This may not be the appropriate time to inquire but i am curious if there are any plans to further update this shader?
However, interestingly enough, i think that i might have found an alternative solution to all of this. I will post some results soon.
I used an Alpine fractal shader with reduced details for better looking erosion calculation. It seems to work quite well using the fractal patterns from that shader and enhancing it for a better sense of realism.
Sounds interesting. For distant alps I often (well, often... I rarely use it) raise the 10m minimum anyway to 100m or so, and add some fractal.
It was working at first but i noticed what appears to be a limitation in the Alpine fractal, and that is true Hanging Valleys. The patterns are very good for such features but the valleys do not look higher then the lower larger scale valleys. Eroding is still a bit hard because not all of my smaller valleys are well defined. I will note that i am working on functions to imitate somewhat such mountain fractals with more defined hanging valley features so i can share that later to see if anyone has any better solutions to this.
Here are some examples...
Here is another example in red is where the elevation is higher being the hanging valley then the main valley below in green and what i am after.
What comes to mind is using the erosion shader in 2 versions; a normal for the total inclusing hanging valleys, then a lower octave one to make the lower broad valley. I dont know if that would work...
I did erode on top of erode but i did not try the lower octave idea. I'll have to see what happens with that.
Tried it and the features of the Erosion fractal does not allow the gullie lengths to be long enough which is what i was using for the larger valleys. So it produced the wrong type of pattern. It would need the capability of having very long flow channels spreading out and breaking down even entirely flat terrains. In that instance long valleys can be made. Having a flatness feature would help also for the slopes like softer erosion as in plains. Even a way to have the erosion effect the terrain in a way to have rigid mountains to more soft eroded plain would be neat. Another thought is enhance Flow map B to go from a relatively straight river from the higher terrains then meandering along the more flattened terrain regions like having a built in adjustable soft warper.
Too bad. Not every theory works, I'm afraid. There's lots of things I conjure up in my head and try, but fail to work for mysterious reasons.
No problem. There is still enough to experiment with for now. Certainly there will be nice new things to test in the near future within Terragen.
We're always working on new ideas, but so far we don't have a specific and overall beneficial update for the Alpine Fractal.
- Oshyan
No doubt so it's all good. I'm not to concerned with the Alpine fractal. The more i explore, the more other alternatives present themselves. Not all of the time but more times then not.
The lead-in octaves for the Alpine shader, how does one know how large the main features are and does it multiply the Feature scale itself for every octave added?
Anyone?
Not me, sorry. Don't have a clue :-[
No problem. There is no information on the Wiki about this so maybe someone from Planetside can weigh in.
The goal is still the same but struggling with proper scales and patterns. The picture gives a better indication. Red marks long wide valleys, green are those hanging valleys that a hard to get and blue being the glaciers is a might not happen.
Can anyone weigh in on the scaling issue from Planetside? I'm still trying to figure it out. Thank you.
Quote from: Chris on December 08, 2015, 11:41:03 PM
The lead-in octaves for the Alpine shader, how does one know how large the main features are and does it multiply the Feature scale itself for every octave added?
When "Lead-in octaves" = 0, the largest cells occur at the scale of "Feature scale".
Additional lead-in octaves add cells that are larger than "Feature scale". How much larger depends on "Scale step" on the Settings tab (although this setting also controls the scale ratio of smaller octaves as well as largers ones).
For example, if "Feature scale" = 4000 and "Scale step" = 3:
Lead-in octaves 0 => largest cells are 4000 metres across.
Lead-in octaves 1 => largest cells are 12000 metres across.
Lead-in octaves 2 => largest cells are 36000 metres across.
... etc.
Lead-in octave cells have the same displacement amplitude as the cells that occur at the "Feature scale", and they also contribute to the flow of smaller cells in a different kind of way. So when considering how steep the mountains are, where you want to think about the relationship between amplitude and scale, it is better to think about the relationship between "Displacement amplitude" and "Feature scale", but not the scale of lead-in octaves, and that's why I chose to parameterise it this way.
Matt
Thank you for the in-depth reply. It gives me a much better idea of what the scale perimeters do. By the way you describe the largest cells, visually what i had seen did give me a fair idea of the sizes i needed but a good explanation of that so it certainly helped a lot. The eye tends to pick up easily what looks fake and what looks real so understanding scales can be important even on a localized level. However, i'm getting closer now. :)