HOME | DD
lyc — rise
Published: 2007-09-12 23:04:00 +0000 UTC; Views: 14018; Favourites: 86; Downloads: 492
Redirect to original
Description
inspired by the incredible demo from 2001: [link]it was quite a challenge to render this thing, in several ways:
1. the deep fractal iteration is pretty difficult to handle numerically, and getting a normal to this thing's surface (which it technically, in the limit, doesn't have) is pretty ugly.
2. it was rendered with 3 different builds of my renderer (big thanks to *Aexion for help with the rendering!), which means keeping track of lots of system variables and state files in various states of completion...
3. the builds i used to render didn't have proper state-combining functionality, so i integrated that into my latest code which was rendering a different scene by that time.
4. the temptation to stop rendering this and bake some of the newer stuff was immense!
edit: improved image quality, and it's now png
~feckt wanted to render something like this, so i hacked together a little application for him in like 5 minutes that will produce an .obj file of the sierpinski tetrahedron. it takes a commandline parameter for the recursion depth, the default is 4; interested parties can download that here: [link]
Related content
Comments: 90
Nice pic - do ypu still have your app for creating the sierpinski tetrahedron ?
I would like to try it out : -)
👍: 0 ⏩: 1
sorry it's definitely way too old to be usable by now, it was hardcoded for 4 threads and crashes on anything else, besides many other problems. oh, and it isn't even "usable", the exe just runs and spits out this bmp file.
on the plus side, indigo renderer will eventually support implicit surfaces, and this guy will definitely feature in there
👍: 0 ⏩: 1
appreciated 
👍: 0 ⏩: 0
unfortunately i can't oblige, most of my images are once-off (produced by code alone, since modified).
👍: 0 ⏩: 1
do you still have hittest formula? my brain hurts when I am trying to derive those things.
👍: 0 ⏩: 1
thanks, glad you like it 
👍: 0 ⏩: 0
thanks, it's an older work but still quite interesting i hope
👍: 0 ⏩: 0
Still need a little more space above and below, though. NEED.
👍: 0 ⏩: 1
i should do a new version just for you <3
can you gimme a sketch of the composition? the object is basically an icosahedron, it's just very distorted here because of a wide field of view.
👍: 0 ⏩: 1
Yeah, I know what it basically is. I just need to see a bit more of the background above and below it. Like, I feel the inexplicable urge to add 2 cm of background above and below the icosahedron's top and lowest edge, respectively.
👍: 0 ⏩: 0
Ding! - I Return.
This one piece, it keeps haunting me. Pulling me back. Argh.
👍: 0 ⏩: 0
Depth-of-field for the win, or something. Unique, but not abstract, the shape is surprisingly simple, despite looking complex from a distance. Good stuff!
👍: 0 ⏩: 0
It's inspired by the Menger Sponge? We've been studiyng the Menger Sponge at school this year, and it looks symilar
👍: 0 ⏩: 1
whoops, sorry i missed this comment!
yes, it's quite similar: instead of the centre missing from a square (and repeated on the remaining parts), it's the same process starting with a triangle.
👍: 0 ⏩: 1
thanks 
👍: 0 ⏩: 0
thank you
👍: 0 ⏩: 1
great. I quite like the blur, gives it more depth when it only focuses at a certain distance
👍: 0 ⏩: 0
It easy to make sierpinski octahedra and tetrahedra since squares and equilateral triangles tile a plane.
But tiling the plane with pentagons is a different thing altogether. The largest object seems to be a 5 sided pyramid. But it doesn't look like the smaller pyramids meet each other along a line lieing in the pentagon base face.
The top triangle face looks like an isosceles triangle similar to the original but the bottom two seem like shorter scalene triangles.
Wish I could move this thing about in 3 dimensions so I could get an idea of what's going on.
👍: 0 ⏩: 1
thanks, glad you like
👍: 0 ⏩: 1
xD well, its still an awesome shot ^^
👍: 0 ⏩: 0
You are crazy ! How did you manage to render this ? How long did it take ?
Did you use a fractal program or a 3d one ?
👍: 0 ⏩: 1
lol yes i am crazy 
👍: 0 ⏩: 1
You have a quadcore and took you hours ? 
👍: 0 ⏩: 1
hardcore optimised 64bit build, 4x3ghz
yes, i try to make the most of my available resources
👍: 0 ⏩: 0
my renderer, allura. spectral, physically-based rendering, metropolis-hastings sampling and all that jazz.
👍: 0 ⏩: 0
actually my program's made a lot of progress since i rendered this
👍: 0 ⏩: 0
Wow. Thats incredible! It looks so real, like a plastic model almost.
👍: 0 ⏩: 1
i'm using a prety advanced rendering method now, based on this dude's work: [link]
the plastic look worked out kinda ok here (glad you like it!), but i want to investigate other - more interesting - reflectance models (esp. metals) after my exams
👍: 0 ⏩: 1
Sounds neat!
👍: 0 ⏩: 0
| Next =>