The Next Big Tutorial
I’ve finally reached a point where I have about an hour of dedicated time set aside each day for CGFigures projects. That means I can finally turn my attention to the biggest tutorial undertaking I’ve done since my beginner series. Namely, a comprehensive introduction to the new geometry nodes system coming in Blender 3.0. Today I want to outline the basic idea for that tutorial and give some sense of how long these things end up taking.
The new ‘fields’ version of geometry nodes is fortunately very similar to the well established shader system. For tutorial purposes that means that I don’t have to learn an entirely new workflow from scratch and that’s a huge time save. Though this version is great (and actively improving) my understanding is that it still lacks some of the features available in animation nodes and Sverchok (both extra parts of Blender for those unaware). Likewise, the incredible procedural capabilities of Houdini are a cut above, which is to be expected. Even so, it’s been harder to justify teaching those systems. For starters, I am not experienced with any of them. Stuart Jantzen of biocinematics has a fantastic starting course for Houdini (I intend to finish it soon enough). The bigger issue to me is that animation nodes has long been an ‘experimental branch’ and Sverchok is independently developed and is somewhat more directed to architecture. Geometry nodes, though a bit less feature rich right now, is part of the main Blender package. No extra downloads, no multiple software platforms, just download Blender and use it directly. That accessibility is important to my mission specifically. I want scientists to have better figures. I don’t expect them to commit dozens, if not hundreds, of hours to learning 3D. Reducing barriers to entry is an important part of that process.
At this point it would be fair to ask, why introduce geometry nodes at all? Isn’t procedural geometry complex to the point where it is beyond the “beginner level” I just described? Well yes. And no. Many of my current tutorials are already beginner accessible walkthroughs to create more complex effects. My asset releases bypass the need to model anything at all. For me, geometry nodes perfectly straddles the two. The tutorials help to teach a workflow or to recreate an exact effect or model but usually with limited customization. The models save you the time of having to make a specific thing so you can get to using it. With geometry nodes I can release the final model with a beginner friendly control interface for increased customization. The tutorial shows exactly how to make the desired product and gives extra control or understanding to anyone who needs or wants it. I’ve advocated for this before with procedural materials.
Right now the tutorial layout is pretty simple. Start by establishing the version of Blender. Discuss the basics of using geometry nodes and some of the approach. From there, launch into several major examples that would each normally be their own tutorials. Each example will showcase a different set of capabilities in geonodes and collectively they will demonstrate how broadly useful the system can be. The examples will all be cleaned up and released so anyone could use them for their own work. I’ll then highlight several other creators who I enjoy learning from as well as some useful tools and add-ons. Finally, I’ll close off with a discussion of what geonodes is going to bring to the Blender scientific community, and some of my own future plans.
