Altera Friday Workshop: AI Composition for Cinema, YouTube & Vertical
Most people generate an image, then crop it for wherever it’s going. That’s backwards, and it shows. In this session, Elizabeth Kealoha teaches you to compose for the screen before you ever hit generate, so your work actually holds together across cinema, YouTube, and vertical formats instead of getting cropped into something that doesn’t read.
You’ll build one story as three keyframes, open, peak, and resolution, and learn the five decisions that shape a frame before you write the prompt: where it plays, which direction the story moves, where to leave room to grow, how to size your subject, and what stays locked across every frame.
Who it’s for
Anyone generating images or video who wants their work to read clearly no matter where it ends up, cinema, YouTube, or vertical. No prior experience needed.
What you’ll leave with
The five decisions that shape a strong first frame, before you touch the prompt
How to compose the same story differently across 21:9, 16:9, and 9:16
Three keyframes that tell one complete story, built live in the room
A working sense of what breaks when a frame gets cropped for the wrong screen, and how to avoid it
Before you come
Bring your own image generator with an aspect-ratio setting, we don’t provide tools or credits. Please don't forget your laptop.
Schedule
12:00 Welcome, tool check, asset link
12:05 Tool stack and prompt recipe recap
12:15 What’s wrong with this frame?
12:22 The five decisions for frame one
12:30 Built live, together
12:39 Build your own, three keyframes
12:50 Shares
1:00 Close
About Altera
Altera is a creative innovation ecosystem in Atlanta where filmmakers, technologists, and companies explore emerging tools and build the future of storytelling. We teach this work and we do this work. If you want to go deeper, our workshop series runs on from here. If you’d rather have it made for you, our studio produces generative campaigns for brands.
Lunch not provided. BYOL