Introduction: An Aesthetic Generator

Inspired by a series of Surrealist Tactics and Protocols, the Aesthetic Generator, allows for a series of reconfigured and completely unique realities to be generated and analysed in a game of play between the active user and generative machine.

This is part of an on going research project called 'An Architecture of Lumetric Causality' which explores the generation of aesthetic realities, through the deployment of coding, architectural model making, film, and analogue transnational drawings.

Things you'll need:

  • A Digital Camera or Film Footage
  • Rhino / Grasshopper
  • A Basic Understanding of 3D Animation

Step 1: Surrealist Tactics and Protocols

Picture of Surrealist Tactics and Protocols

Above: Yves Tanguy, Man Ray, Max Morise, André Breton, cadaver exquis, 1928. Pen and ink, graphite and coloured crayons on paper, 31.1 x 20 cm, The Art Institute of Chicago (reproduction, the Art Institute of Chicago).

In early Surrealism artists such as Man Ray, Max Morise, and André Breton (among many others), were known for their exploration of the unknown through elaborate games, tactics, protocols and automation. Through this play, reconfiguration and collisions of worlds a prominent design methodology was discovered, leading to the strange yet enticing imagery that we know as Surrealism. Even though surrealism has fallen out of favour in more recent years I believe theses techniques are as relevant today as they were nearly 100 years ago and it is technology that can now help as a critical resistor for new playful exploration.

If the Cadaver Exquis (The Exquisite Corpse) drawings of Morise and Breton is an example of;

'automatic drawings ... [which] deployed machine-like depersonalization and an insistence on the unpremeditated mark to disable the twinned authority of unique signature and compositional order' - Susan Laxton 1

then through creatively deploying similar ideologies in relation to more contemporary means of production we might unlock a new way of viewing our ever changing design practice - For this reason I will refer back to the game of Cadaver Exquis throughout the guide, as a reference point for our journey.

By the end of this guide I hope to have explained a few of my interests as well as enabled you to explore a similar creative generation process through the deployment of relatively simple and playful techniques.

Step 2: Playful Scripting

Picture of Playful Scripting

The first stage of the process requires us to operate in grasshopper - grasshopper is a free node based generative modelling language for Rhino3D - which can be downloaded here

This script is the first part of the process in creating an image such as the cadaver exquis. The basis of this scripting comes from a much larger project, which will form a series of Instructables over the next few months, but in the simplest explanation the RGB script analyses the colour information in an image file and from the Red Green and Blue colour channels re-plots this information into XYZ 3D space.

To start with you should drag and drop you're chosen image into the node called 'Image Sampler'. From here the main variable of the script is the 3 sliders located on the bottom half of the grasshopper file, these are the strength by which you'll be deforming you're image. As you change these values the mesh's will change dynamically - depending on your computer setup this may take some time to respond, if this is the case, try disabling the image sampler and re-enabling once you have selected your values.

In the example rhino screenshots the variation between strength gives way to vastly different outputs. Due to the node based nature of grasshopper these variable sliders can have their values adjusted beyond the provided range for different results and also changed completely to take different types of inputs - which will be something we explore in a later guide.

Step 3: Initial Outputs and Placement

Picture of Initial Outputs and Placement

Now that we've created a single mesh we should move onto creating a series of meshes and see how they interact between each other, in this sense we are creating the next step in our cadaver exquis, the unknown folds, where each image we sample and mesh we add to this collection will be another 'fold' in our exploration.

For this example I will place the meshes in a linear line using the Red Green Blue Images from the previous example. This process of placement can be automated in grasshopper, (see script below for one example) or done manually, this allows for the potential to place these meshes in different configurations to the linear line or even for further deformation once created.

As you can see in the example image, the meshes aren't currently intersecting, this is due to the fact that they're block primary colours and deform in a very standard way - to counter this, we'll have to look to a more complex set of still images in the next step - which will be an example project, The North Korean Experiment.

Step 4: Example: the North Korean Experiment

The North Korean Experiment is a recompiled film made from a documentary that was shot at a gallery opening in the North Korean Embassy in London. This film revealed an opportunity, due to its mundane exterior existence, to be re-imagined as a place of exploration through the RGB script.

By breaking down the film into sampled stills, a new narrative was revealed overwriting the current existence of the Embassy. Through the intersection of these images new configurations are understood, characters who once appeared in the film now place themselves into different configurations across time and space.

The video also shows a simplified diagram of the linear placement of images through to deformation as well as showing an example of 'replaying' the film. This will become the next stage of our process, the unfolding of our combined cadaver exquis.

Step 5: Re-Animation

Now that you have you're complex set of meshes, arranged in whichever order you have chosen to pursue, its time to re-animate you're new creation, or to refer back to the cadaver exquis once again, this is where we unfold our drawing to reveal what we have made. I've chosen to play this game with a series of recorded images of an architectural model I made to show a different range of film types that can be produced by this technique.

For the animation we can use a variety of methods inside rhino or in other pieces of animation software depending on personal preference. In both scenarios, we'll be animating a clipping plane that moves through the meshes. (See above video diagram) In Rhino the two options I would recommend would be Horster Tools, located here, for Grasshopper or a stand alone plugin called bongo, located here. Both options will give you a standard set of animation tools for the Rhino interface. If learning animation software is not something you wish to do I do have a simple script that re-animates the linear line that I can upload, but the hope is to create something more complex then this.

To animate the clipping plane in rhino create and place the clipping plane at the start of you're mesh, then select the auto-keyframe option in bongo, and move the clipping plane through the mesh to the other side. It's as simple as that, you've now dissected you're cadaver exquis. The next step is just to render out the images and see the results of you're completely unique creation. You can even group the camera to the clipping plane once setup in the correct orientation to create a mechanical replay of time against space, similar to the North Korean Embassy Film.

When coming to render the film, I've found that the fastest way to produce a series animations is to actually use rhino's internal renderer, which is mostly a personal preference for speed of play, so feel free to use whatever rendering software you have access too or are most familiar with.

Step 6: Examples of Play

Now that we have a way of unlocking a creative language, we can start to play games freely with the system. You now have all the components in place to repeat, alter and adapt this game to suit your own needs.

This step is less of a instruction then and more of a creative suggestion for types of play, showing examples of games that I've played - these range from preformative scale models to full size performances - each with a similar process or series of steps to abstraction. These then range in output, from architectural model to drawing or even sculptural pieces.

Step 7: Future Developments

Picture of Future Developments

As a teaser for the next guide I've included some images of one of a series of replacement analogue device that I will go onto show, this device being known as the RGB-3.

This device attaches to the camera and senses the environment using LDR (Light dependant resistors), these then go onto replace the variable controls, adding more levels of complexity to the game that can be played.


Fuzzy-Wobble (author)2017-08-22

insanity. love it.

Swansong (author)2017-08-22

That's a really interesting intersection between surrealist art and technology. The images it creates are compelling and beautiful :)

About This Instructable



Bio: Thomas Parker is an architect, designer and filmmaker who recently graduated from the Bartlett School of Architecture, his interests lie in experimental architecture through exploration ... More »
More by tparc:An Aesthetic Generator
Add instructable to: