Tools

Adobe's experimental tools transform photos into navigable 3D scenes and let creators interview AI characters

Adobe MAX 2026 showcases Project Screen Test, Project Sight Span, and Project Sound Bender—tools for character development, spatial image navigation, and audio customization.

Last verified:

Adobe MAX 2026 unveils three conversational AI experiments for creative workflows

Adobe demonstrated three experimental AI tools at its MAX conference on November 11, each designed to expand creative control in domains where AI generation has traditionally offered limited directorial power. According to The Verge AI, the tools—Project Screen Test, Project Sight Span, and Project Sound Bender—prioritize iterative conversation and user guidance over single-pass generation, marking a shift in how Adobe approaches generative creativity.

Project Screen Test: audition-driven character development

The most distinctive of the three is Project Screen Test, which treats AI character creation as a conversational performance rather than a text-prompt exercise. Creators begin by defining base personality traits (such as “anxious, shy, awkward”) and selecting or describing a voice, either through presets, audio prompts, or reference samples. The tool then extends character depth through FaceTime-like dialogue, allowing creators to refine backstory, temperament, goals, flaws, fears, and movement patterns through natural conversation. Once satisfied, users can generate animated screen tests by describing an environment or scenario.

According to The Verge AI, the resulting characters exhibit more recognizable and consistent personality than typical AI-generated animations, addressing a core limitation of current video generation tools—their reliance on deterministic prompts offers far less nuance than traditional frame-by-frame animation. The feature currently exists as a standalone creative exercise, with no announced integration into Adobe’s existing video suite.

Project Sight Span: spatial image expansion

Project Sight Span converts flat images—photographs, paintings, and illustrations—into navigable 3D environments by generating plausible out-of-frame content. The tool preserves artistic style and recognizable details (including tattoos and facial features in photographs) while allowing aggressive repositioning and camera angle changes. The Verge AI notes that generated content remains pixelated during navigation but restores resolution when users finalize their chosen angle. The feature has similarities to Apple’s Spatial Reframing, though with greater flexibility in repositioning; critics may argue the result is “far less deserving of being called a photo” once heavily manipulated.

Project Sound Bender and the audio customization paradigm

Project Sound Bender introduces a cursor-based audio mixing interface where users upload a video and describe desired sound effects via text prompt (such as “bubbling sci-fi cauldron noises”). The tool visualizes audio customization as a color-wheel interaction, enabling intuitive sound design without traditional mixing expertise.

Why This Matters

These three tools suggest Adobe is betting on conversation and iterative refinement as the next phase of generative creative workflows. Rather than perfecting single-prompt generation, Adobe is building tools where users guide AI through dialogue (Screen Test), spatial navigation (Sight Span), and intuitive parameter adjustment (Sound Bender). This mirrors broader industry trends toward human-in-the-loop AI rather than fully automated systems. For video creators, character designers, and audio engineers, the question is whether these experiments graduate to production tools—and whether the conversational overhead justifies the improved creative control relative to faster, traditional generative approaches. Adobe has not committed to integrating any of these into its main product suite, leaving their commercial viability unresolved.

Frequently Asked Questions

What is Project Screen Test?

A conversational tool that lets creators develop AI characters by talking to them in an audition-like environment, defining personality, backstory, and movement quirks through dialogue rather than text prompts alone.

How does Project Sight Span work?

It transforms flat photographs or illustrations into navigable 3D environments by generating out-of-frame objects and backgrounds that feel natural, allowing creators to reposition and explore beyond the original image boundaries.

Is Project Sound Bender already integrated into Adobe's products?

No; all three tools are experimental features shown at MAX 2026. Their integration into existing Adobe video and audio tools is not yet confirmed.

#adobe #generative-ai #video #3d #audio #creative-tools