I’ve been redoing animations for Nucleate using Blender, focusing on this crystalline, forward-looking style. I started by creating a base crystal and then made variations using geometry nodes, which has been working well so far. I’ve finished 7 out of 10 animations, each taking roughly 5 to 8 hours. Most of these are simple top-down 2D images designed to look 3D, which helps keep file sizes down and performance smooth. Each animation frame is about 240 KB, running at 12 fps to fit the pixel art vibe I’m going for.

The splash page animation is way more complex, involving a lot of movement and crystallization effects. It’s been a bit of a headache because it uses Blender’s NLA track editor, which I’m still getting the hang of. I’m also figuring out how to render and composite different layers properly for that. Right now, animations take time per object rather than per track, and I haven’t started using the NLA Editor for independent object tuning yet. There are over 240 frames in total, which is causing slow load times and even some GUI hangs when loading bigger GIFs.

To keep things manageable, I decided to stick with mostly simple top-down 2D images for the animations. I’m planning to test them on both high-end and lower-tier PCs to see how performance holds up. I’m also considering exporting animations as GIFs directly from Blender to speed things up, even though that might reduce quality a bit.

Pricing has been on my mind too. I initially thought about $40-$50 but now I’m leaning toward $30, with a lower entry price during beta to attract early users. I still need to get some feedback and do market testing before locking that in.

On the technical side, I’m still tuning color, contrast, and outlines to nail the crystalline style. The Blender animation pipeline needs some optimization for game design, especially around handling complex animations and large file sizes. I’m not totally confident yet about how bigger animations will affect boot-up times or how unique animations will impact user interest in Watchlight Studio games. Also thinking about adding a feature to permanently delete cached files to improve transcript summarization quality, but that’s still just a thought.

Overall, I’m making steady progress but there are still a few things I’m figuring out—mostly around Blender’s tools, performance trade-offs, and final pricing decisions.

Two-Line Takeaway

Key step forward

The most important forward progress step is completing 7 out of 10 animations using Blender’s geometry nodes and focusing on a crystalline, forward-looking style that balances file size and performance.

Significant barrier or risk

The most important barrier or risk is managing the complexity and performance issues of the splash page animation, including slow load times, GUI hangs, and the challenges of using Blender’s NLA track editor effectively.

Keynotes

Decisions & Observations

• Decided to focus on simple top-down 2D crystal-style animations to keep file sizes small and performance smooth.

Issues & Next Steps

• Needs to resolve challenges with Blender’s NLA track editor for complex splash page animation, including rendering and compositing layers properly.
• Will explore exporting animations as GIFs directly from Blender to speed up workflow, balancing quality and efficiency.
• Planning to test animations on both high-end and lower-tier PCs to evaluate performance and load times.
• Considering lowering app price to around $30 with a beta discount to attract early users; requires market feedback before finalizing.
• Needs to optimize Blender animation pipeline for game design, especially to handle complex animations and large files without GUI hangs.
• Will continue tuning color, contrast, and outlines to perfect the crystalline visual style.
• Uncertain about impact of larger animations on boot-up times and user engagement; needs further assessment.

Generated with Nucleate by Watchlight Studio — distilled from full transcription.