Himalayan Mountain Study, Round 3!

This is by far the most complex terrain I've ever tackled, the behavior of deep snow is incredibly tough to replicate in CG. This project alone took "only" a week, but that isn't the full story- this is built upon custom Houdini terrain simulation tools that have been in development for the better part of a year. Without them, I would not have been able to get comparable results. I'm working on detailed technical & conceptual breakdowns of the various simulations I wrote, more to come on that front soon.

The base shape of the mountain is nothing special, the primary reference for this mountain (k2) has a relatively simple shape, but the complex behavior of deep snow on extremely tall mountains proved extremely tough to approach. Fine grain material (sand, dirt, etc) alone provide their fair share of troubles, but snow adds a few levels of complexity on top of that.

The snow had to clearly have thickness, most CG snow is basically just plopped on top of a terrain, and it usually works, which makes it generally an afterthought. This completely turned that on it's head. I had to consider the behavior of snow as it responds to wind, extreme slopes, surrounding snow becoming unstable and breaking off the terrain, etc.

More investigation into this is warranted, as I want to push this even further, while many of these techniques technically work, implementing them felt unsatisfying... I want my tools to always have a physically inspired solution, for example, simulating clumpy/glacial features on the snow properly is not something I can currently do, so I had to resort to a workflow that involved carving flowmaps into the snow to generate that harsh, sharp detail. These solutions work adequately, but they also expose an area where my toolset has room for growth. More on this philosophy in future posts.