ProFantasy Community Forum

Fractal Terrains observation and suggestion.

illuminatispyilluminatispy Traveler
in Mapping with ProFantasy's Software
One thing I've noticed about FT's world generating is that brownian noise makes great land masses but horrible mountain ranges and ridged multi-fractal makes great mountain ranges but bad land masses. Is there any way to do a two stage generation where the land mass is generated by brownian noise but only up to a certain altitude then a ridged multi-fractal is generated then the two are merged with the ridged multi fractal going over the land mass when sea level or higher but then would go "under water" when bellow sea level.

I really think it would generate more realistic land masses and create the illusion of plate tectonics.
Share on FacebookShare on Twitter

Comments

  • MonsenMonsen Administrator 🖼️ 81 images Cartographer
    Slayton knows more about the depth of the program to tell you if this will ever be possible, but as far as I know about the current version of FT, there is no way to accomplish this in the current version of the program.
    Share on FacebookShare on Twitter
  • jslaytonjslayton Moderator, ProFantasy Mapmaker
    There are lots of things that can be done, but I don't know that what you describe is really the best overall option because the ridges will be placed oddly. Plus, the ridges in the RMF type really aren't much like mountains except in a very simplistic way. You should be able to get something similar to what you're after by computing an fBM world, burning it into the surface, switching the world type to RMF, and then doing a global set of roughness to 1.0 (or maybe 0.5). That will add a layer of RMF noise on top of the existing fBM noise. I have a feeling that it won't quite be what you were hoping for.

    FT users fall into two camps: the "I need to recreate this exact world that I've had for forever and I want F to do all of the heavy lifting" and the "I want lots of plausible random worlds". The first camp wants the ability to be able to paint the world in broad strokes (continent here, mountains here, forests here, etc.) and have the system compute plausible results from their inputs. I have an approach that would work for this technique, but it's not terribly compatible with how FT likes to operate and takes a long itme to compute. The random worlds folks seem to be reasonably happy with FT, but they'd like more plausible details than the software generates now. Trying to find useful compromises between these two vocal groups is a tough task, so I often just do what I think will be fun. That sounds awful, but people seem to be able to put my new toys to use in each release (consider the Terraformer project; it was an effort by a user to elaborate on what was a simple toy that I put in on a whim).

    If I were starting out to develop FT today that I would do things quite a bit differently. For example, I would store editing data differently (I'd use a pseudo-equal area projection to store data) and I would change the computation model (I'd make sure to take advantage of GPU computing). One thing that I can't do is break the last 15 years of people's work (it happened once with a timing-sensitive bug and I'm trying not to let that happen again).
    Share on FacebookShare on Twitter
Sign In or Register to comment.