welcome!
here you can find all my projects. many would be in so called beta state and i will post new versions sooner or later. feel free to comment on anything.
deworm is a flexible node for creating wormlike meshes.
| idea: | create wormholes or ankles and knees for IK bodies |
-
version download | dd.mm.yy | ||||||
deworm v1.3.zip (52.45 Kb) | 16.05.07 | new features: | Straight Cut pin - fixes appearance of displaced segments | ||||
Deworm v1.2.zip (51.26 Kb) | 12.04.07 | new features: | TwistFix pin (get rid of twisting segments). hints about How Deworm Works added to Deworm module | fixes: | sections can be oriented in relation to each other - TwistFix. some performance improvements tnx to WestTricks collection | i've learned: | a lot about trigonometry in vvvv more... |
Deworm v1.1b.zip (46.16 Kb) | 22.03.07 | new features: | Bin Size input (idea from diki & elektromeier ) | bugs: | twisted segments bug - discussion | ||
deworm v1.0.zip (44.34 Kb) | 08.03.07 | new features: | create worm body from custom geometry 2D sections | fixes: | index table creation. texturing is all right | i've learned: | pulling out index tables from ready made meshes. mesh apperance with flat shaders is really ugly and you can fix that - Reindex Module |
deworm.zip (12.97 Kb) | 03.03.07 | bugs: | texturing is totally wrong | flops: | building the index table from scratch is really smart idea, but also - the very complicated one |
(Click here if you want to add some comments)
It would be very cool if you can draw more than one worm with just one module... maybe add a binsize inlet which defines how many segment long one worm is. and what should happen with the end of the worm? open, filled, rounded? - elektromeier
polygonising a scalar field
the challenge
polygonising a scalar field is an old idea from early ages of CG and solutions are easily found on the web. well known 'marching cubes' algorithm is really fast in code because it works almost entirely on lookup tables.
here is one good tutorial.
things are very different with vvvv. working with large spreads of data is slow and when I started to build the patch i didn't expect it to run in real time. then it looked it might become fast enough with some optimization. finally I landed somewhere in between (majortom to ground control...).
this is kind of pre-release. there is no help patch but two examples named TRY... one is with metaballs. i couldn't find better approach for metaballs yet so that one is very slow.
because large 3D fields are huge spreads of data i have included somehow CPU expensive 'smooth' algorithm so it acts like a tradeoff for smaller fields. take a look:
8x8x8 field | 16x16x16 field |
512 cubes | 4096 cubes |
285 vertices | 1047 vertices |
469 triangles | 1887 triangles |
normal mesh
|
|
130 CPU | 477 CPU |
. | |
smoothed mesh
|
|
180 CPU | 685 CPU |
. | |
normal mesh (marching cubes algorithm overdefines mesh)
|
|
. | |
smoothed mesh
|
|
smoothing algorithm helps somehow but produces some trembling during animation.
inspired by swarm theory (also check Learn More - Related Links)
balancing just a few simple rules for all members can lead to significant behaviour patterns of the group.
simple simulation:
Who you gonna call?
anonymous user login
~1d ago
~1d ago
~9d ago
~11d ago
~13d ago
~16d ago
~16d ago
~23d ago
~30d ago
~30d ago