Please introduce yourself and your team. How did the story of colormass begin?
Tas [Sóti] and I both have a technical and 3D background: we actually worked on face reconstruction technology (which was the high-tech spinoff of the Humboldt University of Berlin) before colormass. That experience gave us direct insight into the scalability challenges of the 3D industry. Many production workflows still depend on manual file naming and managing 3D assets within disconnected silos. It felt hugely inefficient.
At that time, tools like Excel and Word already had seamless online versions (like Google Docs) that made collaboration easy – no more sending files back and forth. But in 3D, nothing like that existed. So in 2016, we founded colormass with this exact same vision: to bring scalable, cloud-based 3D CMS to the industry, and that has remained unchanged.
As we were developing the 3D CMS, we realized that many of our clients were struggling with digitizing materials, sometimes spending days creating a single material manually. About a year into development, we started building our own svBRDF scanner and tiling toolset to automate both material digitization and post-processing, helping clients speed up and simplify their workflows.
How big is your current team? How is it organized? What divisions do you have? What experience do your teammates have?
We’re a team of 28 people, split between two offices, one in Berlin and one in Budapest. Our software development and sales teams are based in Berlin, while our Account Management and 3D Artist teams work out of Budapest. The Budapest team focuses on supporting our clients in using the platform and helping them scale their 3D production overall.
Our development team (in Berlin) is still small but quite strong. Here are a few of these team members:
- Dr. Tobias Nöll and Dr. Johannes Köhler: Before joining colormass, they served as Senior Researchers at the new ZEISS Innovation Hub in Karlsruhe (@KIT), where they specialized in 3D digitization for over six years. They were also recognized as a recipient of the Google Research Award.
- Lennert Schneider: He was part of the core development team of Crytek.
- Ziga Bostner: During his time at Zuse Institute Berlin (Department of Visual Data Analysis), he was a main contributor to Amira, a software package for the visualization, analysis, and processing of 3D data.
- Roman Priebe: He holds a PhD in Quantum Information Technology from the University of Oxford.
- Matt Hughes: Prior to joining colormass, he was the System Architect for Line 6 (Yamaha) products.
The Budapest team is organized into Account/Project Management and three core 3D departments: Modeling, Shading, and Lighting & Layout, each with around four people. This 3D team plays a crucial role in helping our clients scale their 3D production. Whenever clients don’t have the in-house expertise or resources for a specific task, our team is there to assist. They also provide training to help clients get up to speed with the platform.
Even though we are a software company, our 3D team is at the core of everything we do. They work with the platform daily, offering invaluable feedback on which features are needed and where improvements are necessary. This close connection between the product and real production work is something that, in my view, sets us apart from many other 3D tech companies. Building the team this way was a conscious decision from the very start.
What services do you provide? How can you help 3D Artists with their projects? Could you share some successful cases?
We provide both software and hardware solutions across two main verticals, both designed to be fully self-service:
- Browser-Based 3D CMS: A platform that lets you manage your 3D assets, render images, create configurators, and more, all in one place.
- Material Digitization: We operate scanning stations in Germany and the US where anyone can send in materials for digitization. For companies with high-volume needs, we also offer on-site scanner installations. On top of that, our Tiling Toolset simplifies and speeds up the post-processing workflow – not only for our scans but for any set of PBR maps. We also recently released a new tool called Material AI, which allows you to convert standard top-down images into PBR maps. Unlike typical AI tools, this one is trained on a dataset of 16,000 svBRDF scans, making it one of the very few AI solutions capable of delivering high-quality, production-ready results.
Our technology is not limited to specific use cases. Overall, I’d say our tech is ideal for anyone who wants to:
- Streamline collaboration across 3D silos and remote teams by setting up a cloud-based system that connects with various software (e.g., a central material library accessible via plugins in Blender and Cinema 4D), complete with built-in versioning and referencing.
- Create online 3D experiences, such as 3D viewers, configurators, and AR, without needing development expertise.
- Easily share 3D scenes with clients.
- Digitize and process materials in a far more automated and efficient way.
As Claire Mann, Director of Digital Products and Strategy at Designtex, puts it: “colormass is an industry leader in material capture, 3D rendering and visualization technology with hardware and software solutions that allow us to accurately and beautifully represent our product catalog online. One of our favorite partners, the team is engaged, collaborative, and always willing to translate highly technical aspects of the process, brainstorm future possibilities and suggest solutions”.
