Editing a world map

[Deadshade]

As I have bought FT3 and Terraformer and can save .png, it is time to show you what I created. If I can download the file here that is



EDIT :
There is of course still plenty of things to edit - the continents are basically plateaus with uniform 100 height and the coastline is not sufficiently fractal.
As I can export now the FT3 map into Wilbur I tried (and failed) to do something there.
- Everytime I click help in Wilbur, I get a message that it can't locate the help file. Is it because it doesn't exist or because it can't find the place where it exists ?
- The biggest problem I have is that I can't fractalise the coastline so that it stays in that handdrawn form with steps and straight lines. When I try the method you proposed (select coastline, add noise) it just transforms it in a succession of very small round islands but the shape at bigger scales stays unchanged. I think that it is because selecting the border of 5 pixels, makes only fractal structures of 5 pixels or less so that when one zooms out, nothing changes.
Is there a way to (re)draw a 2D fractal line in FT3 or Wilbur ? Either to fractally modify the whole coastline or to draw a fractal between 2 points on the map plane ? Or some third method I can't think of to modify a border so that it looks fractal even at scales of 1000 miles ?

[waldronate]

There are lots of ways to go about roughening a coastline. Using variations on the CSU Johnsondale techniques, Wilbur can get you results similar to the attachment in half an hour or so. It's not perfect, and the plains and mountains are a bit out of scale, but the coastlines are roughed up a bit (if suffering from some erosion).
It turns out that I have reduced the image size a bit (3000 wide instead of 4000), but that should be fairly minor.

[Deadshade]

God the left one is nice ! - what is this CSU Johnosondale technique you mention ? This is really the direction I would like to go. Well the mountains are not on the "right" places but you couldn't know where they are supposed to be.
Of course I'd have to edit with some gradients in GIMP (to get the polar colors right) but I have an idea how to do this part.

I already found out how to do something similar to the right one but it is not really fractal on higher scales :
So I am not very satisfied with it because it doesn't look very realistic.

Thanks to have shown me what can be done ! Now the how
I am rather awed because what I only partly approached in 2 weeks after starting to learn the cartography software, you did it in a half an hour ....

[waldronate]

One way to get rid of that speckled noise all over your mask is to try a median filter or a minimum followed by a maximum filter.

A few posts up from here in this thread is the example tutorial that I did using Wilbur. It's the same basic process as the CSU Johnsondale tutorial and that I used here. Even though the search function at the guild is broken at the moment, a Google search on "CSU Johnsondale" turns up the link. The only things you might want to try on the tutorial in this thread is to drop the land level a little bit to remove the smoothed edges of the original selection.

The attachment is what Wilbur's V2 shader can do and when things are upped to 8000x4000 resolution (I cropped it for upload). The V2 shader has the mistake that it's not symmetric about the equator, but I was trying out some procedural shading code. Note that the "half hour" doesn't include the hundreds of hours of coding and determining the workflow - I have been at this for a very long time. If I ever get actions working properly in Wilbur, then the process could take as little as a few minutes...

The image on the right shows a little Photoshop magic to blend two v2 images (one flipped vertically with lighting reversed), one ambient occlusion, and one direct overhead light. A shader tree would be a nice feature to me to implement in Wilbur, too...

[Deadshade]

Oh yes, this is great. Thanks ! I just found the Johnsondale tutorial and start to experiment with it. It seems to be exactly what I wanted to do.
I'll see about the shaders later (don't overload the grey cells ) - note that I have not the slightest idea what a shader tree is.
I didn't understand what you meant with the "median" filter but it's easy to get rid of the white pixels by just selecting and painting.

I would like to come back to my obsession - a fractal self similar coast. Your second examples (especially the right one) are much better than the B&W above. How did you go from the latter to the former ?
I have thought a bit farther about the necessity of self similarity on the mathematical level and made this post http://www.cartographersguild.com/tu...stlines-6.html.
Do you know of a method how to do this ?

