Texture Acquisition Techniques

Manual Techniques

Manual techniques started by painting pixel by pixel (referred to as pixel art). These techniques evolved and nowadays there are numerous sophisticated painting tools available for Manual texture creation.

• Adobe Photoshop
• The Foundry Mari

Process usually involves finding different textures online and merging them together while painting over.

Manual techniques are not used as much anymore.

Procedural & Generators

Procedural textures creation is creating texture with math. Just as we can create lines and curves with math, 2d and 3d parametric nodes can generate range of useful functions from simple noise till procedurally created wood.

Advantage of such procedurally created textures is its infinite resolution and variations. Disadvantage is limited pre-made nodes and possible lot of time spent to get a texture of needed fidelity.

• Substance Suite
• Quixel Suite

Another powerful feature are geometry aware generators that can for
example generate dirt texture around edges of 3d object.

Procedural & Generators are the most popular texture creation methods right now.

Optical Capture

While manual techniques and procedural / generators can create desired texture, these techniques are many times not suitable if the goal is to digitize or represent surface one to one - or in other words if the goal is to create Digital Twin.

That is the difference as optical capture is currently the only way of getting true texture from reality to virtuality - while removing subjectivity. Manual and procedural techniques are suitable for cases when direction and creativity are more important.

Simple Photograph

A simple photograph is suitable for some surfaces - especially big ones with homogenous reflections. Using color checker and diffused flat lighting is important. Processing involves gradient removal and texture has ambient occlusion baked in that can lead to less precise virtual representation.

Taking simple photography is probably the fastest way of texture acquisition. Disadvantage is lack of any other maps like normal, reflection and glossiness and therefor this technique is suitable only for limited range of surfaces.

Photogrammetry

Photogrammetry is a technique of taking multiple photographs of an object / surface and then computing its 3d shape and 2d texture with trigonometry. Textures are dependant on environmental lighting however there are multiple techniques like flash cross polarization and color checker that can ensure texture is fairly precise.

Reality Capture

Maps like reflection and glossiness are usually guessed and derived from diffuse and normal maps. Photogrammetry requires quite robust pipeline and workflow but provides very good quality data and true height map.

Surface Scanning

Compared to other techniques, surface scanning provides widest range of maps in shortest amount of capture and processing time.

TMAC

Disadvantage is that this technique is suitable only for nearly planar surfaces.