Knotwork Program

Page 3: Days 7 to 9

All images on this page © copyright T. Shermer. All rights reserved.

Up to Knotwork Index

On to Day 10 (page 4)

 

Day 9: 14 Nov 2001

Day9a1

I didn't have much time today, so there's only one image. It shows the first of a new class of substances for making threads: SubstanceImage. These substances are simply based on images (earlier substances were based on code that computed the shape of and rendered the snippets; here the code just copies pictures that have been supplied). For this substance (SubstanceImageTrial, a subclass of SubstanceImage), I used two images: a "straight" section (which actually makes sort of an s-curve), and a "turn" section. These images are rotated to the correct orientation for whatever snippet I'm rendering, and have a mask that lets the background and lower snippets show through. I just threw together the two images in Painter; I haven't made any real effort to make them nice. SubstanceImage is currently rather bare-bones and I will probably be spending some time enhancing and expanding it (so that it's easy to subclass into useful substances). I've also been thinking that I should be rearranging the internals of the whole program a bit so as to make it easier on me in the future. Sadly, hours spent rearranging internals don't yield fun pictures...

 

 

Day 8: 13 Nov 2001

Day8a4

Just because I can, I used two separate PixelShaderImage instances: the background is "Disrupted Fractal Number 11" and the threads are "Europa". Aren't you happy I didn't use some other image for the black outline of the threads?

 

Day8a3

PixelShaderImage using "Europa."

 

Day8a2

A test of my new class, PixelShaderImage. This class is a PixelShader which takes its pixels from an image. In this image, I'm using my "Disrupted Fractal Number10" as the source for the color on the threads.

 

Day8a1

SubstanceEarly on a more involved pattern.

 

 

Day 7: 07 Nov 2001

Day7a2

Same pattern as Day7a1, but with an increase in pulliness.

Day7a1

I've added another parameter to SubstanceEarly, which I call "pulliness." The pulliness indicates how much to pull back a u-turn (sLLs or sRRs). The u-turns in the pattern above are at the center of the crosses and the corners of the image. (My apologies for the image quality.)

 

Back to Day 6 (page 2)