This add-on is meant as a way to create content for RAMSES using Blender. It does this by exporting a RAMSES scene graph right from Blender itself. It is still under development.
RAMSES is a 3D graphics framework concerned with the efficient distribution of graphics among multiple screens for automotive purposes. While the framework is almost complete, it lacks a tool that can load, modify and export 3D assets in a suitable format so applications developed with RAMSES can use them at runtime.
This exporter makes leveraging Blender - a powerful open source 3D modeller - possible for the creation of content to be used in applications developed with the RAMSES framework.
Rendering of a car in RAMSES from a third-party Blender model. Model is hosted on Sketchfab and licensed under CC BY 4.0
By iterating over the objects in a Blender scene, an intermediary scene graph representation is built. The IR aims to be a middle ground between the abstractions exposed in Blender and in RAMSES. A RAMSES scene graph is then built from this by calling into a set of Python bindings written in C++11/pybind11 and maintained in a separate repository. These bindings are compiled beforehand into a shared library that gets loaded at runtime by Blender's own Python interpreter. After some validation, a RAMSES scene/resources file is saved and can then be loaded by code built for the RAMSES Framework.
As of now only meshes and their transformations - scalings, rotations and translations - and modifiers get exported.
Materials do not work yet. We are investigating our options on this. It will probably use a combination of baking and manual shader editing. It may someday leverage Blender's new 'uber' shader - Principled BSDF - to exchange materials with RAMSES.
Also curves, surfaces, metas and text are not supported yet. You can check Blender's object types here, but we do not plan on necessarily supporting all of them.
As of now, you cannot install this directly from Blender itself. This project also requires building the Python bindings for RAMSES, which is referenced as a submodule. Luckily, we provide a CMake script to simplify the installation process.
First init all submodules, or else the project will not build.
Do this by running the following command: git submodule update --init --recursive
Build the plugin libraries with CMake, e.g.:
$mkdir build && cd build
$cmake -DBLENDER_ADDONS_PATH=<path_to_blender_addons_directory> ..
$make install
Since RAMSES comes with -Werror set by default, it is common for the build process to fail for very minor warnings. If this happens, unset ramses-sdk_WARNINGS_AS_ERRORS via cmake-gui or similar.
Make sure you have the dependencies installed. RAMSES lists its dependencies in its own README, as does the Python bindings.
You need to specify Blender's add-on directory - it is usually under version/scripts/addons_contrib.
For Arch Linux and Blender 2.79, for instance, that would be /usr/share/blender/2.79/scripts/addons_contrib assuming you have installed Blender from the official repositories.
Users under Windows should look for a similar folder inside their Blender installation path.
After running make install activate it via Header Menu > File > User Preferences > Add-ons (or Header Menu > Edit > Preferences > Add-ons if under Blender 2.80). You can filter by Import-Export to make it easier to find.
I recommend using Visual Studio Code with this extension. Then to install execute the following command: >Blender: build and start. This copies your work into the config folder for Blender - i.e. ~/.config/blender under Linux, which makes the addon available in the Add-ons panel (see above). You still need to activate it though. Debugging should work out of the box.
Install the addon using CMake. You will have setup a debugger yourself. See this for more details.
This project includes both unit and end-to-end tests. You can run the tests with python test/run_all_tests.py -b <path_to_your_blender_binary> -p <platform> -a <addon_path>.
When in doubt, run python test/run_all_tests.py --help for guidance. Note that <addon_path> is the full path to where the exporter was installed.
Some tests are vanilla unit tests built with Python's unittest module, while others check the screenshot of the exported scene against a valid image to determine correctness. This project already includes a list of screenshots of exported scenes that were manually checked to be free of errors, but these can be updated with the -g flag in the event something major is changed within the exporter - e.g. when we start supporting materials in the future, these will need to be updated and this is the quickest way to do so.
You can add more tests by appending to the test array in test/run_all_tests.py, and you can add more flags to the tests by reading the instructions in run_all_tests.py.
Run python -m pdb test/run_all_tests.py and set up a breakpoint in p.communicate(). Alternatively, check out stdout.txt, stderr.txt and debug.txt in test/test_results/<your_test_name>