Optimizing 3D

This week we have been optimizing our crates that we created last week. We were divided into groups of three and all got to optimize a crate made by the others and one that you made yourself. There is not much I can say or anything new to show of my medieval and my urban contemporary crate as my colleagues that were tasked with optimizing them found nothing to fix.

I believe that this is because of their simple natures, the medieval box, which Anita Stenholm was assigned, is basically just a box on which I have used extrude, I kept the vertices and edges few and did not remove anything, which could mess up the grid. 

The Urban dumpster that Rebecka Nyström was in charge of, is also very simple, just four standard primitive boxes and a cylinder placed and shaped to the silhouette of a dumpster. One thing that I could possibly do if I was to put it is a game is to decrease the amount of horizontal edges on the handle. The camera would probably not zoom in so much that it would need to be as round as it is.

The Sci-Fi crate that was left for me was another story though, it felt somewhat like a mess created by someone who has pretty much no idea of what they are doing. It took a lot of time and puzzling until I found out a way to fix the problems, which were a ton of n-gons. Although it was very much straight forwards after that, it took quite some time getting all tiny pieces to come together.

I had also managed to create some overlapping faces and there were some edges that had magically disappeared or moved. There was also a few overlapping vertices snuck in while I was trying to fix everything.

  

Due to the use of chamfer on almost all edges, I had ended up with all of those n-gons, the solution I used to fix them was to add an edge through the corners of my chamfered edges, ending in the corner of the adjacent squares. After that I target welded all the vertices in the corners to the centre edges now running straight through them. I repeated that for every corner, one side at a time, to finally check if there were any n-gons or overlaps left. I could have left the loops all alround the crate but it would result in many more faces than this solution.

Sometime when starting out I had managed to remove a large part of the inner square on one of the sides. Since it was not possible to undo so far, I made the best of the situation by finishing optimizing all the other sides, placing an edge straight through the centre and mirroring it. Sine I have details on my crate that has vertices very close to each other, I was unable to simply weld the two together. This meant that I had to target weld every single pair of vertices along the seam. 

The problem I had with three overlapping surfaces turned out to be due to a few overlapping vertices. Once that was taken care of the problem went away. I have learnt today that starting out with fixing any overlapping vertices first as it may help to solve many other problems instantly.

What I should have done was to wait with removing things until after today’s class. Since I did not know what would happen, I managed to create a lot of  n-gons, which I did not know what it was either. 

In the future I think that I will be able to avoid many of these mistakes, simply by now having learnt some of what problems can occur and how to find them. I should also complete my model before starting to remove things, or at least make sure that I have some clue as of what I am doing.