Rebuilding and optimizing

 Posted by at 22:40  Add comments
Oct 162012
 

This post is part of a meta-series. Click here for a list of all posts in this series.

I started by adding a few more details to the machinery above the secondary shuttlebay, and beveling/finishing the detailing on the roof of B Deck. Then I spent some time experimenting with different lighting setups, abandoning my three spherical emission objects in favor of three point lights of varying radii, colors, and intensities. I like the final result a lot more, and it’s slightly faster to render. But I also decided that render times didn’t really matter at this stage, since I could kick them off to run over night. To that end, I cranked up my sample count from 1500 to 4000. My 1500-sample renders were taking ~30-45 minutes. This one took 1:40 (that’s hours and minutes, not minutes and seconds), but it fixes a lot of the grain. Not all, but a lot.

In the further interest of reducing the footprint of the model, and the corresponding render time, I decided to start optimizing the grid troughs on the saucer. When I made them, I had just started figuring my way around that particular method. I’ve since refined it quite a bit, but the saucer still suffers from the old method. As such, I’m going through and cleaning up unnecessary geometry in those troughs. I did about half of the work on the upper saucer (the part “above” the sloped middle band) last night, which netted me a triangle reduction on the order of over 55,000, once you factor in mirroring and such. That’s a lot of tiny, unnecessary polygons — and they are, indeed, completely unnecessary; the triangles I removed added no surface detail whatsoever. I think doing this drudge-work will pay off handsomely once I get to the lower saucer, which has even more gridlines than does the upper saucer.


I’ve spent the last several hours rebuilding the saucer grids from the ground up, combining the optimization work I was already doing with new grids that are about 1/3 as wide (and will be 1/3 as deep — 5cm as compared to 15cm) as the previous grids, which will match the panel/grid lines seen on C and B deck. Nothing interesting to show for it yet, and only about 1/4 of the way through optimizing the grids on the dorsal saucer surface so far.  May be a bit before there’s anything perdy to look at.

 Leave a Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>