Week 38: Optimising Edge Flow

This week we have been focusing on optimising our models, working together in groups and cleaning up each other’s objects. It’s a step easily overlooked at times, but certainly among the most important. Despite living in a time of unprecedented information storage it still has its limits. And not everyone can afford the cutting edge. Without optimisation you’d exclude a large potential player base simply because you didn’t bother making models in your game less information-packed with polygons.

Science Fiction Crate by William Mayer

Before

After

All n-gons appear eliminated. The lines are now in greater structural detail and you can see the outline much more clearly. While its design was relatively simple from the beginning there were a great number of vertexes along the rounded bars on its roof. On one of the corners the lid wasn’t as evenly broadened as along the rest of the box. While remedied it was still unable to acquire the continuous depth same as the other sides. This was no doubt my own fault back when I first created the crate, as I possibly did not thoroughly ensure that all sides were extruded in unison. The uneven corner likely only complicated it further and the chamfer had made it even more difficult to properly link it with the rest of the lid.

Cartoon Crate by Andreas Calmius

Before

After

The cartoon crate was considerably simplified. Some polygons on the sides removed due to being over abundant for the purpose of maintaining its form. Some loose vertices were also discovered and welded together accordingly. The n-gons that were spotted (possibly created when adjustments were made to the edges) were rectified with the help of the cutting tool.

Urban Crate by Myself

Before

After

A simple design, but not without its issues. Connecting the vertexes became a problem early on and several of them would not respond to commands. There were openings in the structure’s corners as well which could not be filled with the cap poly for some reason. Apparently there was some sort of second layer which interfered with the outer one. I eventually managed to fix it using the cutting tool after having redone the entire face of the corner. Apart from that there wasn’t much else to do with most polygons consisting of the required amount of vertexes. There were some stray vertices to remove at places, but nothing major.

Looking back on the work I did, I’d say I could have done a lot better. I spent days optimising the crates without the best understanding of the tools available to me. I should really have paid more attention to this week’s class, which was difficult due to a lack of sleep. Nothing seemed to work, with connecting loops not always working and it wasn’t until late into the weekend that I discovered the usefulness of the cutting tool. While I was able to complete it in the end it did take much longer than it should, and I quite frankly exhausted myself spending countless hours trying to make the software do what I told it.

For future reference, I should put more emphasis on being consistent. Whenever I can make a wholly encompassing structure without differentiation I should do so, as otherwise I just risk it becoming more difficult and time consuming to optimise the model later on. Such as having to look at every face individually to make sure you didn’t miss a stray polygon. It all really comes down to creating a method for the construction of your object, which I have yet to master. Simply means that I’ll have to keep practising.

That’s all for now, until next time!

Björn Erik Berndtsson. Graphics and Game Design.