Wood
Blender, Material, Procedural, Volumetric
In February 2022 I started exploring procedural wood materials in Blender, trying to capture some of the beauty and depth of the material. I researched types of wood, 'figuring', milling processes, finishes, plywood manufacture, alongside plenty of real world observation to try and simplify the essence of what makes wood look like wood.
I ended up creating between 40-50 materials exploring a number of techniques. Initially most of these had roughly the same elements - a large stretched and distorted noisy texture to represent the basic figuring which comes from the trees annual growth rings, and a much smaller level of distortion representing the fibrous grain, pores & resin channels.
![](img/darkwood-1.jpg)
![](img/darkwood-2.jpg)
I created plank and wood block patterns by dividing the space into cells, generating random values for each cell, and using those values to offset the UV coords by some amount, giving each plank a different texture location. I expanded this into more complex patterns like parquet flooring and other ornate layouts.
![](img/parquet-1.jpg)
![](img/parquet-2.jpg)
![](img/parquet-3.jpg)
Most of these materials looked good from one direction, but on more three-dimensional forms they looked fairly scrappy or broke the illusion altogether. So I started working on a volumetric material which would better emulate a tree's growth pattern.
By defining the properties of an imaginary infinite tree in world space, objects within that volume can appear as if they are cut from it. Planks can each be given a different volume offset so that they appear to be cut from different parts of the same tree.
![](img/wood-solid-2.jpg)
![](img/bowl.jpg)
There's plenty of space for improvements here - one feature I was keen to work on was chatoyance (cat's eye effect), a type of subsurface reflection which is partly responsible for the appearance of depth in wood. I since found this paper which covers some similar approaches, and has given me some more ideas if I return to this.