Editing a world map

[Deadshade]

I don't have an effective technique for handling inland basins and lakes, unfortunately. The process that I typically describe is actively hostile to them because of the basin fill operation. One way that might be useful would be to make a mask that has just your desired inland lakes. Every few precipiton steps and after each incise flow, load that mask as a selection and set the value to the desired lake altitude. This technique will have the effect of enforcing the lake constraints, but the rest of the landscape will continue to slowly reduce in altitude overall (the precipiton erosion filter is a lossy filter). If you'd like to keep the lake at a relative altitude, move the mouse cursor over the lake area and then set the lake altitude to that value.

Is it not possible that I simply draw a canyon by hand from the lake to sea so that the incise has to follow this road too ? Else I will try the mask method.

The Wilbur shader is a pretty ugly beast.

Ugly I don't know . But impossible to learn sure. The texture tab contains 16 items out of which 5 with multiple choices which open farther on sliders buttons and values. So I estimate only this tab at something like 200 + different individual settings. Perhaps all parameters and settings are not all independent what would increase exponentially the combinatorics.

The V2 shader is similar in many ways to the latitude and intensity portions of the Wilbur shader, but with a major difference: the V2 shader combines several parameters that then used to select a color from the color list, while the Wilbur shader uses parameters to select colors from different lists and the colors are then combined. Using the parameters to select the color gives a result that's quite different in quality from the Wilbur shader. It derives from the idea that altitude and latitude are closely related in terms of vegetation (mostly due to temperature). The visible range is defined using Map>>Map Info; with top=90 and bottom=0, the full range of the color list will be used. all coefficients are normalized to the range of 0 to 1. offset controls adjustment to the map info, altitude controls the contribution of altitude, slope controls the contribution of the local angle, and noise is a factor that breaks up the basic linear patterns.

This seems less than the numbers above but still too much for Learning by trial and error. For the time being coloring is where I am completely stuck. I can't use Wilbur so I tried GIMP but seem to have there the same problem that I seem to have with V2.
When I apply the gradient on the map I linked above, I get basically only 1 color in the gradient. That means that for Wilbur there is no altitude difference and for GIMP no black/white contrast what is basically the same thing.
I thought that when I apply a gradient on a map, then it is normalised so that the right edge is 1 and corresponds to highest altitude and the left edge is 0 and corresponds to the lowest altitude. That way, necessarily, the whole Spectrum is applied regardless whether the delta is 200 m or 20 000. Apparently it doesn't work like that and I am stuck now. I will link tomorrow what I get when I click on V2 shader with the map linked above. What does mean V2 btw ?

The maximum resolution in Wilbur depends on your OS version and amount of memory. There are very few (if any) hard limits built into software. The 64-bit version should handle multi-gigabyte surfaces without problem. The main limit is your patience.

Right. I found out what the problem with the white edge around the map was. My fault.

I stop at 4000ish purely as a convenience because my patience is limited. About 4 posts up from this one is an example that went to 8000 (it's the one with the cropped V2 shader images instead of whole-world images). To get larger images, it's just a matter of keeping up the scaling process and the rest of the processing loop. Save regularly because Wilbur will eventually crash when it runs out of memory or hits a system limit.

Yes, thanks, thought so. However the higher you go (in resolution) the more you erode/incise because you do so at every step so that the mountains are less and less apparent.
That seems to be also a problem - if you look at the map linked above, the mountains are hardly more that thin lines. The height contour map which I started with had a much larger mountainous regions .

[waldronate]

You can certainly draw a canyon. I misunderstood that the lakes were intended to be at other than sea level.

Here we have another misunderstanding on my part. The attached image may help a little. The important concepts here are the difference between height, texture, and selection. The Texture menu has several only slightly related functions on it and they are related only in the sense that they dirty up the texture channel. Another case of leaky abstractions, I'm afraid.

One of the things that I'm always planning on making for the shader (and other) dialogs is an effects preview and popup sliders attached to the parameters to make exploring simpler.

[Deadshade]

You can certainly draw a canyon. I misunderstood that the lakes were intended to be at other than sea level.

Yes they do. The question is whether Wilbur "fills up" a lake with water when one or several rivers flow in automatically. I suspect that not because I understand that the fill basin command will precisely destroy any such basin ready to receive a river's water. So I must find a way to preserve the lake's location from being "basin filled" and in the end of the fill/incise loop render it blue by hand. Or perhaps there is a more sophisticated way I can't think of.

Here we have another misunderstanding on my part. The attached image may help a little. The important concepts here are the difference between height, texture, and selection. The Texture menu has several only slightly related functions on it and they are related only in the sense that they dirty up the texture channel. Another case of leaky abstractions, I'm afraid.

