When we open the blender file, you have all the components of the simulator. Each one is made up using normal blender objects carefully adapted or picked to work like its real counterpart (to simulate, not emulate). In the system, we have a simulated point source of light, a deformable mirror and a wavefront sensor. Later on we progressively explain how things work internally and how they were done.įigure 2: System in action (better viewed in fullscreen)

And the vídeo shows the system in action. To better explain the whole process, the next figure illustrate the “usage” of the system. The whole simulator is a little bit intricate because a lot of problems had to be solved in order to make it possible. We really simulate the light and compute interaction matrices, DM commands, etc all using “real” raytraced light data. We could “emulate” an AO system by having a pre-computed interaction matrix for some pre-defined objects and show how SPFs, Dm commands, etc would behave. That means that we really simulate the physics of the light diffracting and reflecting off of surfaces. It is important to mention that this is real simulator not an emulator. We also compute all the cool math that are used in a real AO system with the data generated by our simulator.

We simulate the star light, the medium where the light passes through (atmosphere), the wavefront sensor that measures the phase deformation and the deformable mirror that corrects for that deformation. In summary, we use blender together with python, Matlab and a bit of (OSL) Open Shading Language to simulate a simple (although quite didactic) AO system. For details about what we simulate, please refer to the previous post. In this post we will talk about the implementation of the simulator.

This post is the continuation of a previous post about making an AO system simulation using blender (+friends).

Blender light beam