Maintaining Character Consistency Across Style and Environment
Testing whether a consistent character identity can survive multiple art styles, environments and motion systems.
What I was testing
This experiment began with a personal brief. I wanted to announce that I was leaving Amazon after nine years. Rather than write a standard post, I chose to create a short animated piece that illustrated a journey.
The creative concept was simple. A character moves through different environments and visual styles, symbolising change, growth and the unknown ahead.
The technical challenge sat underneath that idea. Could I maintain a consistent, recognisable character across multiple art styles and motion systems using generative tools?
The goal was not just to produce a polished video, but to stress test how well AI workflows can support narrative continuity, stylistic variation and motion without losing identity.
The questions I was asking
- Cross-tool consistency between image and video generators
- Prompt refinement for consistent character outputs
- Style shifts without breaking recognisability
- Motion realism versus stylised animation
- How much manual correction is required to preserve identity
How I approached it
Phase 1: Building Controlled Character Reference Sheets
I began by using real images of myself as base material. From these, I generated multiple controlled 3D-style character reference sheets including a Pixar-inspired version, a cinematic version and neutral front-facing reference renders.
Each sheet defined facial proportions, hair structure, colour palette, skin tone and lighting direction. The goal was not to generate a cool version of me. It was to create anchor assets that could stabilise identity across generations.
Phase 2: Structured Prompt Libraries
Rather than writing fresh prompts each time, I built reusable prompt libraries organised by category. Character base prompts defined core physical traits. Camera angle prompts covered 38 different shot options to create more cinematic results.
Phase 3: Video Stitching and Continuity
To extend sequences, I experimented with using the final frame of one generated clip as the input reference for the next. In theory this should have preserved visual continuity.
In practice, the model treated each new scene as an entirely independent generation, working only from that single frame. This led to subtle facial drift, proportion inconsistencies, lighting degradation and loss of fine character detail.
After multiple iterations, I refined the workflow by combining the end frame of the previous clip, the original character reference sheet and reinforced prompt controls. Reintroducing the character reference anchor reduced drift significantly and restored consistency.
What I took away
- Structured prompts outperform descriptive prompts. Sectioned components covering character, camera angle and lighting separately, then combined into a single prompt
- Reference images dramatically reduce character drift
- Camera libraries create smoother stitching
- Tool selection should be stage-specific, not aesthetic-driven
- Generative systems require governance, not just creativity
- Frame continuity alone is not identity continuity. Character anchors must be reintroduced deliberately at each stage
What came out of it
The project produced a repeatable workflow. One that keeps characters recognisable across art styles, holds proportion and facial consistency through multi-scene stitched video, and cuts down significantly on unpredictable outputs.
More importantly it reinforced something I keep coming back to: generative AI works in creative production when you treat it as a system to direct. That is what keeps the output from being random.