Cinelca Stage

AI has made it dramatically easier to generate images, video, music, and sound effects. But actually directing those elements still happens through flat timelines, sliders, menus, and layers.

We started with a simple question:

What if a foldable phone could become a tiny production stage?

The iPhone Duo felt uniquely suited to this idea because it gives us two connected surfaces and a physical hinge. Instead of treating the second screen as more screen real estate, we wanted to use the form factor itself as part of the creative interface.

That became Cinelca Stage: a spatial canvas for filmmakers and creators where media can be arranged and manipulated as if it exists on a miniature stage.

The long-term idea is broader than audio: a creator could position characters, cameras, lights, sound sources, or generated elements spatially, then use that arrangement to drive editing or generation tools.

For the hackathon, we focused on making that idea tangible through sound.

What it does

Cinelca Stage turns the unfolded iPhone Duo into a compact spatial sound workspace.

One display acts as the primary media and control surface, while the second becomes a visual 3D stage.

A creator can:

• load or preview a scene
• place sound sources in space
• move them relative to the listener or camera
• adjust properties such as position and intensity
• preview how the scene changes as those sounds move
• use the Duo's physical layout to make spatial editing feel more direct and intuitive

Instead of thinking about sound as tracks in a timeline, you can think about it as objects in a scene.

For example, a creator could move dialogue to the center, place traffic behind and to the left, and position ambience around the listener — all visually.

The same interaction model could eventually extend to visual scene editing, AI video generation, virtual production, and spatial storytelling.

How we built it

We built Cinelca Stage as a native iPhone Duo prototype using Apple's development stack and Bitrig.

The interface is designed around the Duo rather than simply stretching a conventional phone application across a larger display.

The core experience is split into two coordinated surfaces:

1. Media / control surface
   Used for scene playback, selection, and editing controls.

2. Spatial stage
   A visual representation of the scene where sound sources can be positioned and manipulated.

We created a lightweight scene model that represents each sound as an object with spatial properties such as:

[ S_i = (x_i, y_i, z_i, g_i) ]

where:

• (x_i, y_i, z_i) represent the source position
• (g_i) represents gain or intensity

The UI maps gestures on the stage to these parameters, allowing the user to manipulate the mix spatially instead of entering numerical values.

We also designed the app so that the two displays communicate continuously: selecting or moving an object on the stage updates the corresponding controls and scene state on the other display.

A major goal was to keep the demo lightweight enough to build during the hackathon while still demonstrating a larger interaction model.

Challenges we faced

Designing for a new form factor

The biggest challenge was avoiding the temptation to build a normal iPhone app and simply make it wider.

The hinge and dual-screen configuration needed to provide an interaction that would be meaningfully worse on a conventional slab phone.

That pushed us toward the idea of a control surface + physical stage, where the geometry of the device becomes part of the interface.

Scope

Spatial media production can become extremely complex very quickly.

A complete system could involve multichannel audio, head tracking, object-based audio formats, AI source separation, video segmentation, 3D environments, and rendering.

That was far too much for a hackathon.

We deliberately reduced the idea to its core interaction:

select an element → place it in space → immediately understand the result.

Connecting 2D gestures to 3D concepts

Another challenge was making spatial positioning understandable without requiring professional audio knowledge.

The interface needed to communicate front/back, left/right, distance, and intensity while remaining simple enough to understand in seconds during a demo.

Building in parallel

We developed the prototype across two machines with parallel development workflows. That helped us move quickly, but it also forced us to keep components modular and avoid having multiple people modify the same parts of the project at once.

What we learned

The biggest thing we learned is that foldables are most interesting when the physical configuration of the device changes the interaction model, not simply the amount of available screen space.

The Duo can behave almost like a small piece of production hardware:

• one side can be a monitor
• one side can be a control surface
• the hinge can establish a spatial relationship between them
• orientation can change the working mode

We also learned that spatial interfaces can make complex media concepts easier to understand.

Most people understand “move this sound behind the character” more naturally than they understand pan values, automation curves, or spatial-audio parameters.

That suggests a broader direction for Cinelca Stage: using spatial interaction as a simpler front end for increasingly complex AI and media-generation systems.

What's next

The sound-stage prototype is only the first version.

The same stage could eventually represent:

• actors and characters
• virtual cameras
• lighting
• generated objects
• dialogue and sound effects
• camera movement
• AI video generation controls

A creator could arrange a scene visually on the Duo and then use that stage representation to generate or modify the final media.

Our broader vision for Cinelca Stage is a portable spatial control surface for AI filmmaking — where creators direct scenes by manipulating the scene itself rather than navigating increasingly complicated menus and timelines.

Built With

• bitrig