blacklog

We just created channels in Vimeo for oGFx and E15:

• http://www.vimeo.com/ogfx
• http://www.vimeo.com/e15

It helps a lot to understand what these things are about when you look at them in motion.

For my thesis I modified e15 and created a studio web application to log and share my creative process while writing ogfx scripts. To save time, I embedded the studio application within PictureXS. I separated the studio from PictureXS by making a studio controller and adding some functionality to the picture model, like the ability to publish code and snapshots from e15 together at the same time. People visiting the studio website could send messages to the custom e15 I was running, and I could respond to them without leaving the programming environment in e15. It is not very hard to make an application take a capture from the pixels in one of it’s views and post it to a web service, so the interesting stuff to notice is independent from the platforms used, and what really matters is to observe how the creative process changes when it is performed in a digitally mediated public space.

Places like the MIT Media Lab tend to push towards figuring out new ways to make technology mediate between humans and their needs. There are many cases where this mediation might lead to an improvement of human life, but in many others the result is simply alienating. Writing instructions that make pictures instead of making pictures with my own hands is an interesting separation. Sharing the way these instructions change as I search for a different picture might illuminate about some aspects of computational art, but It could also be just another way to produce data where patterns could be found, just as it seems everybody everywhere is doing these days. We live, after all, in a statistical world.

The studio application is called MyStudio. I’ve turned PictureXS and MyStudio static while I find a place to host them outside of the Media Lab, so nothing can be posted in them for now. I will turn the dynamic features on again when I figure out how to pay for the 30+ gigabytes of disk space I need to store all the pictures in PictureXS if I host it on my own.

This image shows the first 110 pictures I published in MyStudio:

Kate has been recently working on an implementation to support complex 3D mesh manipulation in E15, and she asked me if I could recover the experiments with OpenGL lights I was developing a while ago, to get a working simple lighting model for her to play with.

The following image shows the effect of a spotlight hitting on one of Kate’s meshes, where I placed a scanned image of original artwork from the classic Vampirella comic from the 1970s published by Warren that I found somewhere in the web.

It might take a while to properly implement full control of the OpenGL light resources from the E15 python interpreter, but it will be a very nice thing to have, specially thinking about the potential of combining lighting information with Kyle’s new in-progress implementation of GLSL support for E15. It makes me wanna grow hair on the MIT website.

We have always thought that oGFx should feature lights, because they open the way to a lot of resources to play with when using textures and shaders. Bump maps, highlights and all kinds of other effects depend on the ability to calculate how light bounces over a particular vertex or fragment, and light in motion is enough to enhance the experience of a digital environment, just because a change in lighting can be understood as a transition between spaces, the passage of time, or both. Control over lighting models is an important feature to consider when thinking about the creation of a language for interactive graphics.

I recently finished the first iteration of glLight support for oGFx , and already experienced a lot of pleasure playing with it in some of the scripts we have made in the past. Following the advice from J. Popovic, teacher of the MIT Computer Graphics undergraduate class, I wrote some code to turn on and off graphical representations of some of the objects in the oGFx context, making it easier to debug otherwise hard to track things like light directions, light positions and surface normals.

Here is a very rough (and I mean it) video where you can see the influence of the light in motion over a scene where I was manipulating an interactive animation I made using quadratic Bezier curves from Apple’s Quartz2D, and here is another one, same degree of roughness, and a different use of the same curves.

I wrote a few simple methods in PictureXS to let E15 request individual images from it, and used them to build the qbert staircase again, using the average color of each picture to wrap the rest of each cube around it. When the first set of geometry was loaded, I was surprised to see all the censored pictures I forgot to block from access with the simple methods I wrote, and then decided it would be more interesting to show a special censorship label for each censored picture instead of disappearing them from view. After loading PictureXS again, I found out that there are not many censored pictures, but certainly more than what I thought. Naughty people…

you gotta love those normals!