THE EMERGENCE OF 3D IN GEN-AI PIPELINES

And why Adobe's potential is trapped behind a paywall

This piece starts with a case study. Sebastian Marek was commissioned by Adobe's own Substance 3D team to demonstrate what a modern digital twin pipeline looks like in practice. The result is stunning—a single 3D master asset, material variants applied in seconds, AI-generated environments swapped via text prompt, a complete library of marketing-ready outputs produced without a single physical prototype or studio booking.

https://sebastianmarek.com/work/digitaltwin

It's exactly the kind of work that makes you lean forward. And then, almost immediately, it made me furious...Not at Sebastian. Not at the work. At the gap between what Adobe is demonstrating is possible and what Adobe is actually making available to the people who need it most.

I've been working in 3D, motion, and visual communication for over fifteen years. I've watched Adobe circle the 3D space like a nervous suitor—acquiring, previewing, promising, adjacent-integrating—without ever fully working up the courage to go all the way. It is time for Adobe to go all in and stop playing mind games in this love triangle.

That case study crystallised something I've been unable to articulate cleanly until now.

Adobe doesn't have a 3D problem. Adobe has a courage problem.

The landscape has already moved

While Adobe has been deliberating, the rest of the industry has been building.

A new category of tools has quietly emerged—platforms that use 3D not as the finished output, but as a structural control layer for AI to work on top of. The render becomes the scaffold. The AI does the heavy lifting. The creative stays in control.

krea.ai Realtime lets you sketch spatial intent, drop an image reference, write a prompt, and watch the result resolve in real time. Fragment (https://studio.fragmentslab.org/) combines 3D camera controls, doodles, reference images and text prompts into a single canvas—not a node board, not a timeline, a canvas. The distinction matters enormously to how creatives actually think and work.

StyleFrame takes 3D previz renders and uses them as material and lighting reference to generate client-ready imagery without the render wait. Test materials in minutes. Iterate without touching the original geometry. Mago studio goes further—taking 3D playblasts and compressing the entire end-of-pipeline workflow into a single AI rendering pass, preserving camera movement, depth, capturing MoCap, and temporal consistency across every frame. What once took weeks now takes hours.

And then there is WonderStudio—now Autodesk Flow Studio—which turns video footage directly into a 3D animated scene, with full motion capture data, character integration, and lighting reconstruction handled by AI. It launched as an independent product. Autodesk recognised what it represented and acquired it in 2024.

Some of these platforms are already production-grade. Others are approaching that standard rapidly. What they share is a conviction—that 3D control combined with generative AI creates a creative pipeline that is faster, cheaper, and more art-directable than anything that came before it.

Adobe saw this coming. Project Neo exists, but it feels more like a side experiment than a strategic decision. Adobe has owned the components to build this properly for years. It hasn't quite cracked the formula—or committed to cracking it.

A decade of almost

The acquisitions tell a story.

In 2015 Adobe acquired Mixamo—a character rigging and animation library that could have become the beating heart of a 3D character pipeline inside Creative Cloud. Eleven years later, Mixamo lives at mixamo.com. Separate login. Separate interface. No meaningful integration with After Effects, Character Animator, or anything else a motion designer or content creator uses daily. You export an FBX and drag it somewhere else yourself.

In 2017, Adobe's internal Project Felix became Adobe Dimension—a 3D compositor designed for product visualisation and packaging mockups, positioned as the Creative Cloud-friendly face of 3D. It had genuine promise. It was quietly deprecated in 2021—not as an admission of failure, but apparently to nudge users toward a Substance subscription. Adobe removed a CC-native 3D tool to protect a separate revenue line. That is a choice worth sitting with.

In 2019, Adobe acquired Allegorithmic, makers of Substance—the industry standard for materials and texturing, used across AAA game development, VFX, and product visualisation. At acquisition, Adobe's own Chief Product Officer described Substance as "a natural complement to existing Creative Cloud apps." That was six years ago. Substance remains a separate subscription, priced close to Cinema 4D, for a tool that occupies a single step in a pipeline that already costs Cinema 4D money.

Also in 2019, Project Aero promised AR content creation natively inside the Adobe ecosystem. It went nowhere.

Each decision, viewed individually, is defensible. Viewed together, they describe a company that keeps acquiring tickets and never boarding the plane.

The locked door problem

The Substance situation deserves particular attention because it has recently become actively antagonistic to the user experience.

After Effects now supports Substance materials. On paper this sounds like progress. In practice it exposes one of the most frustrating gaps in the entire Creative Cloud ecosystem. You can apply Substance materials inside After Effects. You cannot create or edit them without a separate Substance subscription that costs nearly as much as Cinema 4D.

Adobe has built half a pipeline and called it an integration.

The comparison that matters here is Cinema 4D Lite. C4D Lite ships inside After Effects at no additional cost. It is limited—deliberately so—but critically, the loop is closed. You can model, rig, render, and composite without ever leaving After Effects or reaching for your wallet again. You can complete a thought. The lite model works not because of what it includes, but because it doesn't break the chain at the moment you need it most.

Substance inside After Effects breaks the chain at the most fundamental point. Apply, yes. Make, no. That is not a workflow. It is a shop window. And shop windows dressed as features erode exactly the trust that Creative Cloud's value proposition depends on.

The absence from Photoshop compounds this. Photoshop is already a paint application. The mental model for material creation (texture painting, surface editing, procedural adjustment) maps naturally onto tools CC subscribers already know and use daily. Adobe hasn't followed its own logic.

