Mitsuba Conda Dockers

Dockerfiles to build Mitsuba 0.6.0 with bindings to a python version installed with Conda. These dockers build Mitsuba from a fork I made to build on python3 scons and with all the files preconfigured to build the python bindings with Conda.

Inspired by mitsuba-docker.

The dockers are organized in the folder /dockerfiles.

Currently only the docker for ubuntu18.04 is available but more versions will be added soon.

❗Currently supports running mitsuba through the command-line interface, but enabling mitsuba GUI should be possible following mitsuba-docker

Build and run the dockers

to build a docker:

docker build -t your-name/mitsuba-conda ./dockerfiles/ubuntu18.04

to run

docker run -it your-name/mitsuba-conda

Render a scene

You could run the docker mounting the folder your-folder containing your-scene.xml to render that scene.

your-folder will be mounted in /app, and we can render the scene with the command mitsuba /app/your-scene.xml.

Two ways of doing this:

1. either run the container in the terminal

   docker run --mount type=bind,source="$(pwd)"/your-folder,target=/app -it your-name/mitsuba-conda

   and then run this mitsuba command

   mitsuba /app/your-scene.xml

2. or run the container without a terminal but provide the mitsuba command already

   docker run --mount type=bind,source="$(pwd)"/your-folder,target=/app your-name/mitsuba-conda mitsuba /app/your-scene.xml

📸 Both ways will output your-scene.exr in the folder containing your-scene.xml

Test

To test rendering a scene.

1. download the Cornell Box scene cbox.zip from http://mitsuba-renderer.org/download.html.

2. Unzip cbox.zip, then in cbox.xml change the second line <scene version="0.4.0"> to <scene version="0.6.0">.

3. Finally run:

   docker run --mount type=bind,source="$(pwd)"/cbox,target=/app your-name/mitsuba-conda mitsuba /app/cbox.xml

Mitsuba-python bindings

The dockers are already set up so that conda installs a suitable version of boost-python required by Mitsuba. To test the mitsuba python binding run:

docker run your-name/mitsuba-conda python3.7 -c "import mitsuba;from mitsuba.core import Vector;print(Vector(1.0, 2.0, 3.0))"

Make sure to specify the correct version of python installed with Conda. By default, the command python will invoke python2.7 interpreter.