Indeed The sufficient and necessary condition for efficient communication is to have the same (or very similar) correspondence between words and concepts. As I progressed along this thread in establishing in partial domains a correspondence similar to yours, I learned much. Your page dramatically increased my understanding too - now the texture tab doesn't seem to me impenetrable anymore. I certainly still don't know what the individual parameters do but I have now a structured idea of the concepts instead of utter chaos. It is just formidable that you found the time to accompany me along the way. So thanks.

Now back to my basic problem with the V2 shader that I mentionned yesterday.

Here is what I get when I take my map linked above and do : Load the mdr surface file ->Texture->V2 shader


This might be a good Sahara image but not a planet.
So when I do V2 shader set up then Land colors, I see that the colors on the map correspond to the upper 5% or so of the Spectrum. The command apparently disregarded altitudes on the map and considered that for coloring purposes everything (mountains included) was very low around sea level.
Then I went in the shader set up and put 1 in the altitude parameter because I imagined that this would force the gradient to apply the color at gradient coordinate X to every altitude at X% of Max.
Well it took up a bigger part of the Spectrum (the tops of mountains became green) so it probably explained that the difference was due to the difference between the default 0.75 and my 1 in the altitude setting.
Now if this is right, it would mean that the gradient is not adapted to my planet - it produces a desert planet where forests exist only between some 60 and 80% of max.
If we applied it to the Earth, it would show forests only between 4 800 and 6 400 m. That is practically nowhere.

So if these hypothesis were correct, that would mean that I need to change the gradient and I tried to edit the color list. I saw there altitudes values of 0.25 etc till 1 and thought that I could simply change the altitude values by typing 0.1 instead of 0.25 etc and that way "push" the green to lower altitudes. But somehow I was forbidden to just change the altitude values so I gave up.

If the above is more or less right, how could I edit the color list ?
I dread some method where I should have to type in some array dozens of hexadecimal codes for colors ....
Of course the best would be if I could import gradients from Gimp because creating and editing gradients is extremely simple and fast there.

Supposing this problem solved, it stays that only a small part of problems is solved.
I could have a realistic altitude-color correspondence but as we already discussed earlier, the color depends also on latitude/climate.
So actually the semi realistic color array for a planet is not a vector but a matrix with altitudes in lines and latitudes in columns.
How can I solve this problem in Wilbur ?
I imagine that in FT3 it could be solved by using the climate color textures but I stay with Wilbur for the time being.

EDIT : I just found a function Histogram in the Window tab that I have never opened up to now. If it shows what I think it shows, then it is an extremely useful tool for a realistic planet. Clearly the height Spectrum of mine (corresponding to the map above) has absolutely nothing to do with a realistic planet. Instead of having a broad peak at very low altitudes (say 100 m) and then exponentially falling tail, it has a huge pointy peak at around 800-900 m exponentially falling to low altitudes!
I don't know how to eliminate the sampling of oceanic areas that swamp the land statistics but clearly this planet is a kind of high plateau (like a lower Tibet) everywhere with a few mountains and no Lowlands. It is then not surprising that only a very small part of the color Spectrum (namely the color between 800 and 900 m) completely dominates the planet.
As I applied the tutorial line by line, this would mean that there was probably (?) not enough erosion. I need 90% + of land below 400 m with a broad peak. Or so I think.

Actually wouldn't it be a good idea to monitor the erosion loops by looking at how it changes the statistics ?
To illustrate what I mean, here is a picture :
Attachment 68340

[waldronate]

The second attachment in your posting didn't make it through, unfortunately.

Try applying Filter>>Mathematical>>Exponent at the default settings to see how it affects your histogram.

[Deadshade]

Wow thanks ! This is rich with information - starting to play with it but will need time to get a feeling for all that can be done with it.
I don't know what happened with the attachment - when I click on it, it appears. Will retry :


I played with just noise , filters, erosion and selections to try to get the distribution from red to blue.
It worked, albeit slowly but I got much more interesting colors. As a byproduct I also got a nicely fractal coast but didn't see how and why it happened.

[Deadshade]

I found at last a solution on my fractal coast problem.
There is a nice free software called Autorealm that has several real fractal Tools.
So I redid the contours and have a much better fractal coast B&W map.

