Pt. 18 – Production Setup

Image: Firefly 2026

Technical Setup:

Hardware:

MacBook Pro M3 Max with 64GB RAM and 4TB SSD. Additionally, I have multiple external 2TB SSDs for each facet of my work:

• Design WIP
• Animation
• Video
• Photography

My setup handles Blender’s viewport performance well, though final render times for scenes with heavy geometry and lighting will still be substantial enough.

Software:

Blender 5.0 for all 3D work (modelling, rigging, animation, rendering), Adobe After Effects 2026 for compositing and colour grading, Adobe Audition 2026 for sound design and mixing.

Backup:

Daily backups to a SanDisk 8TB external SSD. Given the project file sizes and the catastrophic consequences of data loss this close to deadline, this is non-negotiable.

File Organisation:

My project structure follows the production pipeline, with numbered top-level folders for each phase:

1-PreProduction contains the animatic, planning documents, reference imagery, research materials, storyboards, and resources.

2-Assets splits into acquired assets (the car models from Sketchfab, etc.) and assets I’m making myself. The made assets folder subdivides into Characters, Environments, and Props – each with their own blend files and texture folders.

3-Scenes will hold individual scene files. Each scene file references the assets rather than duplicating them, keeping file sizes manageable.

4-Renders will store rendered image sequences organised by scene, plus a titles/credits folder.

5-PostProduction has separate folders for After Effects projects, audio work (split into FX and music), and final output files.

6-Documentation so far contains blog imagery, screenshots for presentations, thesis writing, and thesis references.

This structure means I can find any file quickly and understand at a glance where I am in the pipeline. It also makes backing up straightforward – the entire project lives in one parent folder.

Workflow:

The production pipeline breaks into three clear phases:

Pre-production (complete): Script – Storyboards – Animatic – User Testing. This phase established what needs to be built and validated that the story works before committing to production.

Production (current): Character modelling – Rigging – Animation blocking – Props and environments – Final animation – Lighting.

I’m working through this systematically, though I’ve realised I need to adjust my approach.

Post-production (upcoming): Rendering – Compositing in After Effects – Sound design in Audition – Final export.
The colour desaturation effect will be handled in compositing, giving me flexibility to adjust the progression without re-rendering.

Project Management:

I’m using Trello to manage tasks across a Kanban-style board with seven lists: Backlog, This Week, In Progress, Next Week, Testing/Review, Blocked/On Hold, and Complete.

Each card represents a task with due dates, checklists for subtasks, and labels for work type (character work, props, environments, animation, technical tasks, admin, writing). The progress indicators (0/4, 5/5, etc.) show checklist completion at a glance.

I use Trello for my client work as well, so this system reflects how I actually work rather than adopting project management tools specifically for thesis purposes. The workflow is established and familiar, which reduces anxiety.

The “This Week” and “Next Week” columns show my immediate focus. I plan work in weekly blocks rather than formal sprints – it’s more realistic for a solo project where tasks don’t always fit neat two-week cycles. Cards move right as work progresses, and the “Complete” column shows what’s actually finished rather than just “in progress.”

Priority Shift – Blocking First:

My next task is due to be modelling the age variants (50s, 70s, and 90s John). I’m now shifting priorities to block out the entire animation first, using 30s John as a placeholder for all ages.

The reasoning: Martin White, who’s composing the music, needs timing and pacing to work against. He’s based in Portugal and has been without power for nearly two weeks after the tornado, so he can’t work much at the moment anyway. But when he can, he needs scene lengths, emotional pacing, and timing for key beats. Rough blocking gives him that – he doesn’t need finished character models or polished animation to compose.

Blocking means literally placing grey boxes for walls and buildings, sliding characters across floors without walk cycles, moving cars without rotating wheels. I’m establishing what happens, when it happens, and how long it takes. If I find a suitable house model for exterior shots, I can drop that in as a placeholder. All the polish – proper walk cycles, wheel rotation, detailed models, final lighting – will all come later.

This approach also means I’ll know if the story actually works in 3D. The animatic validated the concept, but blocking will reveal any pacing problems, shots that don’t work and need adjusting, or scenes that need cutting. Better to discover that now than after I’ve spent weeks on it.

The age variants, of course, matter very much for the final film, but they’re not needed for blocking. I can swap different character models into the blocked animation later without breaking timing or camera work. This sequential approach also tackles the less intimidating work first – blocking is more straightforward than character ageing – whilst preparing what Martin needs as soon as his power returns.

I’ll be discussing this priority shift with my supervisors tomorrow to make sure it aligns with their expectations and the assessment timeline.

Balancing Thesis and Freelance Work:

The biggest challenge is managing thesis work alongside ongoing freelance and client commitments. I can’t just stop taking client work for months – I still need to pay bills.

My approach: client work gets priority during office hours – they’re paying the bills and have their own deadlines. Thesis work fits around that – when I have capacity during the day, and in evenings and weekends when client work is quieter.

The Trello board helps here – I can see at a glance what’s critical this week versus what can shift when client deadlines hit. The weekly planning approach means I’m not rigidly committed to hitting specific daily targets that might be unrealistic given fluctuating client demands.

This isn’t ideal – a full-time student would have more focused time – but it’s the reality of postgraduate study whilst running a business.

Other Production Challenges:

Solo production timeline management:

Every task is on me. There’s no team to fall back on, no one to pick up tasks if I’m stuck. This means being realistic about scope and ruthlessly prioritising what actually serves the story versus what would be “nice to have.”

Render time optimisation: Final renders will be time-consuming. I’m building this consideration into decisions now – lighting setups that render efficiently, geometry levels that balance quality with render speed, using instances where possible rather than duplicating heavy assets. Test renders early in the week inform whether I need to simplify before committing to full scene builds.

External dependencies: Martin’s situation in Portugal highlights the risk of external dependencies. I can’t control when his power returns, but I can make sure he has the materials he needs for when he’s able to work again.

Current Status:

Week 2 of a 15-week production schedule. Character rigging complete for 30s John (body and facial), IAM badge modelled and shaded, four period-accurate car models sourced and standardised.

Shifting priorities this week to block out all scenes using placeholder geometry and 30s John as a stand-in for all ages. This will establish timing for Martin’s composition work and reveal any structural issues with the animation before committing to detailed asset creation.

Slightly behind the original schedule on age variants, but the blocking-first approach should actually save time by preventing wasted work on assets or animation that might need cutting. Meeting with supervisors tomorrow to discuss this adjusted workflow and get feedback on the approach.

The foundation is solid – planning and early asset work are done properly. Now it’s about maintaining momentum through production whilst staying flexible enough to adapt when circumstances (like Portuguese tornadoes) require priority shifts.