# INTRO

The journey of my latest generative project has been a winding one. One with a lot of interesting creative challenges, personal struggles and problems. I think they are worth highlighting and talking about as there are sparse resources recording and analysing the actual creative processes projects like these go through.

It is a holistic breakdown. Not just explaining the algorithm but also the creative journey. I hope you find this useful, interesting or inspiring.

***

# 1. Discovering the Idea

A massive amount of work goes into long form generative projects. They are technically demanding and dry; staring at 100 bricks of code to make a line curve ever-just-so or batch rendering 100 permutations of your algorithm to purge that one undesirable feature you saw that one time. All the while trying to maintain your creative vision and adapting to the novelty the computer throws at you.&#x20;

Because of this I often find there are several false starts, or ideas that wind up not meeting your desired vision. When you’re presented with this, as an artist, you can either lean into it, taking the ‘meandering’ approach, letting the algorithm and your eyes and your impromptu tastes dictate the direction of the project. Or you can wrestle with the beast, steering it into the direction you want or starting again.&#x20;

Both have their merits and I think in general the latter has gotten some bad rep associated with it. It feels less ‘natural’; more friction, less marketable as the artist-savant ‘oh this piece just evolved from my unconscious dreaming with the computer’ whole thing. I think that a lot of the generative art field is driven by these happy accidents. I also think that if we want to leverage, understand, and master this art form then being methodical and bullish is also necessary.&#x20;

I try to take an approach where I have an end vision in mind and relentlessly work towards that, I don’t aim to hit that target, but just use it as a catapult into uncharted territory. Someplace further than blind playing would lead, and someplace more spontaneous and unanticipated than a rigid design implemented generatively. At least that’s what I intend.

The project started out as a pfp, as a torso-sculpture, as an explosion. Each of these ideas came from different places; a vision I had in my head, an experiment with my previous algorithms I wanted to develop, a design I was tinkering with from scratch. Ideas can come from anywhere and many of these I will probably pick up at a later point in time, though they just weren’t working at this one.&#x20;

![Ideas explored. To be used another day.](ipfs://QmP7n1q5pxe2ab6KpxmzM6tYsxhYYQ5FLciopnY8hW5uhF)

I usually have many experiments going on, coded, drawn, painted, or written down. The important thing is that they are caught and collected. It’s difficult to be constantly working on creative projects but I find this technique to work very well.

Whenever you have an idea, record it. Sketch, code, word - whatever you like. Store these ideas in a central place. Revisit them when you need a new project. That way, when inspiration strikes, you can capitalise on it and develop it later. Instead of staring at a blank canvas waiting for the lightning bolt to strike, you start by iterating on an existing idea you had, maintaining your excitement for new work.&#x20;

![Captures ideas. left: sketchbook. right: digital paintings.](ipfs://QmZqEWRWiFCxTQo5uSTDgCH27hsk1nSWbXNHJFLtVyQTzK)

This works for me but maybe not for you - I encourage you to try new methods, experiment and never be too hard on yourself when you fail, failure is vital. Sometimes we win, sometimes we learn. Treat yourself honestly but also gently. It’s okay to start over. 

Leviathan, as it eventually came to be, started with my third attempt at a goal. It was a design-based approach I hadn’t taken before. I would start by reproducing an existing idea and iterate from there. I gathered research material, images, and artists I wanted to emulate the mood of. Using their images, I started with basic black and white squares and tried to emulate their compositions, energy and mood. Keeping it very abstract and minimal, I didn’t want to be spending a long time on an algorithm, only for the design and final output to look rubbish.&#x20;

![Initial research and generative sketches. software: PureRef.](ipfs://Qma67DRU8uP7wsfXFPs6zKeJrZwtp8ZfuzPSdBapWUbN7q)

A design first approach that will continue to evolve past the initial target. After several iterations I had some basic tonal studies I was really liking and wanted to take further.

During this back and forth between my studies and my references, an idea started to form. I latched onto this feeling of creating something monolithic, something in the sweet spot between abstract and figurative, where I really love to play.

A titanic form falling from the skies, emerging from the mist. I knew I had seen this before. It was vivid in my mind.&#x20;

It was a painting I saw when I was 16 or so at the Tate Modern - Anselm Kiefer’s Lilith (which the project was originally going to be called). I remember I was blown away by the scale. It was ethereal, the painting was so massive that you were dwarfed, made into atoms when viewing it. All you could see was the detail - it was only while standing the length of the whole room that you actually saw the entire piece, an arial view of a city.

![Lilith by Anselm Kiefer](ipfs://QmWBwJ5AABTfSMMYfbR55FAgQfo3gxueFzfm2uQCMWd84W)

I fell in love with this effect. This was the subject matter that guided much of the evolution of Leviathan. A mood, an ethereal form.

![Ethereal explorations.](ipfs://QmWpiZvnhZv9bNCPBdrodYmWjozURzvjf8rVyXgHGpYqJR)

Creating more and more work, I’ve come to realise there is this kind of mirror between you and your design. I’ll often be working on something, only to realise I’ve been making something from my childhood or an idea I forgot from when I was younger. Your work reveals parts of you. You find them along the way. You cannot help become part of what you are making. It’s what separates you. And I love leaning into this.

I was originally going to have much more open ended and varied outputs. I decided against that and went with a mirrored composition that I would add asymmetry to. This appealed to me because I wanted to challenge myself to create something self-similar but with great variety within. It was a decision that perhaps amputated it’s ‘long-formness’ but I think it’s important to listen to yourself as a creative even if and especially if it goes against genre standards and you have a gut instinct about following it.

