I’m currently rebuilding core systems for an in-development MMORPG (Magickness™), focusing on long-term scalability rather than short-term implementations.
This phase has been centered around fixing and restructuring the character preview system — specifically how UI, meshes, and animation data interact inside Unreal Engine 5.4.
The Problem
The character preview system wasn’t behaving consistently:
- Mesh alignment issues between MakeHuman base and Mixamo skeleton
- Character preview rendering incorrectly in UI
- Scale inconsistencies across imported assets
- Animation mismatch due to skeleton differences
This created a chain reaction where UI preview, animation playback, and mesh proportions all conflicted.
What I’m Fixing
1. Mesh + Skeleton Alignment
Using a Mixamo-based skeleton with MakeHuman meshes required:
- Rescaling mesh assets to match skeleton proportions
- Verifying bone hierarchy compatibility
- Ensuring consistent import settings across all assets
Small inconsistencies here break everything downstream.
2. Character Preview UI Issues
The preview system itself needed adjustment:
- Camera positioning inside the UI
- Mesh anchoring and offsets
- Consistent scaling so characters render correctly across all customization states
This is one of those systems where visual bugs often come from underlying data inconsistencies.
3. Animation Pipeline Stability
Animations imported from Mixamo required cleanup:
- Verifying retargeting settings
- Adjusting animation playback within UE5
- Ensuring animations match updated mesh proportions
This is especially important early — animation pipelines become painful to fix later if the foundation is wrong.
AI-Assisted Workflow
I’ve also been integrating AI (Claude) into the workflow — not as a game feature, but as a development tool:
- Reviewing system logic
- Identifying inconsistencies in setup
- Accelerating iteration when debugging pipelines
Used correctly, it acts more like a second set of eyes than a replacement for engineering.
Why This Matters
This is part of a broader foundation phase.
Instead of building forward on unstable systems, I’m:
- Fixing pipelines early
- Standardizing asset behavior
- Building systems that can scale as the project grows
It’s slower upfront, but prevents exponential problems later.
Project Context
Magickness™ is a discovery-driven MMORPG built around:
- System-based progression (not predefined classes)
- Player experimentation and mechanic combinations
- Emergent gameplay systems
This stage is focused entirely on making sure those systems can actually support that vision.
Dev Log
If you want to see the actual implementation and issues in real time:
Site
If you’re working with UE5 + Mixamo + custom meshes, I’d be interested to hear how you’re handling pipeline consistency — especially around preview systems and scaling.
Top comments (0)