I also saw that you got a right and realistic coloring.
Do you think that there is added value to use the editable gradients in GIMP or that it is too much time to invest in Learning for little improvement in the coloring ?

Last question that I already asked a few posts earlier. When I click on help in a pop up menu in Wilbur, it says "Unable to locate the help topic id ... etc". Is it because it doesn't exist or because I botched something when installing Wilbur ?
This is a great handicap for me indeed because for 90% of the options I have no idea what it does and what parameter values I should/could select.

[waldronate]

A median filter looks at all of the pixels in an area around a given pixel and then selects the one that's most likely to occur (the implementation is to collect all of the pixels in an area around the current one into a 1D array, sort those pixels, and then use the one in the middle of the array as the result). For shot noise (single pixels that vary wildy from the local average) or excess detail, a median filter will mostly remove the noise without too much loss of detail and usually without smoothing the edges as a simple blur would.

The notion of a shader is pretty simple: it takes input information such as color lists and height values, and uses those inputs to produce an output color. A shader might color the output based on the difference in angle between a light source and the local surface direction (the result looks like a gray lighted image), it might pick a color from a list based on altitude, it might search the local area to see how much sky is visible, or even combine those (e.g. the basic shader in Wilbur uses a pair of color lists combined with lighting information). A shader tree allows for combining the results of multiple shaders into a final results (e.g. a color list shader can be multiplied with a lighting shader to get something similar to the default Wilbur shader). A shader tree implementation typically allows the users to select the types of shaders and the operators used to combine them, including allowing different shaders for different inputs (e.g. one shader for land and one for sea). It's a flexible solution, but takes a bit of effort to implement.

As far as the description of the sort of coastline processing that you're interested in, I would recommend trying a vector-based solution. Trace along the coastline and then increase the fractalization for the vector paths. I know this is doable in something like CC3 if you use fractal polygons to trace your continents. CC3 is, by design, a fixed-scale implementation and has some limits on the total number of nodes in a polygon (the limit is conveniently similar to the actual Windows limitations). Other than that, though, it would be pretty straightforward, I think. Then you can output the coastline to an image if you'd like further processing. There is the manual tracing step that's in there, though. I think that Illustrator or Inkscape have similar functionality and may also be able to do the auto-trace operation.

I haven't used the gradient in The GIMP, sorry, so I can't really offer much in the way of assistance on that front, I'm afraid. The V2 shader in Wilbur allows a color list (a gradient with an awful user interface) to be used for the base equator-to-pole coloring. On my far-future list of toys to incorporate in Wilbur and FT is the notion of latitude values. The user would get a guide line to set at a particular latitude and then can attach a value to that latitude based on the current context. This feature would be useful, for example, to set a basic equator-to-pole gradient for rainfall or temperature. It would also be useful for something like the V2 shader, where the base color gradient could be directly specified on the map via guide lines.

I haven't gotten around to making a context-based help file for Wilbur, sorry. I usually hint that someone could do that by taking the (admittedly elderly) documentation on the Wilbur web site, chopping it into web pages, and then naming the pages according to the context-sensitive names shown in the error message. For example, hitting F1 on the main page brings up a notice that the help topic ID 131200 is missing and that it even tried to look for an HTML file named "[installdir]htmlhelp\W0020080.html" (other dialogs have different IDs, which allows for a hyperlinked system to be developed). I am trying to avoid mucking about with the documentation file because I'll update it and I have far too many things that I'm behind on to put any time into that activity. There are whole features in Wilbur that I haven't used in many years and there are features on dialogs that I also haven't used in a long time. I don't necessarily remember that they all do, either (I can look at the source and determine what, but why I would want to do that may still remain a mystery).

