Alex Dragulescu: Extrusions in C major (detail) / Blogbot (detail)

[Read pt.1 for completion] Dragulescu’s Extrusions in C major uses music as its input, specifically the “Trio C-Major for Piano, Violin, and Cello” by Mozart. Here the artist rigorously describes his mapping: Different colors represent different instruments, while each segment of the fragmented forms represent a single note, with characteristics such as velocity and duration controlling the development of the form. The final form represents the temporal structure of the piece.

Blogbot and related projects Havoc and Algorithms of the Absurd represent a slightly different approach with a performative flair. Blogbot generates “experimental graphic novels” from content found on blogs. Texts are presented as though being read, appearing line by line accompanied by visual icons.

The online example What I Did Last Summer appropriates pixellated images of war machines and soldiers taken from computer games. They are then used to illuminate a narrative of fragments from two blogs relating to the Iraq war. One is by an American soldier and contains details of raids and military maneuvers, the other is the famous blog of Salaam Pax, the Baghdad Blogger. The introduction of temporal and graphic aspects to the text turns it into a performed narrative. Simultaneously, a graphic composition of increasing complexity is created as the text grows on the canvas.

Lev Manovich speaks of data visualization as the New Abstraction (see Data Visualisation as New Abstraction and Anti-Sublime, Word DOC file). In this context Dragulescu certainly presents an interesting take on info-aesthetics, with complex data sets being appreciated for their structural beauty alone.

Alex Dragulescu is from Romania and currently leads the Experimental Game Lab at the Center for Research in Computing and the Arts at University of California, San Diego.

Alex Dragulescu: Spam Architecture (detail) / Spam Plants (detail)

Romanian artist Alex Dragulescu turns data sets into raw materials for the generation of tantalizing 2D and 3D forms. Rather than scientific visualization intent on clarifying the content of the data, Dragulescu creates graphic and temporal compositions notable for their strong graphic qualities.

Spam Architecture is one project that has garnered much attention recently. Here spam is translated into three-dimensional form by analysing keywords and patterns in the text. Like its sibling project Spam Plants, it explores the mapping of textual data into spatial configurations.

All trace of the original data source is absent in the final result. No reference to the textual material remains, nor of the analytical process involved. Instead, a single coherent form is presented, with no signifiers indicating its origin. In this sense, the spam data could be said to simply constitute an arbitrary pseudo-random data input, with the result bearing no semantic connection to the raw material that it was generated from.

Dragulescu does not provide clues or any rational way of evaluating the nature of the mapping. But nor does he make a claim to producing literal meaning. Hence the viewer is free to enjoy the results as a complex formal experiment in which spam undergoes a process of transsubstantiation, transformed from a source of irritation into intriguing objects of great beauty.

Generator.x has been on extended (and unannounced) one-month holiday, but now summer is ending and blogging will slowly resume. To warm up, here are a few random links that have accumulated:

• AOL reSearch has just managed to release a substantial data set containing 20 million search queries from over 650 000 users. What was probably an eager attempt at scoring Open Source brownie points, has rapidly become a public relations disaster. Americans are rightly paranoid about their privacy, and the data set is likely to include personal data like names, social security numbers, unpleasant searches for porn and violent images etc.

  The original post has not surprisingly been removed from AOL, although a cached copy can be seen using Google's cache. Mirrors posting the data set can easily be found, one of the best bets is to try the Bittorrent download. While the release of these data is bad news, it’s sure to be of interest to information visualizers and dataminers. It would almost be surprising if no art works came out of this debacle.

  Read Techcrunch for a good overview of the whole story.

• Kunstverein Medienturm in Graz will feature a show called Further Processing in September. The show will show software-based works created with Processing, and also give a presentation of the tool itself. Contributing artists are Pablo Miranda Carranza, Fabio Franchino, Ben Fry, Golan Levin, Lia, Mark Napier, C.E.B. Reas, Karsten Schmidt, Martin Wattenberg, Marius Watz. The show is curated by Sandro Droschl (director of Medienturm and one of the curators behind Abstraction Now) and myself.
• Art.ficial Emotion 3.0 is an interesting exhibition at Itau Cultural in Sao Paulo, Brazil, featuring a major presentation of media artists whose works relate to cybernetic theory. See Paul Prudence's writeup on Dataisnature for a summary. Regine over at we-make-money-not-art recently did a interview with Guilherme Kujawski, one of the curators of the show. In it he presents his ideas about the exhibition and its relation to cybernetic theory.

