videos.
Upscale + Skin enhancer
Once we have selected the best shots for our protagonists, We performed an image upscaling with Topaz AI Image Upscale to have very high resolution images to work on.
Subsequently, to avoid the “rubbery skin” effect typical of many AIs, we reprocessed the images using Higgfields AI's Skin enhancer tool.
Choice of frames
We worked on the shots as if on a real film set.
Each scene was designed before being generated: shot, reverse shot, close-ups, coverage.
Using Higgfields Cinema Studio 2 we produced multiple variations of each frame, selecting only those that maintained continuity of light, perspective and position.
The generation was not linear, but iterative: test, discard, refinement.
Only in this way does AI stop being random and become directorial.
Video Generation
There is not just one generation for each scene.
We produced multiple variants of the same time with Kling 3.0 comparing micro-differences in expression, acting, focus and posture.
In this example you can see a comparison between different clips of the same line being delivered.
AI creates possibilities, but consistency comes from direction and vision.
That's where the real work happens.
Soundtrack and editing
The background music was generated using SUNO's AI, starting from a prompt designed to sustain narrative tension.
The editing was done in Adobe Premiere, where we inserted subtitles, balanced the volumes, added sound effects and defined the pacing.
The final output has been meticulously crafted to ensure audio consistency, visual quality, and overall impact.
What videos can we make with AI?
Generative technologies allow us to transform static images into cinematic content and create visuals that were impossible with traditional methods.
But technology alone is not enough. Without a clear idea and a precise direction, AI remains only a technical tool: the result depends on how it is driven.
Thanks to our experience in directing, photography and digital production,, We integrate these tools into a structured creative process, transforming them into concrete, effective content that is consistent with the identity of each project.
Tags
Video
Customer
BigFive
Year
