Thread: [WIP] This Orb: world-building and mapping project

OK, in trying to move on to the next step, "elevation" i've hit a couple problems. I looked through all your tutorial threads (at least looked at all the titles), found some good tips, but ultimately didn't find the answers to my problems.

1)
I'd like to create a realistic elevation map, along the lines of this:


I have the elevation roughed in some detail like so:


But obviously the level of detail is dramatically less. The obvious ways i might try to achieve the effect are mind-numbingly tedious, and i don't want to embark on one of those paths if there might be a better way.

a) Paint it by hand: tedious, but eventually i'd get what i want
b) Copy/Paste bits of elevation from our world: possibly faster, but getting stuff to match with my coastline, and planned mountain locations may be impossible.

It looks like some people use Wilbur to make erosion-effects, but that's a Windows program, i have a Mac, with limited Windows emulation, which can't handle Wilbur.

Advice? Helpful suggestions?

No matter how i do it, i'll probably retouch some of it by hand. Can anyone recommend some good custom brushes for this kind of effect?


2)
I also seem to have a chicken-or-the-egg type problem. Looking over Earth's elevation map, i get the impression that erosion patterns are much more strongly defined in wet areas than dry ones. But of course you need to know elevation to figure out your rainfall patterns. Is the difference really going to be significant on a global scale? How have other people dealt with this?

Originally Posted by jwbjerk

I also seem to have a chicken-or-the-egg type problem. Looking over Earth's elevation map, i get the impression that erosion patterns are much more strongly defined in wet areas than dry ones. But of course you need to know elevation to figure out your rainfall patterns. Is the difference really going to be significant on a global scale? How have other people dealt with this?

For the first question, I'm sorry, I just got the same problem of detail level...

But, for your question, I talked with a geologist about such a similar matter. He said to me that the erosion patterns could be approximate this way: In the beginning the erosion is quite everywhere the same. What really matters is the geological soil. On a long scale it tends to diverge from regions to regions. For exemple after at a certain time, a region will be quite flat because of its soil which is far more tender than another region. And so you've got the first step to define your rainfall patterns! I think you'll have to work those question together.
I hope I was understandable enough...

If you have access to DEMs then B is the way to go...just copy n paste and blur n paint until they fit seamlessly. Otherwise you're with the rest of us looking for a way to emulate a DEM...finding some techniques or doing it by hand. On the erosion, it shouldn't matter too much for a world map because you can only get so detailed in the image. By and large the rest of us just use informed guesswork and make it up as we go.

Originally Posted by Guldaroth

But, for your question, I talked with a geologist about such a similar matter. He said to me that the erosion patterns could be approximate this way: In the beginning the erosion is quite everywhere the same. What really matters is the geological soil. On a long scale it tends to diverge from regions to regions. For exemple after at a certain time, a region will be quite flat because of its soil which is far more tender than another region. And so you've got the first step to define your rainfall patterns! I think you'll have to work those question together.
I hope I was understandable enough...

If i understand you, you are saying i shouldn't worry about eroding high rainfall and low rainfall areas differently, because climate and rain patterns vary, and at some point in the past all the different parts of the planet will have gotten some rainfall. Is that right?

Originally Posted by Ascension

If you have access to DEMs then B is the way to go...just copy n paste and blur n paint until they fit seamlessly. Otherwise you're with the rest of us looking for a way to emulate a DEM...finding some techniques or doing it by hand. On the erosion, it shouldn't matter too much for a world map because you can only get so detailed in the image. By and large the rest of us just use informed guesswork and make it up as we go.

Is there any different between a "DEM" and a "bump map", like these: (1. Terrestrial elevation data)? He's labeled the 16bit versions as "DEM"s, and the 8bit version as a "bump map", but i would have called them both "bump maps". Maybe you would call them both DEMs?

16,200 pixels wide is the largest i've been able to find a bump map, and while it's 4x bigger than my map, would anyone recommend working from something larger (if it's available somewhere)? Seems like my elevation might be much less obviously part of Ethiopia (for instance), if i copy/pasted a piece that was more zoomed in. Of course the scale of the features would have been changed, but i expect such geologic features are "fractal" enough, that they look similar at different scales.

There may be some technical disparities but I don't know what they are...someone here surely would know. I guess bump-maps could be applied to anything like walls or car hoods or snack boxes to give the surface some texture while DEMs are earth specific. I'm guessing there. They're the same thing really in application for what we do with maps.

Originally Posted by Ascension

There may be some technical disparities but I don't know what they are...

Then it probably doesn't matter.

I got to thinking about brushes made from DEMs (or elevation bump maps)... after some experimentation, it seems pretty promising. I doubt they can produce perfect results, but it should look a lot more authentic than round, fuzzy brushes.

It is indeed time-consuming, but actually i'm pretty happy with the copy/paste elevation method. It has a way of including interesting details that i hadn't intended, which makes it more fun, and increases the authenticity of the final product. Of course i still have my larger, more complicated continents remaining. The true test is to see how much of my enthusiasm remains for this method at the end.

I've more or less completed two of my smaller continents. I don't think it will be nearly as easy as i feared to spot the terrestrial origin of the pieces i "borrow". IIRC i used 4 pieces of Earth (plus one for the island on the right)-- anybody care to guess which ones?

Radial gradients are a great tool to pitch the elevation of an area in the direction you want.



P.S. Interesting thing i've observed. Lots of imaginary maps have continents with mountains roughly in the middle, but earth doesn't seem to work that way. Almost all the main watersheds are jammed up against one side of a continent, usually with a smaller watershed opposite.

Originally Posted by jwbjerk

P.S. Interesting thing i've observed. Lots of imaginary maps have continents with mountains roughly in the middle, but earth doesn't seem to work that way. Almost all the main watersheds are jammed up against one side of a continent, usually with a smaller watershed opposite.

As though moving plates accumulate long piles of stuff on the side that's undergoing collision vs. the simple fractal models that use a function that's the same in all directions, which means that the high spots will be in the middle. It's a good observation that many people never make.

Nice work on the altitudes, btw.

Thanks.

I think the "high in the middle" generalization also tends to be true of the more easily observable chunks of land, i.e. small islands, so it's understandable how the generalization takes root.