  For visual impressions of the show, see the following Flickr sets : mrprudence, watz.

• Code & Form is a new blog I’ve started to cover more technical and code-related issues that would be too geeky for Generator.x. This separation of content means that Generator.x will be more clearly focused on finished works and theory, rather than tools and technology.
• Ars Electronica is around the corner, if anyone is going and would like to meet up please send me an email on marius at unlekker net. I had thought of organizing an official Generator.x gathering, but there are not enough hours in the day… Hope to see some of you there anyway!

Built with VVVV: Football Globe Germany / Superformula

Following up on the post about David Dessen, here is some more VVVV news:

The busy boys at Meso have launched a new version of the VVVV wiki, improving considerably on the previous design. Important new functions include the Galeria (where Meso can show off VVVV projects like the Football Globe Germany) and News blog sections, as well as improved Reference and Documentation sections. The main access for users is now the Fan club page, which gathers access to the forums, blogs and shoutbox in one place.

All in all, this redesign sharply reduces the geek factor which marked the previous VVVV site. It should also make it easier for would-be users to find information about the tool. Like Processing, VVVV might be in eternal beta, but that doesn’t mean there is no maturing of the tool and its community.

David Dessens (aka Sanch): Linear / Math surface destrukt

David Dessens’ work with VVVV has been generating a lot of interest since the first appearance of his shell-like objects on the VVVV pages. With the launch of his own blog Sanch TV he displays a range of hugely impressive formal experiments, bursting with voluptuous curves and saturated color. It is proof not only of Dessens’ personal talent, but also of VVVV’s qualities as a production tool.

Most of his experiments involve the use of vertex shaders, filters that affect geometry but which are executed directly by the graphics card (GPU) rather than the computer’s internal processor (CPU). The GPU is a specialized chip dedicated purely to graphics operations, and farming out computation to it results in lightning-fast execution. Some of Dessens’ experiments are based on shader implementations of mathematical "supershape" surfaces. These meshes are then distorted and manipulated further. But even working with standard mathematical formulas as raw material, Dessens manages to produce images with a unique visual style.

At the moment Dessens’ interests lie mostly in live visuals, but it will be interesting to see how his work develops. He is currently artist-in-residence at VVVV developers Meso, which should be a guarantee of more interesting work from him in the near future.

Be sure to see Dessens’ showreel, generated purely in realtime. As an extra bonus for wannabe VVVV hackers, he also posts shaders and patches on his blog.

Fornes / Nowak / Corcilius: From DIN to DIM

For a different take and a different scripting language, go read theverymany, Marc Fornes’ blog on his experiments in computational architecture. 98% of his blog so far is Rhinoscript code for creating generative structures, accompanied by intriguing illustrations. It makes you want to work with Rhino just to be able to see it run.

For those who don’t know it, Rhinoscript is a VBScript language used to control Rhino, a high-end 3D package used for anything from CAD/CAM and visualization to computer animation. Rhino is popular with coding architects, sculptors and CGI heads alike. It’s not as old skool as AutoCAD and AutoLISP, which has been used for computational architecture since 1986. But it’s likely a lot more useful.

theverymany is refreshingly focused on sketches and code, but there is documentation of one interesting recent project: "From DIN to DIM", a “series of experimentations looking at transitions between the German Standard of design to self-similar objects controled by declared variables…”. Done with Vincent Nowak and Claudia Corcilius, it consists of generative formal studies, using nested loops to generate structure.

As with much computational architecture, the results are visually very compelling. The techno-organic tower structures recall fashions in blobby architecture, while simultaneously reminding one of 70s sci-fi book covers. The translation of simple code structures into complex and appealing form seems effortless, it would certainly be interesting to see the slides shown in higher detail.

Marc Fornes is a graduate of the AA's Digital Research Labotary class, and is currently working as an architect for Zaha Hadid Ltd. He indicates in the sidebar of his blog that his rhinoscript library might be available as open source.

Harris / Kamvar: We feel fine

We feel fine is a lovely new project by Jonathan Harris and Sepandar Kamvar. It scrapes blogs, Myspace accounts and similar social networking systems, looking for the sentence fragments “I feel” and “I am feeling”, recording the sentence and the context they appear in, including photographs in the case of Flickr entries. The result is a massive dataset of feelings and moods combined with demographic data.