Unfortunately while this software is relatively easy to learn and has an intuitive UI, it is incredibly opaque as far as exporting images goes.
It is old so only 32 bits and must have some limitation on resolution that is mentionned nowhere in the help files.
So I am again struggling with something that must be a very basic problem - export the right format and the right resolution into Wilbur.
If I just save a .bmp and then load the bmp into Wilbur, I apparently loose the resolution of Autorealm and the coast has too big pixels so that the fractal look is partly lost.
The native Autorealm save formats keep the resolution nice and high but I can't export them into Wilbur.
I noticed that if I make a zoom on the Autorealm map and then save a .bmp , then the resolution is much better but I get only a part of the map .... from Charybda to Scylla.


So what I am trying to solve is how to import bits of map into Wilbur and then merge them together into a single map. When/if it is done, I will still have to bring the result in a 2x1 format for an equirectangular projection what is another problem.
Could you advice how these 2 things can be done in Wilbur ?


Btw the V2 shader tutorial was really usefull, thanks. However it is not possible to "break" the longitudinal color by setting f.ex a desert Inside a forest band , right ? If I need one then editing with Gimp is the only solution ?

[waldronate]

I'm not sure that Wilbur would be right tool for combining the input images. Something like The GIMP would be a better candidate for merging things and getting your desired aspect ratio.

The V2 shader only supports vertical coloring, sorry.

[Deadshade]

OK, these last days I mostly "played" with Wilbur to try to get a feeling of what some commands do . Especially your last short Tuto was useful. While "playingt" I noted the questions that were coming up and if you could shed some light, It'd be great.

1) The filter>fill mound is great for experimenting with mountains. But when I do one range and follow by a second one ,then I loose the first. Is there a way to keep all of them ?

2) In the fill mound there is a button "fractal parameters". Can you tell briefly what those are and do (H, lacunarity etc) ?

3) Why is recommended Left = 0 in the V2 shader set up when the map is Left = - 180 ?

4) Why is the (provided) V2 shader only hemispehrical : polar>temperate>desert ? Is it possible to change it in symmetrical so that the whole shader does :polar>temperate>desert>tropical>desert>temperate> polar ?

5) How can I create other shaders ? Now I can modify the one that is there but would like to create new and keep them.

6) When inserting a new Nth color in the V2 shader , the interval 0,1 is cut in N equal parts. Is it possible to change the interval (e.g some colors are wider than 1/N and some narrower) ?


EDIT : I have neglected FT3 these last day but when I wanted to export from Wilbur to FT3 it either crashes or tries to load but doesn't finish. I tried bmp and png textures formats in Wilbur. png crashes and bmp doesn't finish loading. Under which format shouldI save in Wilbur so that it can be loaded in FT3 ?

[waldronate]

1) The filter>fill mound is great for experimenting with mountains. But when I do one range and follow by a second one ,then I loose the first. Is there a way to keep all of them ?

There is a drop list for "Operation" on the mound dialog (and also on many others). This operation is how the new information blends with the old. Using "Replace" (the default) will completely replace whatever is already in the surface with the new values. Using Max will give you the highest of either the new data or the old data, which is probably what you want. The operations are (where dst is the value already in the surface at the current point, src is the computed "new" input value from the filter, and result is the new value placed back into the surface through the selection):

Replace: result = src
Add: result = dst + src
Subtract: result = dst - src
Multiply: result = dst * src
Divide: result = dst / src
Min: result = minimum(dst, src)
Max: result = maximum(dst, src)

2) In the fill mound there is a button "fractal parameters". Can you tell briefly what those are and do (H, lacunarity etc) ?

The attached "Calculate Height Field" PDF is the relevant section from the documentation. It's technically for the calculate height field dialog, but it's the same dialog used here as well.

The old documentation is mostly just missing new features; few of the older feature descriptions are incorrect.

3) Why is recommended Left = 0 in the V2 shader set up when the map is Left = - 180 ?

Starting at 0 makes the math easier for my little pea brain. In order prevent distortion in the noise function, the vertical:horizontal world space ratio should be match the pixel space ratio. For example, if your world space runs from 0 to 90 vertically (i.e. is 90 degrees high), it should be 90 degrees wide for a square map. For a map that's twice as many pixels wide as it is high, the world space should be 180 degrees wide. Where you start the left edge affects where the noise function starts, but is otherwise a convenience. Do note, however, that if you intend to wrap the results of the v2 shader onto a sphere that there is no way to generate the noise function on a sphere: it is a planar beast in this context. The V2 shader is intended to be a very simplified thing.

4) Why is the (provided) V2 shader only hemispehrical : polar>temperate>desert ? Is it possible to change it in symmetrical so that the whole shader does :polar>temperate>desert>tropical>desert>temperate> polar ?

Coding laziness is the reason. Plus, it was intended to be a close match to a popular mapmaking tutorial here at the guild. No points for looking at the peculiar blue-gray oceans and guessing, though. You can, however, make a color list that is symmetric and goes from polar through the colors and back to polar.

