Tech

Summer 2025

Over the summer I began experimenting with different software to familiarise myself with other animation pipelines.

---

Project 1: 24hr AnimChallenge - Agora Community

For my first project, I decided to team up with a fellow undergraduate studying animation at Bournemouth University. He and I had stumbled upon a challenge posted by Agora Community. The challenge: create an animation in 24 hours based on three words - TIME, CHAOS and SWITCH.

We joined the competition with only a day left to spare.

So as soon as we received the brief, we instantly got started on storyboarding.

We assessed our options with the brief and decided that the setting of an underground train station would suffice for CHAOS and TIME. From there we worked out timing and key narrative points. SWITCH would come in later as our gimmick (we wanted to incorporate a time dilation/freeze moment).

Once pre-production was sorted, I tasked myself with finding rigs to use from the Agora Community page and got started on posing a handful of them to act as a crowd.

Everything was going well, however, we had not anticipated how much time would go into each shot. What ended up happening was the both of us getting burnt out very quickly and having stayed up all night working on this, we had lost our vision.

Still, the takeaway from this project was more valuable than the chance of winning, learning to cope under pressure with the reality of animating to an extremely tight deadline.

What were my strengths? What were my weaknesses?

My strengths were being able to communicate with my team effectively and to be able to demonstrate quick problem solving skills (beneficial to my workflow).

After missing the deadline, we decided to call it quits on the project as we had lost our love for it - this'll happen if you stay up all night working on something for 24 hours.

Here is the result for my contribution to the project:

24 May '25

---

Project 2: MB12 Productions

Experimentation with new software began when a friend of mine studying Film at Royal Holloway University approached me with an idea for a short film series he was hoping to direct. Keen on the idea, I asked him if he'd be interested in having a studio ident done for him since he'd mentioned he would be sticking with the same cast and crew throughout. He was delighted to receive the offer so he quickly gave me a rundown on how he wanted it to look.

The brief was simple; he wanted:

• Old school type font

• Typewriter noises

• Moody lighting

Stoked, I became very excited to jump straight in. However, there was one problem.

I knew I didn't want to produce it in Maya because I wanted to expand my skillset.

What did I do ?

I opened Blender.

I watched a couple of relevant tutorials to help me achieve the vision my friend and I had established.

The first tutorial I watched inspired the way I would incorporate moving text in the sequence. https://www.youtube.com/watch?v=3ykQbF2j5zw&pp=ygUWYmxlbmRlciB0ZXh0IGFuaW1hdGlvbg%3D%3D

Once I had the idea for how I wanted to animate the text, the rest of development was devoted to how I wanted to evoke an 'old-school' vibe.

And it hit me right then and there. To deliver this animation I had to focus on the lighting heavily too.

For lighting reference, I looked at the music video for Hatchet by Archive as I had noted in the past that the way the lights move around in the scene really give a dramatic and atmospheric tone to the video.

Eventually I was able to produce a first draft for the studio animation, and here it is:

08 June '25

After finishing this draft, I knew there was room for improvement.

I realised the graphics were poorly positioned in the frame and the camera movement had janky keyframes. To solve this, I went back to the drawing board and started brainstorming to generate some concepts for title card visuals.

After deciding on a new visual, I communicated with the director and we both agreed that the second graphic was visually a lot better.

Allowing myself time away from Blender for a couple of weeks allowed me to come back to the project and see it again with fresh eyes. This time around, I went ahead and changed the position of the lettering and reworked the camera movement, also incorporating HDRI lights to illuminate the scene better as my main criticism from before was that the letters were too dark. As of this stage, the camera movement is not remotely finished.

Experimenting with Depth of Field was also super helpful as I believe this will be necessary for my projects during the semester.

28 June '25

Since I was working to a time crunch, my current workflow with animating in Blender is very unrefined, with my methodology for creating a sequence of shots being to render each shot and compile them in Premiere Pro. Had I had more time, I would have looked into what kind of camera sequencing is available in Blender and used that approach instead.

---

Semester One | September - December 2025

Over the course of this semester, I will be exploring new tools to help improve my workflow as a 3D animator. These tools will range from plug-ins I have found to help make certain tasks easier to save time; experimenting with simulations; to considerations of how I can implement AI in my work to improve workflow.