Could you tell us more about your cloud-based platform? What are its capabilities? What tools does it feature? Please tell us more about your PBR scanning technology and rendering platform.
Platform (Browser Based 3D CMS): You can think of it as Blender or Cinema 4D in the browser but focused entirely on managing and using 3D data without model creation features. It’s a central hub where you can upload, organize, and manage your 3D assets and then use them across teams to render images, build configurators, or export data to other tools, all fully self-service.
At the heart of the platform is what we call the Single Asset System, designed to make LOD generation as dynamic and automated as possible. For example, any model you upload can automatically be downloaded as AR-ready files (.gltf and .usdz) without any manual conversion.
So while I mentioned different functions like rendering and configurators, these are really just different outputs from the same core system that stores high-quality source data and dynamically downscales it when needed. And if more control is required, you can also create manual LODs and define complex logic for variations of your 3D assets.
The online editing works much like any other 3D software: it includes a 3D scene editor where you can define products and scenes (see the products and scenes documentation), as well as a browser-based material node editor that lets you create or modify materials (see the materials documentation).
One of the major benefits of being cloud-based is seamless referencing. You can define what can be overridden in a parent scene for each 3D asset, and whenever you update an asset, it automatically updates across all scenes where it’s used. The referencing stays intact, even if you rename an item. Additionally, since revisioning is built into the system, you can revert assets to previous versions when needed.
If you prefer using Blender to create scenes and assets, we also offer a plug-in that streams data between Blender and the cloud system, allowing you to easily upload assets and scenes as needed.
When you submit a render through the 3D scene editor, the images are rendered using Cycles and processed on Cloud GPUs. This means that even if you submit hundreds of images, they are generated in parallel. Since the rendering process integrates natively with the cloud system, you won’t need to worry about maintaining plugins or ensuring version compatibility, as is often required with render farms; everything just works.
Material Digitization: It’s important to distinguish between two types of scanners available on the market:
- Heuristic Scanners: These scanners rely on heuristics, capturing images and using simple algorithms or image manipulation to create maps.
- svBRDF (or Inverse Rendering) Scanners: These scanners reconstruct the full reflection properties for each pixel captured.
The first type is much easier to implement but doesn’t result in high-quality materials; it requires significant tweaking and serves more as a starting point than a final, polished material. The second type, which is the one we focus on, produces a perfect material with no need for additional adjustments (and so far in the world we are only aware of a single other such scanner on the market). You can find out more about it here. If you would like to look at an example output, you can check it here.
Since post-processing the materials is just as much work as scanning them, we also built a texture editing toolset for tiling materials more automatically (for technical details, check our documentation).
And what about your Configurator app? How does it work?
The Configurator tool enables anyone, even without programming knowledge, to create configurators and online 3D experiences directly in the browser, no downloads required. You can use the 3D editor to build your scene and then easily integrate the viewer into any website. Through the 3D CMS, you can upload your geometries and materials and even edit the materials online using the node-based material editor.
You can easily integrate the final 3D viewer using just a simple link, with no development skills required. However, if you have programming knowledge, you can also use the API or web components to customize the style and UI elements to suit your needs.
How do you approach the business side of things and promote Colormass? How do you work with the community?
To be honest, up until now, we’ve mostly kept a low profile and we were not really marketing ourselves in the 3D industry. Since we have an internal 3D team, we’ve had the perfect opportunity to test what works and what doesn’t, which has guided our development over the last 8 years (as you can imagine, building a robust 3D CMS is highly complex). Recently, other 3D teams in our network have started adopting our 3D system for their production needs and to send 3D scenes directly to their customers, so we felt it was time to begin involving other 3D teams who might be interested in using it for their own projects.
Looking ahead, we’re considering opening it up to the general public in 2026. Until then, anyone interested can submit a request to use the system via our contact form. The reason for this is that we want to focus on collaborating with companies that we believe would benefit most from our solution. We usually determine this in the course of an initial call, which is more of a technical discussion than a sales pitch. It’s an opportunity to understand the context in which the system would be used and see if it’s a good fit.