![Final initial designs.](ipfs://QmXMfQ4yHubtpyxtTsNyxyBNtSwrESiKPpmiPL7UgGv9hv)

Making something that intrigues you is the most important thing. Take lessons from your peers to get better, but you must love what you are working on. A reaction of ‘200%!’ when you see your work. At every stage of the process if I’m not a ‘fuck yeah’ when I see the outputs then I know I’m in trouble. The ‘fuck yeah!’ test never fails.

***

# 2. The Algorithm

Discovering the main algorithm and design of the project is step one. After that the piece starts to grow on its own. You have rhythm and new, novel things occur every session. Its relentless and addictive. This is the fun part.

![Small changes can have massive effects on the outputs](ipfs://QmVfWFrfrmvM3UdYa7R8oKSXtCFKRsDkzTgLr3CQJctCjd)

The whole project uses sampling. A technique where you randomly pick a point within your canvas and render something at that point. It is akin to ray-casting/marching.

Using this approach, I was able to create a lot of atoms and detail which added to the sense of scale of the piece.&#x20;

I subdivide the canvas into regions, dictated by various noise functions. Depending on the combination of noises, the sample would land on a render method, deciding what was drawn there.

From this, combinations of draw functions dictate what gets drawn (ellipse, square etc) while the noise functions dictate the pattern of those primitives (where they are). There are several templated noise functions with a variety of properties; some taller or twisted, others rectangular, they form a kind of function swatch I dip into and paint with.

![Noise functions and simple applications of them.](ipfs://QmQVf9qB6shESjfvPNBKf7GMu9E731x39WMMmANHAUSuwu)

![Noise functions applied to forms](ipfs://QmaPZQByx7eD6umVsQP7ZjmpMauDeH6yifsuStVaAP3szi)

The total passes used were:

⁃            A function that dictated the ‘sculpt’ - the form set apart from the background pattern.

⁃            A function that dictated the primary patterning on the sculpt.

⁃            A function for the texture overlay.

⁃            A function for negative space overlays.&#x20;

⁃            A lot of functions specific to the chosen render primitives at the given sample point.

All of these noise functions are added, multiplied and distorted with each other, leading to various combinations of effects, some mirrored, some not.

![Final compositions](ipfs://QmbRAYH4s1FvKKA3TVa1cgmAy7shmoRyt7Kdxf5G5uAeVx)

 

***

￼ 

# 3. Algorithms That Scream

When asking from critique from other artists, a lot of people said that it was very aggressive. I was so happy to hear this.

It is. Leviathan is an aggressive piece compared to my other work; Angel’s Egg, more energetic, Gaiatica, more poetic and delicate. I knew I wanted to do something different with Leviathan.￼

![Gaiatica and Angel's Egg.](ipfs://QmfYzZgwLiA4VmYnEoHktkFiZY6w3LPe1sYhJmhAVJ6NmJ)

::tezos-storage-pointer[]{contract="KT1U6EHmNxJTkvaWJ4ThczG4FSDaHC21ssvi" path="token_metadata::602676"}

::tezos-storage-pointer[]{contract="KT1BJC12dG17CVvPKJ1VYaNnaT5mzfnUTwXv" path="ledger::15432" data_spec="FX-GEN-DATA-002"}

During the project, I found out my sister had died, quite suddenly. I struggled a lot. I couldn’t create for a long period. I thought I could find solace in art making but nothing was able to come out. Anyone who has had a sudden loss can understand. The immense shock, grief, anger. So many days spent in a haze. 

But as well as the bad, there’s also a kind of euphoria, a kind of manic happiness. Gratitude for life and having loved them. Everything is saturated, in bloom, brighter. It puts things in perspective. When I eventually made it back to work, I think this bi-polar feeling seeped into Leviathan, it’s like: ‘yeah I’m fucking here, I’m alive, here’s some fucking art to prove it’. It needed to be an aggressive life-affirming thing. 

My artwork always has an emotional drive to it, it’s ultimately got to resonate with me on an instinctual level in some way for me to pursue it - I love to explore this with generative art as I feel that it is underrepresented in the space, I want to see algorithms that scream.

![A bit literal but you get the idea.](ipfs://Qmbkjm3wKMnMGJ7tzJEh4RGWAFVox8Y3g46fbBouGW7UY1)

 

***

￼ 

# 4. Patterns & Meaning

I based the algorithm on a reflected pattern, this was to aid in the viewers comprehension of a form while keeping it abstract, a skeleton to help them project. I think that because of this the piece becomes a mirror of the viewer. They see what they want. I like to provide an ambiance and gentle guide through the title and project description but for the most part I leave it to the viewer. I feel context enhances work and I never want to say exactly what my work IS. ￼

![](ipfs://QmVs7g7fWXbuguFXaYpufruqRvbTUknZtHhwLKrcFXYxhf)

Ultimately the visual either resonates or not, definitions only belittle that.&#x20;

I love that bit of what I do, seeing what part of how I feel has been transmitted to individual viewers and what part is totally them. These marks are indifferent, there is no form apart from the ones we see in our heads when we look at the piece. 

It’s like a chemical reaction that I start and you finish. A colossal dream we’re all sharing together. A leviathan.&#x20;

I hope you enjoy it.

::tezos-storage-pointer[]{contract="KT1U6EHmNxJTkvaWJ4ThczG4FSDaHC21ssvi" path="token_metadata::1211490"}


The creative processes, algorithms and journey behind Leviathan - the latest generative collection from S.F.Batchelor.

stuart faromarz batchelor. generative painter

Leviathan: The Creative Process

Written by

Editions

Share

Details

stay ahead with our newsletter