This is a follow-up to Planet Zoo.
At one point, I decided to show my parents my planetary zoo, and they said that it would be really nice if it were not interactive, but just static. Like a slide show. The 3D effects seemed not to mean anything to them at all.
Well, I took it upon myself to actually try the idea out.
I basically copy-pasted half the code from the other program into a simple, 2D canvas, that used a scan-line approach to draw colours onto the screen. I chose a simple perspective, standing somewhere a few hundred meters above the surface of the planet and looking off towards the horizon.
At one point, I decided to show my parents my planetary zoo, and they said that it would be really nice if it were not interactive, but just static. Like a slide show. The 3D effects seemed not to mean anything to them at all.
Well, I took it upon myself to actually try the idea out.
I basically copy-pasted half the code from the other program into a simple, 2D canvas, that used a scan-line approach to draw colours onto the screen. I chose a simple perspective, standing somewhere a few hundred meters above the surface of the planet and looking off towards the horizon.
It seems that I no longer have any of the first renders, which is a real shame. Doing them pixel by pixel, they were low resolution, took forever, and looked kind of plasticky. So instead, I decided to draw a fluffy cloud at every 10x10 or so point. This gives the images a soft feeling, like they are landscapes of the imagination, made to be painted, rather than planets. Then, adding in some extra effects, like the fluffy clouds having a small amount of randomness to their colour, it started to really look like something!
Evaluation
So it basically looked like this at the end of the first day, but I kept believing, adding small touches and trying to polish it in so many ways, and nothing really changed. It still did the same thing, over and over again.
This whole project, together with the planetary zoo, being a two-for-the-price-of-one is its saving grace. If I had not been able to reuse most of the cody, I would have been a bit dismayed with the results.
Thing is, I had made a generator of just one visual piece: The Romance Landscape. Every single image it spit out was basically a recolouring of the same landscapes repeated again and again. Without a human hand putting skilfully putting things together, seas, mountains and clouds can only ever be so beautiful.
This picture below is actually the most important thing I got out of this project:
It is an error, quite clearly. But despite this or perhaps because of this, it is much more interesting than all the previous pieces put together. Its composition is far from great - the colours clash quite a bit. But at least, it is unique. It feels like it is something on its own.
I am quite convinced that the exploration of these concepts is what led me to the inspiration for the Convolutional art generator.