Blog

Everything in the Arcanum Saga was built by hand in Unreal Engine. No AI image generation, no diffusion models, no shortcuts. Every environment, every character, every light and particle system — authored directly, one scene at a time. The look didn't come from a prompt. It came from years of technical work and a very clear vision of what this universe needed to feel like.

Aeon took two years. Two years of building shaders, sculpting geometry, solving dome projection, iterating on camera paths, rendering and re-rendering sequences until they were right. The fulldome format doesn't forgive — the audience is inside the image, surrounded by it, and any weak seam or awkward transition is immediately felt. Getting that right took time we couldn't rush.

Ai Divine came next and has been a different kind of project — one year in and still ongoing. It's a larger, more complex film, and the world demanded more. New characters, new environments, expanded lore. The production is active; the film grows as the universe does.

One recent addition to Ai Divine: the soundtrack now features original music composed using Suno — music we made ourselves for this specific film. That's the only place AI tools appear in this production: in the audio, created by us, for this project. The visuals remain entirely hand-built.

Building in Unreal Engine for the dome format means solving problems that don't have established solutions. Equirectangular projection at dome quality, real-time vs. offline rendering decisions, how light behaves in a hemispherical frame — these are craft problems, not generation problems. The engine is a tool we've learned to push far past its defaults.

The Arcanum Saga is a ten-film universe. Aeon is complete and available for licensing. Ai Divine is also licensable — an ongoing production that continues to grow as new chapters are completed, available to planetariums and immersive venues now. The other eight films are in various stages of development, building out a mythology designed from the beginning to span the full series.

House of Magic in Macau is not a small venue. A $30 million purpose-built theatrical complex, it houses three dedicated theaters — with the main stage anchored by a 100-foot-wide by 20-foot-tall LED panel installation. The scale is deliberately overwhelming. The venue was built to make audiences feel like they've stepped into something impossible.

I was brought in to create the show content. The result was a 70-minute production built around intricate musical cues, projection-mapped set pieces, and live dance immersions — sequences designed so the visuals and performers exist in the same space rather than behind each other. Getting that to work at LED-wall scale, with timing precise enough for live dance, required mapping every cue to the millisecond.

For a holiday run, I pushed the production further: animated anaglyph 3D sequences, purpose-built for the 100×20 LED setup. Audiences were given red and blue anaglyph glasses for several scenes. To my knowledge, this was possibly the largest animated anaglyph 3D ever created for a live theatrical show — the sheer physical scale of the screen, combined with content engineered specifically for that stereoscopic offset at that size, puts it in territory that simply hadn't been attempted before.

Anaglyph at that scale behaves differently than at cinema or display size. The depth planes have to be calculated for a viewing distance measured in tens of meters, with an audience spread across a wide arc. Getting the 3D to read cleanly across the full house — not just center-orchestra — meant treating the geometry of the room as a variable in the content design, not an afterthought.

House of Magic represents the kind of project where the technical constraints and the creative ambition push each other forward. A 100-foot canvas demands ideas that earn the scale. The anaglyph sequences were the moment in that show where the audience stopped watching and started inhabiting something.

In 2011, we were asked to provide visuals for a Tiesto performance. We'd done plenty of stage visual work by that point — projection mapping, LED wall content, real-time reactive systems. But the production team had something specific in mind: two large-format autostereoscopic displays, flanking the DJ booth, playing custom content. Glasses-free 3D.

Autostereoscopic displays — the kind that create a 3D effect without requiring the audience to wear glasses — were barely mature technology at the time. The consumer versions were small, low-resolution, and finicky about viewing angle. Getting theatrical-scale versions in front of a club audience was uncharted territory.

The technical challenge was real-time content generation in a format these screens could actually display. Autostereo screens require a specific multi-view rendering approach — you need to generate the scene from 8 or more slightly different viewpoints simultaneously, interleaved in a specific pixel pattern that the screen's lenticular layer separates into the correct eye for each position in the audience.

We built a custom real-time renderer that output a lenticular-ready signal. It had to run at full frame rate, driven by audio analysis from the booth, adapting to the music in real time. We tested it for weeks before the show, terrified it would fail in front of 4,000 people.

It didn't fail. The effect — seeing 3D imagery floating in front of the DJ booth without any glasses, responding to the music — was something that specific audience had never experienced. We got reports afterward from people who'd been to hundreds of club nights saying it was the most disorienting and beautiful thing they'd ever seen at a show.

To our knowledge, it remains the first time real-time 3D content was performed live on glasses-free autostereoscopic displays in a concert environment. The technology has advanced significantly since then, but that moment — the combination of the right idea and just barely possible execution — is one we're still proud of.

The most common mistake in immersive media is confusing visual complexity with emotional power. We spent years learning this the hard way — shows packed with intricate geometry and particle systems that left audiences unmoved, while a single slow camera pull through an empty void could leave people in tears.

The dome is a strange vessel. It removes the rectangular frame that every other visual medium uses to manage audience attention. There's no edge, no safe place to look, no hierarchy imposed by framing. The audience is inside the image. That means everything you put in the space affects everything else — and restraint becomes one of your most powerful tools.

The second lesson: tempo. Dome shows that try to sustain a fast visual pace exhaust audiences in minutes. The inner ear is involved in ways flat-screen cinema isn't. We learned to build in moments of stillness — not dead space, but deliberate breathing room where the image resolves and the viewer can feel oriented before the next transition. Aether was our first show that got this completely right.

Scale is a gift and a trap. The dome lets you make things feel enormous in a way no other medium can match. A slow drift through vast architecture, a ceiling that opens onto infinite space — these are genuinely impossible everywhere else. But the trap is using scale as a substitute for meaning. Enormous things need to feel enormous for a reason, or the viewer disconnects.

Sound is probably 40% of the immersive experience — significant, but not the whole story. Audio and visuals develop together, informing each other as both evolve. There's no fixed rule about which leads. Sometimes a sequence drives the sound direction; sometimes a track unlocks what a scene needed to be. The relationship is fluid, and locking it to a formula kills the work.

The last lesson is harder to say politely: don't over-romanticize the audience. The bar for dome content has historically been low — abstract, amorphous, ambient visuals that function more as a screensaver than a film. Audiences raised on that have learned to accept it. The opportunity isn't to trust them to bring their own meaning to nothing — it's to give them something with actual narrative and emotional weight that they weren't expecting. Lead them somewhere. Most dome content never tries.