About this Knowledge Base
Introduction​
This knowledge base is not a random collection of tutorials. It is a structured curriculum designed to bridge the gap between abstract 3D concepts and concrete investigative needs. Understanding how this site is organized will help you navigate the content efficiently, whether you are building a complex simulation or simply verifying a claim.
Key Takeaways​
Site Architecture
Four layers: an orientation section, a Foundation course with four modules, topic courses, and a reference library of tools, add-ons, and techniques.
Design Logic
Built on First-Principles: We focus on the "why" before the "how" to give a better idea of the context in which certain topics belong.
Site Architecture​
The content is organized into a modular file structure, allowing you to switch between linear learning and reference lookup.
docs/
├── intro.md # Entry point: Blender for Visual Investigation
├── introduction/ # Orientation: how to use this guide
├── courses/ # Curriculum layer
│ ├── index.md # Course catalog
│ ├── foundation/ # The baseline course
│ │ ├── syllabus.md # Course front door (Course Syllabus)
│ │ ├── 01-the-3d-world/ # Module 01 — Thinking in 3D
│ │ ├── 02-methodology-and-data/ # Module 02 — Research gate and data lifecycle
│ │ ├── 03-modeling/ # Module 03 — Modeling fundamentals
│ │ └── 04-blender/ # Module 04 — Blender training
│ ├── gaussian-splatting/ # Topic course: Gaussian Splatting for investigation
│ └── site-reconstruction/ # Topic course: site reconstruction
├── tools/ # Reference library: external tools
├── addons/ # Reference library: Blender add-ons
├── techniques/ # Reference library: methodologies and workflows
├── use-cases/ # Applied knowledge: real investigations
├── templates/ # Page templates (internal, not published)
└── archive/ # Retired content (internal)
Each Foundation module has the same shape: index.md is the Module Syllabus, introduction.md is the first read, and the remaining files are the module content. Topic courses use the same pattern at the course root.
The Design Logic​
I have structured the information based on First-Principles. Where most tutorials jump straight into "click here, press that", my aim is to help you build understanding before action and give you strategies to solve specific problems rather than handing you a step-by-step solution. It is inevitable that some processes require a step-wise approach, but I try to minimize this as much as possible. In the hope that when you encounter a novel problem, you can reason through it and use this knowledge base as reference.
Bottom-Up Progression​
| Layer | Purpose | Why in this order |
|---|---|---|
| 1. Foundation course | Theoretical and tool baseline | Four modules: Thinking in 3D, Methodology & Data, Modeling, and Blender. Every later layer assumes them. |
| 2. Topic courses | Applied curricula | Curated paths compose shared modules (Measurements, Camera Matching, …) into courses such as Gaussian Splatting and Site Reconstruction. |
| 3. Reference library | Tools, add-ons, techniques | Look up "how to do X" once the foundation is set. It is written to be searched, not read in order. |
| 4. Use cases | Applied synthesis | Real investigations in which the layers above come together. |
Linear/Non-Linear Approach​
While the Introduction and 3D Context modules are mostly linear in setup, the Tools, Addons & Techniques section is designed as searchable databases.
- Non-Linear Access: You do not need to go through this in order. If you need to know how strategies on how to match a camera perspective, you can jump directly to that technique.
- Categorization: Techniques are grouped by methodology (e.g., Precision Modeling vs. Direct Modeling), tool name and tags.
- Modularity: Each technique page is self-contained with its own resources, meaning you can solve specific problems without reading the entire history of the software. Basic knowledge of Blender is required however.
Teaching Philosophy: Geometry First​
You will notice that Module 1 does not start with Blender's interface. It starts with Geometry and Dimensions.
We deliberately separate the Concept from the Tool.
- Concept: A 0D point has no length, width, or height.
- Translation: In Blender, this is a Vertex.
- Concept: A 1D line connects two points.
- Translation: In Blender, this is an Edge.
By understanding geometry basics first, you learn to think visually and structurally. The idea is that when you eventually open Blender, you see how geometric understanding translates into Blender's world.
A Living Resource​
Origin​
When I started this project in 2024, the world looked a bit different. I wanted a place that brought together my findings, research, and workflows in one spot and I wanted the ability to share that. Sources that would teach Blender and/or visual reconstruction always felt too fragmented to me, and I wanted to fix that.
So this knowledge base is an aggregator: a mix of videos, information, and whatever I find. I want to capture, share, and use it for context, alongside more elaborate workflows and pages.
On AI-Assisted Writing​
I used to write everything myself, but at this scale, and at the pace at which technology is progressing, that is not possible for a one-man band. Also, I don't see the point of that anymore for this context. I feel my own writing tends to drift into anecdotal fluff that no one's waiting for. The goal now is content that's readable and optimized for both humans and agents. Less opinion, more facts and as much 'straight to the point' as possible. I have set up rewriting skills that aim for that, and apart from placeholder content (see below) everything has been checked, verified, restructured and edited by me. The traffic light at the top of each page indicates its status. The goal is that this knowledge base can serve as a very decent contextual as well as practical starting point for both you and your agent.
I don't mind AI writing. I think of it like VFX in a movie. I know the spaceship isn't real, but as long as it doesn't break the immersion, it doesn't matter to me. It only becomes a problem when it takes you out of the flow and you start to notice it. Same goes for AI writing and this is an ongoing process.
Placeholder Pages​
Some pages are placeholders, waiting to be filled in and reviewed when I find the time. I still think they add value to the knowledge base, even unfinished.
Accuracy is a collective effort. If you spot an error, have a question, or want to share a workflow technique, please reach out. Collaboration helps sharpen the tools for everyone.
Let's get started.
Roadmap​
This knowledge base grows with its field. Here is where I am focusing next.
Courses in Development​
Gaussian Splatting. A complete course on scene reconstruction with Gaussian splatting — from capture and processing to Blender integration. A prototype of the full course structure is in development.
AI Mesh Generation. A course on generating 3D geometry from text and images. It includes hands-on evaluations of tools such as Trellis, Hunyuan3D and Pical3D, and how they fit into reconstruction workflows.
Learning with Local AI​
I do not yet have the compute for a personalized 1-on-1 AI tutoring service. The educational side works in two directions instead:
- A knowledge base built for agents. Point your own AI assistant at this resource. The content is structured and verified so it serves as a solid corpus of context for the workflows described here.
- A course on running local AI. Set up capable models on your own hardware. A local model can support you in two ways: it assists with the workflows on this site, and it can act as a training companion while you learn.
Addon Development​
I build dedicated addons for specific parts of investigative workflows. These tools come out of real cases and fill the gaps where general-purpose Blender tooling stops.
Summary​
This is a first-principles curriculum: foundations → courses → reference, with deliberate tracks for builders and analysts. You can now navigate it with intent.
Resources​
- Blender for Visual Investigation — the entry point
- Thinking in 3D — geometry first
- Blender Basics — tool training
- Methodology & Data — the research lifecycle