Archive republication: This article was originally published on 1 April 2021. It has been translated into English and lightly updated for clarity while preserving the original reporting and context.
This article originally appeared in Digital Production 05:2015.
Appleseed was introduced during an open-source technology talk at FMX 2015. After six years of development, the renderer was moving closer to production readiness. Digital Production reviewed the state of the project and spoke with lead developer François Beaune about its future.
A growing number of render engines were adopting path tracing. The established options included Arnold, Iray and contemporary versions of RenderMan, while open-source projects offered several alternatives. Cycles shipped with Blender and was planned for Rhino 6. LuxRender and Mitsuba combined path tracing with more complex algorithms. Appleseed was less widely known, but had been developed for roughly six years and was steadily approaching production maturity.
The variety of open-source renderers reflected their different intended uses. Mitsuba, for example, originated in academia and exposed implementations of modern algorithms so researchers and developers could examine them early.
Renderer for Animation and VFX
Appleseed was designed specifically for animation and visual effects. Its motion-blur implementation was intended to be robust enough for production, while technical directors could build programmable shaders with Open Shading Language and use Disney’s SeExpr for procedural materials.
Permissive Licence
The project’s licence was another distinguishing feature. Some open-source licences impose restrictions when code is combined with proprietary software. Code distributed under the GNU General Public License, for example, cannot simply be mixed into a closed-source product.
Appleseed used the more permissive MIT Licence, which allowed broad reuse and made the renderer suitable for distribution with commercial software. Image Engine’s look-development application Gaffer provided an early integration, although Gaffer itself also used a permissive open-source licence.
Feature Set
Appleseed supported major surface models including multiple microfacet BRDFs, an extensive set of light sources, spectral and RGB rendering, render passes and layers, polygonal and image-based bokeh, instancing and other production features.
At the time of the original report, subsurface scattering and volumetric rendering were still absent. Both were planned for later in 2015, so this limitation should be read as historical rather than a description of current builds.
Integration Challenges
Appleseed could be tested on Windows, Linux and macOS. The distribution included Appleseed Studio, a graphical interface recommended for initial evaluation. It accepted OBJ and Alembic geometry, included a Cornell-box scene and was accompanied by additional sample projects on the Appleseed website.
Plug-ins existed for Maya and Blender, but their integration did not yet match Cycles inside Blender or V-Ray inside Maya. They exported scenes rather than providing interactive rendering directly in the DCC. Interactive work remained limited to Appleseed Studio and the forthcoming Gaffer integration.
Conclusion
Appleseed and Gaffer were already worth examining, despite the renderer’s incomplete DCC integration. The engine showed clear production potential, but commercial adoption depended on Maya and Blender workflows progressing beyond export-only operation.
This assessment reflects the 2015 feature state. Consult the current project documentation and test shader, scene-export and render behaviour before production use.