This is not an isolated frustration. In a recent piece I argued that After Effects is currently undergoing essential infrastructure work—three unresolved 3D engines, a UI not built for serious spatial work, foundational pipeline repairs before AI features are layered on top. That reading makes the Substance situation simultaneously more understandable and more urgent. If Adobe is genuinely fixing the foundation, Substance needs to be part of the structure when it's done—not an optional extra bolted on after the fact. Adobe Max 2026 is where that argument either holds or collapses.

Adobe has learned this lesson before

This is not the first time Adobe has fragmented a product ecosystem and paid the price.

Before Creative Cloud, Adobe sold Creative Suite in vertical packages. Design Premium had Photoshop, Illustrator, and InDesign but not After Effects. Production Premium had After Effects and Premiere but not Illustrator. The gaps were arbitrary. The pricing meant professionals were constantly buying around the edges of the tools they actually needed. It was a support nightmare, a sales nightmare, and a loyalty killer.

Creative Cloud solved this by unifying everything under a single subscription. Adobe grew substantially as a result.

Maxon learned the same lesson with Cinema 4D. Prime, Visualize, Broadcast, Studio—MoGraph locked to Broadcast, rendering tools fragmented across tiers. A motion designer couldn't access architecture tools they occasionally needed. When Maxon collapsed it into a single unified subscription, adoption broadened. Again, not a coincidence.

Adobe has the case study. Inside its own history. And it is repeating the mistake with Substance—and Mixamo, and Character Animator, and the entire 3D layer of Creative Cloud.

The platform nobody else could build

Here is what makes this genuinely frustrating rather than merely irritating.

No other company is positioned to do what Adobe could do here. Firefly handles generative image and video. Project Neo is edging toward generative 3D. Substance has the material and lighting intelligence. After Effects has the compositing layer. Gaussian splat pipelines are emerging as the bridge between real-world capture and synthetic environments. Mixamo has the character library. Character Animator has the performance capture layer.

All of it exists inside Adobe. None of it talks to each other in a unified creative canvas.

The tools being built independently by Krea, Fragment, Mago, and others are impressive. But what they cannot replicate is ecosystem coherence. I chose Redshift over Octane as a renderer not because Octane was inferior—it was arguably better. I chose it because it lived inside my existing pipeline rather than alongside it. Integration that breaks the chain is not integration. A tool that requires you to leave your ecosystem to use it will always lose to a tool that doesn't, even if it is technically superior.

That is Adobe's structural advantage. And it is being squandered.

The vision that should exist—that Adobe is uniquely positioned to build—is a CC canvas where you sketch spatial intent, drop an image reference, write a prompt, and Substance materials resolve the surface logic while Firefly generates the environment and Neo handles the volumetric geometry. Non-destructive. Not a node board. A canvas. The kind of real-time, reference-driven, prompt-assisted workspace that Krea and Fragment are building as independent tools, without the ecosystem Adobe already owns.

Instead, Krea builds it. Fragment builds it. Mago builds the playblast-to-final-render pipeline. Autodesk acquires Wonder Dynamics and builds the video-to-3D-scene pipeline that Adobe's own Mixamo acquisition was pointing toward in 2015.

The Blender problem

There is one more force Adobe is underestimating.

Blender is not better than Cinema 4D, Maya, or Houdini. It is good enough, and it is completely free. The distinction matters because the next generation of 3D artists is learning on Blender—not because it is the best tool, but because it removes the financial barrier entirely. This is the DaVinci Resolve pattern. Resolve went free, editors learned on it, and suddenly Premiere had to justify itself to people whose muscle memory already lived somewhere else.

3D applications demand enormous learning investment. Users who commit to a tool commit deeply—switching carries a real cost in time, retraining, and rebuilt workflows. The window to earn that commitment is early, when habits are forming. Adobe is losing that window to a free tool, not a better one. Once those habits are formed outside the Adobe ecosystem, the likelihood of those users ever becoming CC subscribers for 3D work diminishes significantly.

The switching cost has inverted. The paid tool now requires justification. Adobe is losing a generation of 3D-native creatives before they have ever opened a Substance trial.

The ask

Dear Adobe Creative Cloud and Adobe 

Including Substance in Creative Cloud would not fix all of this. But it would give every CC subscriber—every motion designer, product photographer, event designer, brand designer, and content creator who already pays Adobe every month—a reason to build their 3D material habit inside the ecosystem rather than outside it.

Substance was marketed and priced as a specialist game-industry tool. It is not one anymore. The generalist creative who already lives inside CC needs materials and texturing. Adobe is charging specialist prices for a tool with generalist reach.

Substance is sticky for the same reason the rest of Creative Cloud is sticky: a unified ecosystem with dynamic linking between apps means your assets, your material libraries, and your project files all live in one place and talk to each other. Once that infrastructure holds your work, leaving becomes genuinely costly.

There is also a quieter argument here that Adobe's sales team should be making loudly. CFOs and IT procurement leads like fewer subscriptions. Every additional tool a creative team needs outside of CC is another procurement conversation, another vendor relationship, another line on a software audit. Folding Substance into CC removes a barrier that finance departments are often the ones blocking. This is not just a creative win—it is a commercial one.

Include Substance in Creative Cloud. Make material creation available inside After Effects and Photoshop without an additional paywall. Properly integrate Mixamo with Character Animator and After Effects. Commit to Project Neo as a canvas, not a side experiment. Connect the pieces that have been sitting in separate rooms for a decade.

Adobe has been flirting with 3D since 2015. The industry has stopped waiting. Autodesk is executing. Blender is capturing the next generation. Fragment and Krea are building the canvas Adobe should have built three years ago.

The components exist. The ecosystem exists. The only thing missing is the decision to connect them.

Go all the way, Adobe. The creative community has been patient long enough.

Inspired by the work of Sebastian Marek, whose Adobe Substance 3D digital twin case study is well worth your time: sebastianmarek.com/work/digitaltwin