object ...

ive just put some images up for the blog ... theyre kind of oput of sequence though. how about this for the object?

quartz vs tungsten

originally the group agreed to scan some quartz, but after getting a timing crystal from maplin and hacking the casing off we decided to go with the bulb filament after all ... especially since they had 'one they prepared earlier'! heres the images i'd found before that had inspired me to nag for scanning some tungsten... sadly we didnt manage to zoom/focus in so far .... ive found some electron microscope images of quartz too, but i think we would have had the same trouble getting a good shot.

images borrowed from:

http://www.ralphmag.org/microspaceN.html
http://www.istockphoto.com/imageindex/391/9/391924/Tungsten_Light_Filament_2720_x_Magnification.html http://www.istockphoto.com/imageindex/391/9/391928/Tungsten_Filament_624_x_Magnification.html

microscope images and master plan!

these are our microscope images that ive prepared for processing to create the volumetric object. plan was to make each image a different orientation, then compute the density for each voxel using the pixel data of each image... it isnt as complicated as it sounds tho. If each image represents the x, y or z plane then take the sum of pixel(x,y), pixel(y,z) and pixel(z,x) if the normalised average is greater than a certain threshold (0.35 seems to work well) then plot a point (cube).

space station sized object?

another experiement- this ones probably not buildable, or would 'decay' throughout the show ... probably as we are mounting it ... but the atom array option makes sure all the points are connected, so it is technically possible. I just liked how it looked because it made the organic looking shape mechanical again. Excuse the volumetrics ;p

bit sketchy

i sketched out a few ideas of how i wanted it to look .... ;p ... honestly ... sketch render option to be blamed for this one- seemed a bit dull to just show the wireframes for the different views.

nurbs of the extruded object ... ive been using a fresnel shader on most of the renders so far to show the surface and to try to make it look like the microscope images. Un-coloured version should be more accurate, the demo objects looked creamy yellow ... may try another render later.

heres the actual matric extrude object, its fairly chaotic as the platonic i extruded from didnt seem to be symetrical.

this is a slightly tweaked version, i prefer it because it reminds me of the microscopic bits of mank that were on the microscope objects. Cant build this one though because theres lots of objects that arent attached ...

bit of colour

i made this using a matrix extrude on a shape, then made it generate a nurb version and metaball version. The metaball is the green shape, i think it looks kind of organic especially the tentacle/ spines which are because i had to turn down the metaball distance/influence to get some sort of shape in the rest of the object. You can just about see the purple object between the 'bones', this is the nurb shape... we would have to have the 2 objects to get it to support the floating segments.

object?

how about this for an object? its not exactly what i was thinking of... ive been working on something to create a volumetric object from the 3 'images' we got out of the electron microscope, but ive been having some trouble exporting it back- cinema 4d has an xml file format, but its over 1000 lines to set up the scene- before you put any objects in it! Ive been exporting a vrml file, but its also having trouble doing this as the object ive made consists of roughly 125,000,000 cubes at full resolution :( i can generate it at a lower res but its not so impressive then ;p been working on an alternate plan- using mograph and xfrog cinema 4d can generate fractal like iterated objects. I havent yet managed to get it to generate anything that look like it is buildable tho so ive been playing with c4d's other features.