Tooling Around: Sculpting With Skymatter's Mudbox

For the next part in Gamasutra’s ‘Tooling Around’ feature, which profiles and interviews middleware and tools developers about their products, we talk to Andrew Camenisch, co-founder of Skymatter, developers of 3D sculpting and modeling tool Mudbox. The software was developed by the New Zealand based company’s founders during their time working on The Lord of the Rings at Weta Digital as a way of expanding their own toolsets, and was first used as a complete product on King Kong. The initial beta was released in May 2006, and was followed by version 1.0 in mid-February of this year. Camenisch indicates that Skymatter is currently working at releasing version 2.0 later in the year. Mudbox is a brush-based sculpting tool designed “by production artists for production artists” with a toolset for designing, shaping and detailing geometry, as well as creating normal and displacement maps. “With Mudbox 1.0,” says Camenisch, “we introduced workflow-changing features such as 3D Layers, Local Subdivision, Asymmetrical Mirroring, and Real 3D stamping. And with Mudbox 2.0 we will be expanding the toolset with the ‘Chameleon Engine’ - production-class texture painting.” We spoke to Camenisch recently and asked about Skymatter, the Mudbox software and its use in the industry. When and why was Skymatter formed? Skymatter was formed in 2005 to bring solutions to production artists. Development of Mudbox, however actually began a year earlier. The Skymatter founders met at Weta Digital while working as lead modelers on The Lord of the Rings, and we began developing Mudbox initially, basically for ourselves - to extend our own toolsets. With overwhelming feedback from other artists using Mudbox Beta in production, we knew we had an opportunity to form a company and create a business model that would allow us to continue developing Mudbox and realize the larger vision for artist friendly, production-class tools. What were the aims and goals of the company at this time? The goal of Skymatter then was basically the same as it is now: Solve production problems by making "best in class" tools that artists enjoy using. How did you realize the need for a product like Mudbox? Our backgrounds are in fine arts - painting, sculpture, drawing, etc. - and I think, like a lot of artists, when we first began making things with 3D software many years ago, we intuitively felt there had to be a better way. Like other production artists, we improved our workflows by writing lots of scripts and customizing our workspaces - but this only goes so far. To try to improve our tools, while working on The Lord of the Rings, we were in touch with software developers explaining our needs, providing direction and suggesting tool implementations. Ultimately, though, we weren't getting the response we felt we needed and didn't see developers heading in the direction we wanted to go - so we decided to make our own tools! What was the development time on the product, and what challenges did you run into in preparing the product for industry use? From the beginning, we were so focused on addressing real production issues that after only five months in development Mudbox, was already being used in production. We spent the next year and a half developing Mudbox in multiple productions before making it available to the public. One challenge in preparing the software for public use was testing the software on different platforms and diagnosing issues that appeared on machine configurations we didn't have in-house at Skymatter. In this regard, we were fortunate to have a great beta team with many top artists and companies from around the world, who provided valuable feedback on features and stability. Many of these team members ended up using the Mudbox Beta on productions for almost a year. How has the product developed over the time you've been producing it? Mudbox is under constant development. We just released a version 1.0.6 with a fairly significant performance boost for certain computer configurations. How have you acted on feedback to improve Mudbox? We have a really great community of artists and companies who regularly offer feature suggestions, report bugs, etc. This helps us improve Mudbox. Sometimes responding to feedback involves simply tweaking the UI to make things clearer; other times it involves implementing a new feature or different control structure - as we did recently with the addition of the requested option for trackball-style camera control. One thing that has been confirming for us, though, has been comparing requests and feature suggestions with items that are already on our own internal list - it is validating to see that we're on the same page with other Mudbox artists and heading in the same direction. We understand their workflows. How does the product work on a technical level? Mudbox sports a high performance architecture and scalable design. We spent a lot of time on our brush code to achieve good feel and to enable accurate, fast sculpting on dense geometry. Mudbox is a node based system, which means a number of things: cleaner code base for future development, better foundation for integration with other 3D pipeline applications, and - when exposed to artists in upcoming versions - lots of power under the hood for tinkering artists. We also developed Mudbox to be multi-platform compatible, though we have not yet completed ports to the Mac and Linux platforms at this time. Do you feel your product works best when combined with other middleware? Mudbox is intended for use within a pipeline. We have created a familiar workspace so that artists can move back and forth between software without having to switch paradigms! This can be important for high pressure production schedules - not to mention, significant to the artist's sanity. New users keep telling us how easy Mudbox is to pick up. What are some of the more notable examples of the product's use? Who is currently using the product? Mudbox was first extensively used on King Kong at Weta Digital. Today Mudbox is being used on major films and game titles produced by companies like Epic Games, Blur Studio, Pandemic, Radical Entertainment, Tippett Studios, Cinesite, and many other companies that we can't list here. After only four months on the market, it's terrific how quickly Mudbox has been adopted by the industry. What do you see as the next evolution of Mudbox? Mudbox 1.0 is really only the first step. Mudbox was developed from scratch and brought to market in two years and we intend to maintain a rapid pace of development. Sculpting is a core aspect of Mudbox, and making sure we provide the best sculpting tools possible will always be important to us. At the same time, there is a lot we had planned from the beginning for Mudbox that we just haven't gotten to yet. If you look at all the different aspects involved in creating 3D assets - we're interested in all of that. We'll be able to share more about this later in the year. We're aiming for a strong version 2.0 release in 2007.