5) How can I create other shaders ? Now I can modify the one that is there but would like to create new and keep them.

I'm not sure what you mean by "create other shaders". For the V2 shader, you can use the Save and Load buttons on the shader setup dialog to save and later reload the setting. For the Wilbur shader, you can use File>>Save As with the type "lighting definition (*.lgt)" to save the shader settings and later reload them using File>>Open. The other shader types don't have save or load operations. It's not terribly consistent, I know, but it was almost 20 years between the implementation of those two shaders.

6) When inserting a new Nth color in the V2 shader , the interval 0,1 is cut in N equal parts. Is it possible to change the interval (e.g some colors are wider than 1/N and some narrower) ?

No, it's a limitation of the color lists that I've never gotten around to fixing. There is nothing that prevents you from adding two or three or ten of the same color to increase an interval's width or to proportionally decrease the width of another. One customer has a color list with 300 entries to give them a specific color-to-height ratio over a 0m to 3000m range. The color list files are just text files that can be modified with a text editor such as Notepad; the Data section is a list of RGB triplets ranging from 0 to 255 for each component. There are times when it's a lot easier to edit the color file than it is to do things manually in the shader editor (e.g. duplicating and reversing a list of colors).

[Deadshade]

"Starting at 0 makes the math easier for my little pea brain. In order prevent distortion in the noise function, the vertical:horizontal world space ratio should be match the pixel space ratio. For example, if your world space runs from 0 to 90 vertically (i.e. is 90 degrees high), it should be 90 degrees wide for a square map. For a map that's twice as many pixels wide as it is high, the world space should be 180 degrees wide. Where you start the left edge affects where the noise function starts, but is otherwise a convenience. "

I don't understand that. My map is an equirectangular projection so 2 wide and 1 high. Its resolution is 5000x2500 so also 2:1. And if I use the degree measure then it is -180,180 x -90x90. Again 2:1 so everyting seems consistent.
Now if I tell that (in degrees) left is 0 while up and down stay 90 and - 90. Then I define a map 0,180 x -90,+90. This is a square 180x180 with a ratio 1:1 what is inconsistent both with the projection ratios and pixel counts.
Or what do I miss ?

"I'm not sure what you mean by "create other shaders". "

I meant having an empty color list and then by clicking on colors fill it up. Then save it with a name. I realized that I could do it by removing and inserting colors in an existing shader (like the V2) but wondered whether there was a way to just have a blank page and fill it (like in GIMP).

"There is a drop list for "Operation" on the mound dialog (and also on many others). This operation is how the new information blends with the old. Using "Replace" (the default) will completely replace whatever is already in the surface with the new values. Using Max will give you the highest of either the new data or the old data, which is probably what you want."

Yes I saw that. I realize I confused global and local operations. I thought that the fill mound command created a mountain on a well defined place (a selection) and it stayed there unless i modified it. But apparently this command even if it is restricted to a selection (e.g a local place), calculates the whole planet what I didn't suspect. So if I understand well, creating, 1 range at north pole and then another on the s. pole, the "replace" means that the first one on the n. pole will disappear right ? This would (partially) happen with the max command too. The mode "add" should keep them both - it creates a range on selection and adds nothing to what is outside. Right ?


"There is nothing that prevents you from adding two or three or ten of the same color to increase an interval's width or to proportionally decrease the width of another."Yes that's what I figured out and tried first. The problem being in the "proportionnal" - I wanted to decrease only some colrs, not all. Hence the question whether you contemplated to make the range editable.
Btw Wilbur has 20 years age ??? In 1994 I had my first Pentium and RAM was was a few MBs - not much power to process images


Just for info. I invested a good 8 hours to create a true 2D fractal coast in Autorealm, then had to export it in 25 cut out pieces (because Autorealm has a horrible format handling) into Gimp where I put it again together in order to get an exact equirectangular 2:1 fractal map (Learning a lot about GIMP in the process) and last to export the finished result into Wilbur. So now I have a true 2D fractal coast in Wilbur. You will probably think that this is trivial and uninteresting but I am quite proud to have been able to figure it out and do it alone although I ignored everything about these things a couple of weeks ago

EDIT:
Despite all that I have suddenly a problem with FT3.
When I follow your first tutorial on this thread (creating a B&W mask), loading a selection in FT3, then doing raise - nothing happens. My selection is just blinking, everything has a light purple colr but no land appears. Tried to invert, nothing. I can't apparently act with no tool on the selection? Have you any idea what that may be ? Or should I write to Profantasy support ?