Hello,
do you think it could be possible to make something like a ‘Fraktal Perlin’ device?
Hello,
do you think it could be possible to make something like a ‘Fraktal Perlin’ device?
Is not a Perlin Noise a type of Terrain related fractal.
I don’t know… but I don’t think so.
And even if - it’s probably not the recursive style fractal X-Pax thinks of.
I think that is possible… for randomisation you’s have to use pseudo-random numbers (as I mentioned in the river-thread about the implementaion of L-systems)
What exactly are you referring to by “Fraktal Perlin”?
I found some information here.
A perlin is a Fractal Procedure but like nikita said perhaps he is thinking of a Koch snowflake which more complex versions of this mathematical curve can create the Lindenmayer system or L-System, example River Networks, Vegitation, Frost, Lightning etc…
Overall Good Information on Fractals.
Here’s what I guess I’m not understanding about the question:
If you want a perlin fractal, put down a “Perlin Noise” generator in World Machine.
“Perlin Noise” as used in WM is actually a perlin fractal; Perlin noise technically refers only to the single-octave noise pattern. The Perlin FBM combines multiple octaves of noise at different frequency scales to create the terrain.
So I was curious if the original poster meant anything different from regular Perlin FBM noise?
I believe he either means a set of calssic fractals like the mandelbrot or julia set.
Or he thinks of a more random fractal landscape.
To build up such a fractal landscape you could use a recursive function. This is a part of the mandelbrot set:
http://www.jaggi.ch/informatics/fractals/mandelbrot_old_style/m16.gif
If you’d choose the colours in a way that the fractal became a greyscale map you could use it as a heightfield.
What I noticed is that after erosion (without channels), most heightfields “look fractal” if you look at them as greyscale maps.
A fractal generator therefore would produce landscapes that already have that eroded look without having used any erosion.
@nikita
You are right. I really think about a landscape which looks like a julia fractal or a mandelbrot fractal.
And therefore I think about a new perlin device.
Oh. Madelbrot and Julia classic Fractals.
A Madelbrot and Julia only Generator would be intresting to produce nikita. I observed these simularitys as well and i can see alot of these types of fractals looking at standard erosion with Rigid Multi-Fractal Perlins.
Isn’t the Rigid Multi-Fractal Perlin just a more advanced variant of the older classic Madelbrot and Julia fractals with the FBM Noise?
http://www.ashundar.com/datas/users/1941-madelbrot%20and%20julia%20like.bmp
Speaking of which i came upon this as well. Looks like development has stopped for quite a while now and the user interface in kind of set up wierd so i never used it for anything serious.
http://raieh.homelinux.net/cgi-bin/raieh.pl?FractalWorld
http://raieh.homelinux.net/cgi-bin/raieh.pl?FractalWorld/SampleImages
hmm, well I think there is “being” some confusion arround here. The Perlin noise is closer to “using fractal stuff”, compared to what we are probably thinking of Mendelbrot and Julia “are”… What I mean to say is that when I think of a Mendelbrot or Julia devices, in the context of WM, I am thinking in a heightmap that resembles them. So to say that If I looked at a greyscale image of a terrain seem from above, I would see the same as I see in a Mendelbrot or Julia fractal image…
The Perlin does not use this fractal similarity on a 2D-image basis like the fractal image generators… It uses in 1D for height!.. …I think… I mean, the Perlin is a sum of waves (noise), almost like a modulated radio signal… You add up a bunch of “frequencies” and get a messy terrain-looking shape
Julia and Mendelbrot images are computed from colouring the regions of “space” with the same colour for a same value… So I’d expect a terrain made purely like this would span from a height of -infinity to +infinity, kind 'a like poeticly connecting the Heavens with the deep… However, one might think of clamping it or using a function that converts the infinity scope to a ranged scope (like a sine function :), it might look interesting but there my be problems with hight frequencies)
Something to think about… But I agree some shapes look pretty nice and usable!..
One last thing: My knowledge of Julia and Mendelbrot sets is more than 10 years old… So I might be a bit too narrow minded about this…
In 1975 Mandelbrot coined the term fractal to describe these structures, and published his ideas in Les objets fractals, forme, hasard et dimension (translated into English as Fractals: Form, chance and dimension[2]) in 1977.
Neat Stuff. I love Fractals!
@Sethren
Have you heard about Chaoscope?
It is free tool to create some nice looking fractals (in this case the word attractor is used).
You can find it here:
http://www.btinternet.com/~ndesprez/index.htm
The link section on that site contains some interesting links.
Yes i have, it’s very neat and i need to break out my copy on one of my CDs
I am keeping a close eye one Xenodream as well. It’s 80$ but well worth it IMHO. This software actually creates 3D exportable geometry from layers of it’s fractals. Version 1.6 is due out very soon.
http://www.xenodream.com/mainindex.htm
Another nice freeware Fractal Program as well is Apophysis. This one does flame fractals. This is my third favorite fractal software next to Ultra Fractal which is second.
For more information about fractal flames.
Perlin noise can be used to build stochastic fractals, but the noise itself is not fractal. Fractals by definition are self-similar over a number of frequencies.
Recurrence fractals (like Mandelbrot and Julia) are beautiful, but unsuitable for terrain generation because they produce values near min and max at all frequencies. Imagine Mount Everest as a billion rock needles 1mm in diameter pointing up out of the ocean; that’s a Julia fractal as a terrain.
Ahhh, I get what you are saying. It makes sense to me.
Welcome to World Machine Lee.
nice explanation, Lee!