When I did the parts above, I started from the blue and green JPEG that you posted and then I converted it into a black and white mask GIF that I loaded into Wilbur as a selection and I drew a mountain mask to use as a second selection. (note that Wilbur will rescale your input selection images to map the whole world space, so it's easy to shrink things for rough processing and expand for later processing). I set the whole world to -50, then loaded the coast selection, set things to 100, then loaded the mountain selection and set things to 500. Now I had a basic two-plateau world on which to apply iterations of resize/noise/fill basins/incise flow/precipiton activities.

[Deadshade]

As far as the description of the sort of coastline processing that you're interested in, I would recommend trying a vector-based solution. Trace along the coastline and then increase the fractalization for the vector paths. I know this is doable in something like CC3 if you use fractal polygons to trace your continents. CC3 is, by design, a fixed-scale implementation and has some limits on the total number of nodes in a polygon (the limit is conveniently similar to the actual Windows limitations). Other than that, though, it would be pretty straightforward, I think. Then you can output the coastline to an image if you'd like further processing. There is the manual tracing step that's in there, though. I think that Illustrator or Inkscape have similar functionality and may also be able to do the auto-trace operation.

I found quite confusing the term "vector based solutions". As any geometric shape (curve, surface, volume) is a subset of the R^3 vector space, there can't be any solution that would not be "vector based". Even a random function acting directly on individual pixels must use their coordinates so is vectorial too. Anyway I have not CC3, Illustrator is not free so I installed Inkscape. It looks pretty much like Gimp but more geometry oriented where I found GIMP more color oriented. Inkscape has no fractal polygone function but I found a function that is supposed "to deform a path into a fractal". Unfortunately it returns the message "The fantastic lxml wrapper is required. Download ...." so I stopped at this point. I bought FT3 but I don't think I would buy CC3 only to get a fractal.
So the problem of a fractal coast still stays open.

I haven't used the gradient in The GIMP, sorry, so I can't really offer much in the way of assistance on that front, I'm afraid. The V2 shader in Wilbur allows a color list (a gradient with an awful user interface) to be used for the base equator-to-pole coloring. On my far-future list of toys to incorporate in Wilbur and FT is the notion of latitude values. The user would get a guide line to set at a particular latitude and then can attach a value to that latitude based on the current context. This feature would be useful, for example, to set a basic equator-to-pole gradient for rainfall or temperature. It would also be useful for something like the V2 shader, where the base color gradient could be directly specified on the map via guide lines.

I only mentionned GIMP because it has a relatively easy and intuitive way to apply and edit color gradients. So as I am facing a mountain of things to learn the very existence of which I have been totally ignoring only 2 weeks ago, I tend to minimize the amount of Tools to use. Intuitively what you describe would be perfect for realistic planet coloration - indeed as colors depend on climate and climate depends only on latitude and humidity (of course we consider only rotating planets), then getting a color considering these 2 parameters (latitude,humidity) and creating a smooth gradient would do exactly what is needed. What I take from your message is that I have to experiment with the shaders to find out how it works (there are tons of them in the Wilbur menu). What about the alternative of using the climate types in FT3 and tying them somehow to a color selection ?

I haven't gotten around to making a context-based help file for Wilbur, sorry. There are whole features in Wilbur that I haven't used in many years and there are features on dialogs that I also haven't used in a long time.

I understand and this is precisely the problem. If you don't know yourself what a feature does, then how high is the probability that somebody else does ? Ultimately it boils down to the target followed.
If it amuses you to write a program because it is a hobby and you have fun with it, then to hell tutorials and help functions. The tool is free after all so who would dare to tell you that it is not user friendly ?
However in these 2 weeks I installed and (partly) tested some 10 softwares that are more or less usable for realistic map making on planetary/continental scales - notable exceptions CC3 that is not realistic at all and Photoshop which is too expensive. And I slowly converge in spiral to a duo (GIMP,Wilbur). I am not sure how FT3 fits in but it looks superficially like a more user friendly Wilbur.

