3D Modelling
Learning Blender, animation fundamentals, and game feel through small finished exercises.
This lab is a public log for learning 3D modelling and animation without turning it into a performance. It is here so the process has a place to land: notes, exercises, unfinished questions, and the occasional finished render.
The center of it is simple: I want to understand how motion starts to feel alive. Not only how to model an object, but how timing, weight, squash, stretch, anticipation, lighting, and interaction change the way something feels.
Why This Lab Exists
Some interests come back often enough that they deserve a real place in the week. Games, animation, and tactile software all point at the same craft: making systems that respond with character.
The goal is not to begin with a dream game or a polished short film. The goal is to build a small ladder of exercises that can actually be finished.
Working Ladder
- Finish simple Blender modelling exercises and keep screenshots of the important stages.
- Animate a bouncing ball without a rig, focusing only on timing, spacing, and volume.
- Animate the same idea with a simple squash-and-stretch rig.
- Move to a ball with a tail or legs, where follow-through and overlap become visible.
- Try a small lamp study only after the motion basics feel less theoretical.
Current Notes
The first Blender exercise was useful because it made the tool feel less abstract. Modelling, materials, lighting, UVs, and rendering are still conscious effort, but at least they are no longer only words.
The strongest lesson so far came from lighting. A small scene does not need to model the whole world. A few deliberate lights, blockers, and shadows can make the mind infer a larger space. That feels close to animation itself: suggesting enough that the viewer completes the rest.
The confusing parts are still the basics: navigating the scene, remembering where tools live, and trusting that modelling will become more instinctive with repetition.
Bouncing Ball Plan
The next exercise should stay tiny. One ball, one ground plane, one camera, one export.
First pass: no rig. Animate location and scale directly, keep the volume believable, and use the graph editor to make gravity readable.
Second pass: use a proper squash-and-stretch rig and repeat the same motion. The point is not to make a better-looking clip immediately, but to feel what the rig makes easier.
Third pass: compare against simple video reference. A phone recording of a real ball is enough. The difference between guessed timing and observed timing is the lesson.
Log Format
Each session should leave behind a few breadcrumbs:
- date and time spent
- exercise or object
- screenshot or short export
- what became clearer
- what was confusing
- next concrete step
- where the file or render lives
That is enough. The lab should stay light enough that writing it down does not become a second project.