Three weeks since I last posted, and I'm just back from my summer break.
I've had a couple of weeks to taking stock of my Houdini training. I've realised how deep the program goes (deep) and how deep I'm able to go with my existing understanding of Autodesk Maya (not as deep). To get the most out of procedural modelling, it's useful to have come from a programming background, and I don't. That doesn't mean I can't use Houdini, just that I know where to draw the line for this research.
I had a great chat with my collaborator, Johannes, about what is achievable in the remaining time we have. We talked about economy and labour in VFX, about maths and art and science. We tried to assess the challenges we've met so far in our collaboration (time, ways of working, what the end goal is) and how two different mindsets (the artist and the engineer-maker) might work best in harmony.
We also talked through the Stately Home project and realised that although the aims are admirable, the project might be a little code-heavy for me to be able to work with. We've come up with a project we're both really interested in that could turn into something really special: we're going to create a set of tools that will allow any 3D model to undergo a transformation as if it was a building subject to decay, overgrowth and renovation. This will involve making:
- a set of controls to partly destroy a 3D model using impacts and decay
- a tool that will grow vegetation on and around a building (ivy, weeds, trees, or bushes).
- a procedurally-driven scaffolding generator that will envelop any mode
Altogether, these tools will allow me to load any 3D model and subject it to a life-cycle. I've long been interested in the qualities of 3D models as virtual and atemporal objects. Like most digital information, they are not subject to conventional decay and therefore don't have the chance to accumulate patina.
I've talked about and worked on this concept for a while - the idea that western game artists often create models of wrecked Middle Eastern villages for use in combat games, whereas architectural visualisation artists make plan-perfect buildings for Asian and Middle Eastern property developers. Both types of model exist in the same marketplace and are perfectly finished whether destroyed or not. With this Houdini project, we'll be playing with an oscillation between these states of perfection and perfect destruction. Here's an early visualisation of the scaffolding tool:
We'll continue working on this while I look at other Houdini projects that I can construct without too much custom coding. I'm also collaborating with Altair in Leamington. They are helping me crash cars using their heavy duty engineering software. This sort of physically correct calculation can't run on any normal 3D workstation.
That's all for now!