Print magazine August 2008 cover design
| Late in April the super friendly people from Print Magazine commissioned me to create a cover design which would conceptually support the main feature article about “kinetic typography” of their special summer issue. And here kinetic isn't even so much about motion and animated typography as about the process nature driving the creation of letter forms. It's an interesting (apart from the marigolds and fresias, heh!) article written by Emily King who is currently also curating A recent history of Writing and Drawing at London's ICA. The article too features Hudson-Powell's Responsive Type project, Juerg Lehni (of Hektor fame), Peter Cho and the always inspiring Jonathan Puckey. So our early conversations about this article and the proposed phrase “type, form and function” became the starting point for my further explorations. | From the beginning, the cover design needed to reflect the following for me: the creation process made as visible as possible (that also meant a probably quite raw end result), it needed to be sculptural, use generative form giving and involve some form digital fabrication to indicate the rapidly changing scope & meaning of the word “print”… I've looked into various possible means & approaches suitable for generating not only typographic shapes, but then later allow us to turn them into (solid) forms too. To be able to do just that, I needed to concentrate on particular, sufficiently generic tools/processes which could be combined and customized for this particular project. This blog post is a fairly detailed memory of my journey without map into this terrain of mashing up typography with MRI scans and digital fabrication / 3D printing… |
Gray-Scott model
| After some initial research I've chosen a process called Gray-Scott reaction diffusion to become the heart of the idea. This model is used in chemistry & biology and is related to the field of Cellular Automatation, although is not rule based and can't be used for computation. As the name partly implies, a reaction diffusion exhibits certain behaviours (reaction) and also has a space filling property (diffusion). Both of these combined should make some exciting conceptual source material for experimentation - and so it did! :) The Gray-Scott model has only 2 parameters in its most basic form, but can be used to simulate and generate a wide range of visual patterns by evaluating simple equations for each cell within a 2D grid. | The patterns generated range from a blurry soup-like character to alive dot patterns to maze like structures. The range of expression and temporal stability for each pattern is also quite varied and lies somewhere between liquid and absolutely solid. Here's a video sketch of the process combined with a typographic mask to enable different rules for parts of the image. Letters will adapt an almost static maze like pattern with a actively changing neighbourhood around them. The following configuration map by Joakim Linde not only shows the possible range of expressions quite clearly, but also that there's only a small strip of combinations which produce meaningful results at all - a typical property of a generative system. |
Image by Joakim Linde
| The segregation of foreground from background is one of the basic Gestalt theory principles. So only choosing a single pair of parameters for the GS simulation would only have produced one type of homogeneous patterning (i.e. the background), without a suitable contrasting foreground (e.g. the lettering). | I needed some kind of a “mask” to introduce an alternative configuration to the simulation for areas supposed to become foreground. Because I didn't want to use any existing typefaces for this project I wrote a small sketch (source at the end) to generate the core paths and mask for the “type & form” phrase and later could fine tune the kerning & perfect line weight for the sculpture. |
Once applied the simulation then took its course and generated this maze of shapes:
The level of fairly complex emergent behaviours present in this process is to me one of the most mind blowing things, like the ever self-dividing dots (reminiscent of cell division) who are steadily filling up the space and “try” to align themselves in fairly equal distance, wipe each other out, split again etc…
From shape to form
| The fundamental issue with the approach so far was, that even though I discovered an interesting process, everything still essentially happened in the flatland of 2D typography and didn't really address the “form” aspect of the phrase very well (at all!). However, having played a bit with visualizing MRI scan data before, I figured I could use the combined 2D output slices/frame of the simulation to form a 3D volume, in which each pixel becomes a voxel and grayscale levels are to be interpreted as density. | This new 3D volume dataset can then be fed into the same algorithm used to visualize MRI scans (amongst many other things) and create a super complex 3D mesh. Initially this started out at 5.23M triangles, but then due to various re-designs and changed parameters could be reduced to 1.9M triangles for the final model. Below are some explorations of the customized Gray-Scott simulation combined with the Marching cubes algorithm used to slowly create a volumetric form from the 2D frames. (Btw. I've used the same algorithm for my contribution to the Advanced Beauty project. More about this project soon.) |
Building up the volume, slice by slice.
View from behind. Time becomes depth.
View from below. See how the dots are splitting in two repeatedly...
Frontal view after several thousand iterations.
Reversing the process...
Now using the mask image as seed for the simulation to improve legibility…
Frontal view after several thousand iterations.
View from behind, spikey!
Wireframe (view in highres)
Wireframe (view in highres)
Form to object
| Once I'd decided on this route for the cover, using 3D printing to create a physical object from this process seemed like the logical conclusion for this whole exercise. Thankfully, the magazine's team agreed and after running the idea past my friends @ ThingLab - scaring them slightly with this monster of a model (in terms of detail) - they went ahead anyway and put all their skills into it to produce it. | All in all, it took 2 attempts to print it (using a Z450 printer) and infiltrate it successfully - and make me super happy about my first foray into this realm. The first attempt failed due to the extreme level of detail and amount of filigree structures within the object. |
1st failed attempt.
Final object
The off white colour, texture and rough finish give the printed sculpture a somewhat ancient, almost archeological feel & remind me a lot of the chalk cliffs at Birling Gap, my favourite place on the south coast of England. That and the organic look caused by the Gray-Scott process are also reminiscend of David Cronenberg's vision of biotechnology in eXistenZ, for example.
Below are some more detail shots of the sculpture as well as an alternative version of the cover design on black.
Behold, sucess!
YAMO (Yet another making-of)
If you're still willing, you can read an interview with a slightly different focus/scope about this and my other work on the Print magazine website. There're too much more images in the related set on flickr. Enjoy…
About
This blog is an attempt to document our work in progress as much as possible and share the knowledge and all the things encountered on the way.
Archives
Feeds
Contact us
Email:
info@postspectacular.com
Join us
Network
License
Discussion