Written in Processing, We feel fine is a delightful combination of data mining and typographic treatment. Like Golan Levin’s The Dumpster, it is at once poetic and somehow serious. The playful use of color and typography supports the content of the piece, making it both beautiful and wondrous to explore.

While We feel fine goes a little further than The Dumpster in trying to project scientific axis on the data, both projects make a claim at scientific impact which is not really held up by the work. Ultimately, both are arbitrary visualizations of data which is hardly quantifiable. The success of these works (and they are successful) then stems from their ability to project a snapshot of human emotions in multitudes of permutation, as evidenced on blogs and social networking services. While the viewer may get a slight insight from the axis on which the data can be projected, the sheer size of the dataset is much more signficant in its impact on the audience.

Jonathan Harris presented at the OFFF Festival yesterday. See his site for his other visualization projects. Sepandar Kamvar founded a search engine called Kaltix, which was acquired by Google. He is a consulting assistant professor of Computational Mathematics and Engineering at Standford.

Jill Walker’s excellent jill/txt blog about electronic literature, social networking and, well, blogging, has won an award from the Meltzer Foundation at the University of Bergen. She will receive NOK 100 000 (approx EUR 12 000) in recognition of her “excellence in research dissemination”. This is exciting, as it reflects an official recognition of the potential of blogs to create and disseminate knowledge. Read Jill's blog post for more info, and her publications page for more of her work.

Congratulations, Jill, keep up the frontier work!

Apropos: If you are hip blogging scientish, you should have a look at the Hard Bloggin' Scientist manifesto. It provides operating guidelines for a new movement. While a bit tongue-in-cheek, it’s got some valid points.

They even have sticker images (in pink!) that you can put on your blog to represent the blogging massive. What better way to feel like a part of something important…

Despite our enthusiasm for folksonomy and tag-based navigation, the Generator.x blog has not utilized tagging to manage its own content. While all entries were tagged, those tags could not be used to navigate the blog itself. Until today. Help has arrived in the form of Christine Davis' nifty WordPress plugin Ultimate Tag Warrior 3.0. As the name would indicate, this plugin presents a complete solutions for all things tag-related.

As a result, there are a few new features, although they can be hard to spot right away. Where previously the tag links under the post titles would lead to a search on Technorati, they will now present you with an archive of all entries tagged with that keyword. You can also access the tag search directly with URLs of the form http://generatorx.no/tag/thetag, or choose Browse Generator.x by tags from the Categories menu.

There is also a small collection of the most popular tags in the sidebar. It doesn’t look so great right now, but it provides the basic functionality. The Generator.x blog templates are due for an overhaul in the not too distant future, the tag list will be more elegantly resolved then. For now, the new tag features add to the ease-of-use of navigation.

A behind-the-scenes change is the migration from tracking visitors using Statcounter.com, to installing the excellent Mint stats package. At $30, it’s hardly much of an investment, and the design and presentation of information is spot-on. In fact, one could argue that Mint features some of the best table design on the, for proof, see the demo over on Designologue.com. Add to that the fact that like WordPress, Mint is specifically built to be extended upon and there are already a host of great plugins, it was an easy choice.

LEGO: Mindstorms NXT product range

Following up on recent LEGO posts, Simen at Bengler pointed this out:

LEGO just released information about their Mindstorms NXT product line. To launch in August this year, it introduces USB and Bluetooth connectivity, a smarter Mindstorms NXT Brick and new servos and sensors to play with. There’s an ultrasound sensor for detecting barriers, a sound sensor for responding to sound commands (it even has tone recognition), as well as improved touch and light sensors. This makes a good starting point for simple physical computing.

All in all Mindstorms NXT looks like an exciting package. If the 3D rendered press images are anything to go with, this is definitely a futuristically styled update to a LEGO range popular with the over-20 demographic. There’s a "Call for NXT Great Developers" if you’re feeling lucky.

LEGO also announced that it’s sponsoring a new blog: nxtbot.com. Written by LEGO veteran Jeff James, nxtbot promises to “provide coverage of the entire spectrum of consumer and personal robotics”. Note that while nxtbot handily references the NXT name, it’s not wrapped in LEGO graphics and explicitly states that it won’t just cover LEGO products. That could mean one of two things:

1. It’s a genuine independent blog.
2. It’s a LEGO blog planted to leak viral marketing.

You be the judge, we’re just paranoid.

Wired News has a feature on the development of Mindstorms NXT: Geeks in Toyland. Favorite soundbite: “”Imagine Flickr for robotics.” That’s from Mindstorms Director Søren Lund.