The secret of Martens's transformations--any hints?

Maurice Meilleur's picture

No doubt like many, many others, I've been captivated by Karel Martens's symbol-generated transformations, having recently gotten ahold of a copy of Printed matter. And probably again like many, many others, I'm wondering how he made them.

By every account, Martens is a generous teacher and colleague, so I'm certain that he had a pedagogical purpose in mind when he answered evasively an interviewer's question about how he makes these:

Martens: A birth announcement for my grandson (2002, Zeno) seems to be a moment in my live that I start to work with icons. The desire had been there for some time. Every designer or student seems to rediscover the crude halftone at some point, but I’ve always thought how nice it would be to create different shapes instead of dots. My father’s typewriter always fascinated me. I’d hit a button and the letter ‘a’ appears. But that ‘a’ might as well be a bird or another picture, something I’ve developed with Roelof Mulder for Emigré magazine ... Ten years later I had the opportunity to figure out how to reconstruct Zeno’s portrait in a different way.

Harmen Liemburg: Many people would like to know the secret ...

Martens: Well, it’s relatively simple, and I’m sure others have discovered how to do it as well, but I’m not going to tell you anyway! This assignment (2001, façade Cultureel Centrum) is about the perception of separate colours at close range and the mixing colours that one perceives at a distance. The fact that you can create a third colour out of two, is something that never ceases to excite me. It’s nothing less than a miracle! The problem here was to translate this idea (that was developed in Photoshop) into prints on large glass panels. Screenprinting proved to be too expensive, but after a while I found a company that could print on transparent material. As they needed vector outlines, this caused a necessity to come up with a workable system to make that technical translation, which we finally did.

I've been thinking of doing images similar to these with glyphs from typefaces, and while I'm all about learning things the hard way, I could use a hint about what he's up to here. Martens supplies a few of these himself here, of course--not Photoshop, and 'relatively simple'. And I know that at least one version of these transformations--creating greyscale instead of color--doesn't take very long to do. In 2004, there was an exhibit of Martens's and Paul Elliman's work at Seoul University; Hans Gremmen and Min Choi worked out a way for visitors to the show to have an instant symbol-generated portrait taken of themselves:

I've thought about making these with vectorized halftones made from color-separation layers, or from ranges of different levels of greys in an image. But note that Martens seems to be using specific vectors in combination to create specific colors--he's not pulling them in at random, or it doesn't look that way. (The same goes for the Gremmen/Choi images.) And since I know he's not doing these by hand, I'm stumped for how he gets these in a way that doesn't take months to do.

So, if there's anyone out there who knows Martens's secret: I don't want to disrespect Martens's right to keep his methods to himself or undermine his goals in doing so here, but can you at least give me a start--software, approach, anything?

aluminum's picture

What specific secret method are you looking for? In terms of the icons, they are elements repeated radially then combined with various masking/combining/cropping. Doesn't seem too top-secret.

As for taking them all and combining them into an image, there is software out there that does this. Google 'photo mosaic software' to find some options.

blank's picture

He probably built a huge library of stock shapes and then wrote an Indesign plugin that analyzes a photo, his shape library, and then combines them all into a color image. I had a design teacher who did this to create art out of concentric B&W circles.

Rob O. Font's picture

"Incubator experiment 1" on youtube.com, shows that a few simple overlapping bilatrally symetrical shapes, circularly oriented, when under the influence of interpolation, and animated, as in this example, can lead to 1,000s of still shapes of interesting design. I personally, find the still variety to be a bit boring and repetitive, but I can see how massing them into arrays and then printing them, could be interesting, to some.

hrant's picture

While my cup of tea is more like this:
http://www.underconsideration.com/speakup/archives/002777.html

hhp

Syndicate content Syndicate content