Week Summary - BA1b Week 13(+)

Time Machine

I finished all my UV maps and applied them to my model this week, meaning all I have left to do for the time machine at this point is put everything together for submission. I had some issues with my normal map that I couldn't identify the cause of, as each section was baked separately, so instead of trying to fix it during baking I just painted the problem areas flat in photoshop, and it seems to have fixed the problem.

I also turned the edge padding down when baking to stop the overlap issues. Below is the finished normal map.

My other maps, in particular the albedo and emissive maps, were something of a trial-and-improvement process. I had a rough idea of what I wanted the colours to be, and as I was using fill layers and masks for the maps it was easy enough to quickly change the colours, make a new map and refresh marmoset to see the effect. This process could possibly have been made quicker by doing more detailed, colour concepts, but seeing it in front of me on the finished model was very reassuring and satisfying, and didn't take too much longer if at all. As for the other maps, seeing how my changes affected the model in close to real-time was a really helpful learning process, and again, didn't take too long. It offered me a very quick way to experiment with things like applying texture masks to my maps, and I think I learnt a lot from the process.

Example of photoshop file structures

I applied a very high-contrast rust texture to my metalness map to add some character to the spinning top section. There are some grey sections on the map, but I raised the contrast to make it largely black or white.

I did realise that I hadn't made an ambient occlusion map for my tree pillar, and ran across the same problem again that I'd thought I'd fixed when baking the normal maps for it - xnormal would analyse the mesh forever without making any progress. After running through several troubleshooting solutions with a tutor, it started producing solid white occlusion maps among other issues, so I re-exported all my meshes, which was apparently enough to fix everything. There have been a few problems like these over the course of the project, which I think is to a certain extent a given in this area; I hope with experience I'll be able to start fixing more of them myself instead of relying on outside help but for now I'm fairly happy with the state of these projects and in seeing how far I've come in 3D.

Finished model