[WIP] Not-a-Glock — Game Ready Asset Optimization (feedback is appreciated)
Goal
To improve my understanding of non-destructive game asset optimization and topology, I'm modeling a pistol with the intent to work effectively in VR. I like this as a stress test of my current skills, as guns are perhaps the most crucial asset in a visually appealing shooter. VR is notorious for it demanding constraints, with systems like the Quest using mobile chips to render everything twice. This combination requires an artist to consider how to best utilize low polycount models and efficient image textures to ensure the pistol is visually appealing, without straining the computational budget. I am working in Blender 4.1
My current poly count is 4,264 Triangles, not the lowest possible, but it gives the pistol a pretty nice silhouette IMO.
I opted for unwrapping by hand, allowing me to mark seams around high detail areas and avoid wasting precious pixels. My normal Image texture is 1024px*1024px
I wanted to take consideration for future revision and implement a strong focus on a non-destructive development pipeline. To do this. I used a geometry-node modifier to Copy the optimized geometry, followed by a mesh deform. This allows me to go back and forth between modifying the optimized geometry and deforming its corresponded cage; any changes will automatically update between the corresponding objects.
I would greatly appreciate feedback on how I can make this process more efficient/effective. Share you're experiences, opinions, polite jabs. I'm receptive to critique and would rather improve my work than boost my ego
Replies
Here's the latest with a little more personality. Again, any comments would be much appreciated. I'm determined to get better everyday and prove myself as a valuable asset in the industry.
Thanks for taking a moment to look at my progress!
Thank you for the invaluable feedback! It's really encouraging to get response after chipping away at this for longer than I'd like to admit.