Portfolio Task 1 - Body Mechanics (MAYA) - Workflow & Tools

This tool allowed me to track the arcs of my animation and see if the movements are readable. The example above shows my early experimentation with the Blue Pencil to plot the arcs of my character's weak hand to ensure its behaviour isn't overlooked in conjunction with the movement seen with the strong hand throwing the grenade. This tool is great, and I'll definitely be using it whenever I am animating in future as it's improved my workflow for animating body mechanics

---

PLP Task (BLENDER) - Camera Sequencing

In this video, I generated a basic squash and stretch animation to test Blender's in-built camera sequencer methodology. This process involves having my cameras set up, then creating markers on the timeline for when I'd like the camera to switch, and binding the correct camera to its corresponding marker. This process is effective at what it does, but I am unhappy with how time consuming it is. Overall, I am not a fan of this method, compared with the camera sequencer in Maya.

---

Portfolio Task 2 - Previs (BLENDER) - Video Sequencer Tool Experimentation

In this video, I showcase my live experimentation with learning a new tool while working on my portfolio task for previs. After being unhappy with the camera sequencing method Blender has built in from my previous experimentation, I decided to look online for tools that could help me speed up my workflow. That's when I discovered Tin2Tin's Scene Strip Tools. This tool made my process of generating shots previs a lot easier, considering I was attempting to create something in software I'd hardly used before. The process here, as opposed to the default methodology, is having a camera selected in my outliner and running a python script that allows me to insert my selected camera as a scene strip in Blender's Video Sequencer Editor (VSE). No more faffing around with timeline markers!

---

Portfolio Task 2 - Previs (BLENDER) - Non Linear Animation Editor Tool Experimentation

Here is a screen recording of me implementing use of the Non-Linear Animation Editor for the first time. In this instance, I was stuck trying to shift the keyframes of my camera across my timeline. I attempted to move the keyframes using the graph editor but due to a lack of accuracy while plotting them, the camera started to behave strangely. I knew what I was trying to achieve, so I looked online for assistance and came across the Non-Linear Animation Editor (NLA Editor). This approach enabled me to edit my keyframes the way I intended them to be, then subsequently opening the NLA Editor and turning the range of keyframes into a block that I was then able to manage accurately on my timeline.

---

Portfolio Task 2 - Previs (BLENDER) - Non Linear Animation Editor Bug

In this video, I encountered a major bug with the NLA Editor in Blender. After my sequence finished rendering in Eevee I discovered that my rig randomly disappears in the shot it's supposed to be visible in (close up of the feet as he's climbing the stairs). I could not figure out what was going wrong, as the rig was visible in the 3D viewports in the same shot. After nearly pulling all my hair out and losing hours time scouring the internet for solutions, I discovered through last ditch efforts involving me trying random things that it was a bug with the NLA Editor. After I solved that, I encountered an issue with the rig's geometry extruding really strangely for no apparent reason. I ended up going back to a previous save and redoing the animation. I still do not know what was causing the issue with the rig's visibility in the NLA Editor, however, I do know now that if I encounter similar issues like this in future I can try poking around in the NLA Editor again.

---

PLP Task - Lighting Tutorials (MAYA) - Viewport 2.0 Lighting

Lights in use: Ambient Light + Directional Light

Lights in use: Spot Light + Arnold Area Light

Lights in use: SkyDome HDR + Directional Light

---

Portfolio Task 3 - Previs (MAYA) - Workflow

Above is a screen recording of my first block pass. At the top of the screen, you can see I have my timeline snapped to the top of the screen -- I did this to improve my personal workflow as I found my eyes were being strained looking at the bottom of the screen for too long. Keeping the timeline at eye-level meant that I could work for longer periods at a time. I also prefer the look of having my different workspaces organised, with my viewport having its own dedicated space on the screen rather than being centred around everything. In terms of my workflow generating this sequence, I began to block it out using a master shot. I intend for this shot to act as the main coverage of the scene, with cuts to different cameras that will capture certain action elements that deliver my major story beats effectively. My workflow for this project involved me animating the vehicles in splined + linear tangents to understand timing easier. Since the cars are constantly moving, I needed Maya to interpolate the frames that I hadn't keyframed for reference on how the cars should interact with each other, something I didn't think was possible in stepped tangents.

Using concepts from Alan's Viewport Lighting tutorials, I incorporated Screen-Space Ambient Occlusion and Multisample Anti-Aliasing to reduce pixelation on the geometry in my viewport. I also enabled Depth of Field for certain shots to draw the focus on the main action in the scene, keyframing this where necessary to maintain that focus.

Using my research into Speed Racer (2008), I borrowed camera techniques for certain action elements where the car is speeding on the road, such as keyframing a high focal length and quick-panning to give the impression the camera has real-world behaviour.