In order to have Render workbench fully functional, you need to install one or more external rendering engines on your system, and to set up this/these external rendering engine(s) in Render workbench.
To use Render Workbench, you must first install one of the supported rendering engines on your system.
At the moment, the following engines are supported:
- Pov-Ray
- LuxCoreRender
- Appleseed
- Blender Cycles
- Intel Ospray Studio (experimental)
- Pbrt v4 (experimental)
The precise installation procedure for each of those engines is beyond the scope of this manual, but detailed installation instructions (adapted to your OS, your distro etc.) can be found on their respective websites. For Linux users, you may also usually find off-the-shelf packages in your distribution repository.
As an exception, you will find below some more information about Cycles renderer, which is a special case.
After installing, please make sure your rendering engine is fully functional: test scenes are usually provided with the renderer for that purpose, please refer to its documentation.
Caveat: Installing Cycles Standalone can be tricky and time-consuming — especially on Microsoft Windows platform. If you have no taste for poorly documented installation procedures or if you have no time to waste, you should rather consider using Appleseed or LuxCoreRender, which both provide ready-to-use binaries, along with excellent rendering features.
To use Cycles renderer with Render workbench, you need a standalone version of Cycles, named Cycles Standalone. This version is distinct from the one embedded in Blender. You will find some more information about Cycles Standalone in the dedicated Blender wiki page.
Cycles Standalone usually requires compilation from sources, as no precompiled binaries are generally available in standard environments. Sources and compilation instructions can be found here.
As an alternative, in the (fairly rare) case you already compile Blender by
yourself, you can enable WITH_CYCLES_STANDALONE and
WITH_CYCLES_STANDALONE_GUI in cmake variables (I also had to add -lGLU to
CMAKE_EXE_LINKER_FLAGS) before your build process. You will then get a
separate 'cycles' executable compiled together with Blender.
Arch Linux users may avoid all this hassle by using the package available in AUR.
Mac OS Big Sur users may also find some valuable instructions here.
Once you have a rendering engine installed on your system, you have to set it up in the workbench.
Each renderer has some configurations to be set in Edit > Preferences > Render before being able to use it.
At least, you must fill in the path to your renderer executable, in the corresponding section. This is a mandatory step, otherwise the workbench will not be able to run the engine when required. Some renderers may provide two flavours of their engine: a command-line and a GUI; in which case you should fill in both.
Optionally, you may want to add some command-line parameters (for instance, to activate GPU rendering, or to specify halt conditions etc.: see your renderer's documentation) to renderer invocation. In that case, you can use the dedicated field 'Render parameters' in your renderer section.
Optionally as well, you can set a few renderer-wide parameters:
Prefix: A prefix that can be added before the renderer executable invocation. This is useful, for example, to add environment variables or run the renderer inside a GPU switcher such as primusrun or optirun on Linux. This parameter is fully optional and can be left empty if not needed.Default render width,default render height: the default dimensions of the rendering output. Default values are 800x600 and can be left as-is if no special dimensions are required.