Recreating Call Me by Your Name's Fireplace Scene With AI
Aug 2026
The fireplace scene in Call Me by Your Name was one of the scenes that I just couldn’t forget after seeing it. I wanted to recreate the blocking, framing, and camera movement that Guadagnino brilliantly implements here. So here’s how I did it, for free, with AI.
If you haven't seen it in a while: the film's final shot is one unbroken take. Elio sits in front of the fire, the family sets the table behind him, the credits roll over his face, and nothing else moves. That's exactly what makes it the perfect first scene to rebuild: one locked camera, one subject, one light source doing all the acting.
- Find the ideal stills from classic film scenes.
- Edit the image with Nano Banana Pro to create a realistic human.
- Generate the video.
- Detach the video and audio, then rebuild the sound.
Step 1: Choose your frame
Find the right still. Find a frame where the camera isn't moving, the subject is alone in the composition, and the light has one clear source.
One thing to be clear about: this frame is a reference, not an asset. It never appears in your output. It's there so you can answer to it, the way a DP answers to a lighting diagram.
Rebuild your favorite scene
One still, one new face, one short clip.
Step 2: Rebuild the still with a new person in it
Now recreate it. In Flick this is Nano Banana Pro edit work against your reference. Describe your person, then correct one thing at a time the way you'd direct a set: I changed the sofa to blue, and swapped out my protaganist.
Step 3: Generate the video
Run image to video with your finished still pinned as the first frame, and write the motion prompt as a list of refusals. Camera locked. No zoom, no drift. The fire is the only real motion. The person breathes; they don't act. Five seconds, one take.
This restraint is the entire trick. The original scene is famous because nothing moves except what the fire does to a face. Every extra motion the model invents (a head turn, a camera push, flicker in the room light) breaks the spell, so budget motion like money and spend almost none of it.
Step 4: Rebuild the sound
Generated clips rarely land audio timing, and this scene lives on its sound: fire crackle, room tone, the family somewhere behind. Detach the generated audio, then rebuild it in layers and sync the crackle to the visible flicker by hand, even roughly. Score it with your own music; the original's needle drop belongs to the original. Five seconds of picture with honest sound reads as cinema. The same trick is step four of the method Zoey published with her original fireplace recreation back in February.
The result
Look at the frame again. Long dark hair, red sweater, blue sofa, the hearth burning at the edge of a cool, book-lined room. Not one pixel comes from the film, and you knew what it was before you read a caption. That makes for a great homage.
When you want a second scene, the same four steps produced remixes of In the Mood for Love and the Atonement running shot. Pick a scene with a locked camera first; save the tracking shots for your third.
Frequently Asked Questions
Can AI recreate a scene from a real movie?
As a rebuilt homage, yes: same composition, lighting, and camera grammar, with a new person in the frame. That's the four-step method behind the published Flick TV recreation in this post.
What tools do I need?
The clip used Nano Banana Pro for the still edits and Flick's image-to-video models for motion, all on one canvas. The still carries the scene; the video pass only animates the fire.
Why does the fireplace scene work so well for this?
Locked camera, single subject, one motivated light source. Those three properties are exactly what image models reproduce reliably, which makes this scene the best first project before you attempt anything with camera movement.
Why not just prompt the director's name?
Describe the aesthetics instead: warm single-source firelight, cool Italian interior, film grain, a held static frame. Models execute described light better than name-drops, and style-by-name requests for living directors are where content filters get twitchy.