Socorro -- Refined New Method w.o. DEM

[Peter Toth]

Hi Guild,

Here is my latest progress with the rendering of semi-realistic topography without relying on existing satellite DEM data. I admit my specialized method takes 10 times as long, but the sweet satisfaction of blazing a new trail keeps my motivation to use this particular approach very strong. There has got to be a way to do this, and I'm determined to find out how.

By the way, I believe I might not have enough "lowlands" in this rendering, but here it is.

If you have any comments or criticism, I'd love to hear it.

Thanks,

Peter

[Redrobes]

Look good. What were the kinds of refinements being used to make this one over the previous attempts ?

[shadixdarkkon]

I would generally agree that there aren't enough lowlands, but it is hard to tell without knowing the scale and climate of the region. Overall I think you have the naturalistic shapes of valleys and such down, which I am envious of. Keep up the good work, I admire your projects.

[waldronate]

Without some sense of scale (both horizontal and vertical), it's tough to make a guess at how realistic this image is. The canyon vertical profiles and how they force the rivers to run directly down the centers of the valleys offer some suggestions about the scale, but that's largely an artifact of your erosion sim, I would guess. One of the larger fails of Wilbur's erosion sim, for example, is that terrain is the same hardness everywhere. It's like eroding a big hill of mud (or a large block of solid, uniform material). It's also missing a true mass wasting or thermal erosion sim, but you can sort of fake those by using its precipiton erosion feature with a short max length parameter (1 to 5) or lots of selections using relative elevations.

Like Redrobes, I'm curious as to what your processes are and how this one differs from previous attempts. I am particularly interested in what you're starting from (that is, what you provide to the first tool in your work chain) and what you get out of each tool before feeding it into the next one. The teaser about taking 10 times as long is interesting, but I can't even begin to guess what your 1x baseline would be.