Building organic forms

In order to create the organic forms of my virtual organisms I am following strict rules to conform with my concept of computer generated mathematical forms. I can't manually shape them like a 3D modeller would do point by point and polygon by polygon, as they should all be based on expressed mathematical equations. In the examples on the left I went about this by starting with a simple perfectly mathematical platonic solid with each of its side equally subdivided and extruded using sound. The textures are based on 2d maps of fractals that are resolution-free as they are pure mathetical data.

In this first video you can see one of these first organisms reacting to a Mozart's Requiem. The reason I have used classical music in these experiments is because it is more organic and more random than beat-based electronic music. The result is quite an organic feel which is achieved by the innate geometry of this object.

Darkfield Illumination

In these experiments I worked on reproducing a microscopy technique known as darkfield illumination whereby a single beam of light is shot at the subject and its reflection is captured by the microscope revealing details that would not otherwise be visible. This gives the illusion that the subject is glowing itself instead of being illuminated by an external light source. This illusion is also the solution when trying to recreate this in 3d: turn off the lights and apply a glowing texture on the object. Adding a lot of noise gives the low-res effect of a microscope.

Besides the treatment I worked on further refining the technique I was using for generating my abstract microscopic organisms and their reaction to sound. As you can see from this video, the movement is much more fluid and indicative of the sound input. This was done by assigning the sound only to position of points but also to the scale of polygons so as the sound escalates the tentacles appear to inflate and vibrate until the music really kicks in and the whole thing inflates like it was injected by some sort of chemical. I like the idea how digital input can replace chemical input in this simulation of organicity.
Below are models of simple algae cells covered in more complicated textures as I start to get the hang of organic texturing.

Sub-polygon texturing

Successive trial and error to achieve a porous membrane in the likes of diatoms and their skeleton structure. The underlying mathematical geometry of these organisms is becoming very apparent when “dissecting” them and restructuring them in the 3D visualising program. The way to go about it is sub-polygon displacement where you lay out a two dimensional grayscale pattern. The darker areas denote depth while the lighter areas height, and in this way you sculpt a contour of the skin that is then going to be calculated and projected in 3d by the program. The patterns can be very complicated and elaborate with the limitation that the resulting skin cannot overlap itself. To create these skins I am using shaders which are resolution-free fractals – in effect mathematical equations – so as to avoid pixel artifacts and to create a mathematical organicity. The process in this way remains completely computer generated.

In these examples I am finally becoming more proficient at complex sub-polygon texturing but also at a more realistic simulation of darkfield illumination. With some post-production these could look quite realistic. In the examples on the left I used the input of sound to split the 5 polygons of the 5 sides of the star above and altered their position according to frequency. In the example on the bottom left I am altering the position of points on the surface of the diatom with the use of sound to make it seem as if it is bubbling inside. In this video you can see a 4 sided diatom whose sides have been mapped to sound and made to move and rotate bringing the diatom into digital life. In this second example, using the same principle but with different parameters to show the range of motion.

In these renderings I am finally starting to master the art of texturing - the textures are now highly detailed and organic and have a glassy tone to them just like a diatom’s glassy shell. I found that the trick is to get a balance between geometry and chaos – the way to do this is to create something fully geometric and then apply a layer of random distortion to it. This again reflects what I had noticed before about nature being mathematical with a slight distortion of chaos.

I used these high quality renders to produce this video, where I wanted to emulate how a microscope works. In videos I have seen of diatoms under a microscope, different layers of the same diatom can be seen depending on the focal point of the microscope as if they were under x-ray. This happens partly because the diatoms are actually made up of transparent glass shells but also because at such miniscule scales we reach a state of almost 2-dimensions where the normal rules of 3-dimensionality start to fall apart. I wanted to highlight this principle in my animation but also to emphasise the “inadequacy” of microscopes that only allow for a 2 dimensional, low-res, blurred, far removed view of the microcosm. This would then come in contrast with the capabilities of the digital tool that allows us to visualise, simulate and even immerse ourselves in worlds that we would never be able to experience otherwise.
I did not succeed in this video to portray the short-comings of microscopy as it fails to refer to the workings of a microscope, still referring to the digital medium that made it. I’m starting to wonder whether the digital tool can refer only back to itself.

Cytoplasm

This was a completely different sound experiment to the rest as I tried to find more inconspicuous and subtle ways to input sound. The first idea was to animate the actual pores upon the surface of the diatom. The proper way to do this would be to create an array of cylinders that scale up and down according to sound and then subtract them from the surface of the diatom. Unfortunately the number of calculations that have to be performed to achieve this is too much for any computer. The work around I found was to render the cylinders as a black and white map and then project this upon the surface. In principle it should work but the pixelation that occurs from this does not yield satisfactory results.

In this video, I successfully animated the cytoplasm of a diatom that in real life is secreted from tiny pores in the external environment looking like tiny hair poking out. I mapped the entire scale of each hair to sound frequency so the louder the sound the longer they are, and in the absence of sound they retract back into the shell. I also mapped individual points upon each hair to sound so that they reflect the input of sound through vibration and oscillation. This level of sound input adds yet another layer to my pieces, where literally everything from the smallest piece of hair to a huge swarm of diatoms are affected by sound. The only thing that I cannot achieve in any way (without sacrificing resolution) is to animate the textures themselves to follow sound input.