Still working with the ink trails. Most of the tweaks are just me trying to figure out a more robust flocking algorithm. Kinda moving in circles but I am not going to let that stop me.
The problem remains: How can I use perlin noise as a flow field to control the movement of the objects without it looking like it is perlin noise controlled? I haven’t quite figured that out yet. I am thinking I am going to add a second influence (well, actually a third influence because I have added magnetic repulsion collision avoidance) to help out with the movement. The first thought is that I can use an image as a modifier for the flow field. The objects will still move primarily due to the perlin flow field but they will also reference the part of the image that they are ‘above’ and perhaps the brightness of the image will influence the turning speed or altitude. Haven’t tried it yet.
These images have some successes in them. Some look ghostly or scratchy, but only in sections. The overall appearance is lacking and it isn’t until you zoom in that you can find decent compositions and nice textures.
I will continue to tweak. I hope to have some progress to post shortly.
looks amazing so far.
would be nice to see it bigger.
i assume this happens over time, so why not making it soundreactive?
not in a way that it reacts in realtime to a beat, which is obviously the wrong thing for an image. i’m not talking about visualizing the sound.
but i found out that sound is pretty good data for driving things. it changes over time and you usually have lots of sounds to try to make it look different. so timebased data might be an alternative to static, spatial data you get from an image….
Woaw Robert!
Crazy dude! This one is really incredible.
Reminds me of old zen caligraphy.
The perfect inperfection of the stroke.
Something I always thought impossible in computer graphics but it seem you like to push the limits each time a bit more far. Keep up the good work man! It’s always great inspiration for me.
Cheers!
Simon.
When I was helping develop a fur system for a film I was working on a while back and we needed to break up the fur with noise to make it look scruffier we found that using 2 or 3 sets of Perlin noise with different prime numbers for frequency and amplitude helped get something which looked random and didn’t have a “Procedural noise” quality to it. I don’t know if that would help in your case but I thought I’d throw it our there in case it’s of some use.
Lenny: Audio is definitely something worth trying. More than anything though, I am trying to make the flocking algorithm stand on its own as something that creates organic movement based on very few rules. The current flock essentially does three things. 1) Find a direction of flight based on a perlin noise flow field. 2) Push away from other objects that are really close to avoid overlap or collision. 3) Push away from the boundary sphere to keep it from straying too far. I guess it wouldnt be too weird to have the perlin noise be effected by audio. Maybe I will try that out this week.
Simon: Dankë! Glad you like. I thought of calligraphy too, especially when viewing the detail images.
Mr. A: Thats something I had not considered. Makes sense. I should probably incorporate some sort of toggleable perlin visualizer because its hard to imagine the flow field… I need to see it to see how to tweak it. Will try multi-layer noise this week.
These are phenomenal, good work!
Hi Robert,
I was searching for a late night inspiration hoping that I could find something that would help make sense of my own project… FOUND! Im currently in my first year of spatial design working on the generation of a “Space”. So far I have used the idea of random occurances actually ordering systems. I used ink splatters as co-ordinates for a site, and also used the density to decipher how tall the ceiling is at that point (so far it looks like a pavillion with a roof made of triangles!)
My feed back was to study the ink more in not just visual terms… so here I saw your work mentioning audio and magnetic (went way over my head) but could you recomend any experiments or tests i could do to give my “space” a bit more validity?
Thanks!
Carrie