One of the ways I got back into modeling after a pretty long hiatus, and specifically, vying for game-ready assets, was through trying to sculpt rock formations.
Stones and rock formations, for me at least, embody a somewhat primordial order. There are rhythms and beautiful hues and textures hidden in the chaos of thousands of years worth of erosion and slippage.
There are symphonies of a seemingly chaotic morphology and stories of tectonic significance there. In other words, rocks are pretty awesome and very good practice.
In the models featured throughout this post, I'm illustrating different methodologies that I discovered while trying to learn how to manually unwrap UVs. For the slideshow above I used the following workflow :
- I started off with a very simple low poly base mesh - we're talking 120 polys or thereabout - unwrapped it in 3dsmax, and imported into zbrush.
- In zbrush I subdivided to around 300k and spent probably 15mins adding some superficial deformation and geometric detail
- I exported the high poly mesh as obj and generated normal maps using xnormal.
- Next up was importing the lowpoly asset into Substance Painter and applying aforementioned normal maps. Following that, I used a series of rock texture images to projection paint the surface detail as an albedo or diffuse map. I then employed Painter's awesome particle based brushes and bombarded the model with a midgray/slightly metallic material so as to simulate the effect of water streaks.
- The final step was importing everything into Marmoset Toolbag for map attribution and previewing.
I've included the same model with different maps applied, so from right to left we have the base low poly mesh (nothing added), the same mesh with the normal maps applied, and the final mesh with the majority of maps that Surface Painter spat out.
The most important lesson from this endeavour is that a more 'industry correct' methodology was followed which yielded a pretty acceptable end result without too much fuss and time investment.
In terms of the hard edges of the rock - Ideally those would be completely smoothed over should this be considered an import ready asset - but for the sake of brevity I chose to not modify the base mesh too much as my focus was on the texture mapping aspect of things. More to come over the next couple of days!