When I did the parts above, I started from the blue and green JPEG that you posted and then I converted it into a black and white mask GIF that I loaded into Wilbur as a selection and I drew a mountain mask to use as a second selection. (note that Wilbur will rescale your input selection images to map the whole world space, so it's easy to shrink things for rough processing and expand for later processing). I set the whole world to -50, then loaded the coast selection, set things to 100, then loaded the mountain selection and set things to 500. Now I had a basic two-plateau world on which to apply iterations of resize/noise/fill basins/incise flow/precipiton activities.

I don't understand at all what you mean by "rescaling, shrinking and expanding". But apart of this, I applied exactly a similar method (masks, noise, precipiton) the only difference is that I used 3 levels 100, 500, 1000. I understand incising and basins filling but I didn't use those either.
It gave me my first semi realistic map ever
I say semi realistic because the world was all mountains and very deep canyons. Now what I'd need would be an advice how I can create low Rolling hills and plains. There is a big part of the SW continent which has a large flat sand desert (like West Sahara) so it won't do that the method creates canyons and rugged high ground all over the place. The same goes for jungle on flat ground (like Amazonia).
I thought of creating a "plain and hill" mask which I would use when the mountains get the right shape and then continue eroding only this "plain and hills" selection hoping that the canyons would disappear. It is probably stupid for reasons I don't know yet but as I attack the shaders now I didn't try it.
Can you advice with that ?

May I add that I admire your patience with a struggling newbie ?


EDIT : I didn't find any "median", "maximum" etc filters in Wilbur. There are tons of options Under the Filter tab but none that is saying medium, maximum etc.

EDIT 2 : I got an unexpected problem. I saved (mdr format) the map obtained with the 2 masks which was 0, 100,500, 1000. I wanted to open it right now to edit some and it became 65, 16760, 32800, 65535. What is going on ?

EDIT 3 : Giving up on the shaders. It's simply massively overwhelming for a trial and error method. Spent 3 hours on it and didn't even find how to change a single color in the set up let alone produce a primitive gradient. Is there at least a rudimentary tutorial about how to use the shader set up ?

[waldronate]

Aha! I think that I may have determined one source of confusion: terminology. Drawing software is broadly described in two categories: raster and vector. Raster systems work with uniformly point-sampled 2D functions (typically a red/green/blue/opacity color tuple or a single scalar that represents a quantity such as height or opacity). Programs like Photoshop, The GIMP, Wilbur, and FT3 fall into this category. In contrast, a vector program deal with geometric representations such as lines, polygons, circles, and so on that have attributes such as color, stroke styles, fill styles, and so on. Programs like Illustrator, Inkscape, CC3, and so on fall into this category.

Most display devices are raster devices (blocks of pixels) and vector programs typically need to rasterize their work for display. After the basic rasterization process that samples the geometric entities down to a raster of pixels, the tricks of raster programs can be applied. There are crossover programs that allow you to work with geometric (vector) representations and apply the raster zoo dynamically during operations. Programs like Xara xtreme and (to some extent) CC3 allow for this sort of processing. It's possible to do some level of conversion from the point-sampled representation back to the geometric one, but it usually results in significant loss of fidelity. Geometric transforms such as rotation, translation, perspective transformations, and so on are much less computationally expensive to perform on geometric primitives rather than on the point-sampled representations. This is largely because the raster is missing all of the information between pixels (samples), where the geometric primitives have that information available.

FT3 is an odd beast in that its primary intended function is to point-sample a 3D fractal function representing altitude and generate a vector result that can be input to CC3. You are quite right that FT3 shares many features with Wilbur; FT3 is a work based on the Wilbur code base that ProFantasy contracted with to develop. However, Wilbur uses a fixed-size raster that can be resampled to different sizes. FT3 uses a fixed-size raster for editing, but that raster is interpolated and combined with the fractal function as needed to generate a raster of the desired resolution. For output to CC3, it generates a raster and then scans that raster to generate a vector geometry representation of altitude contours, rainfall contours, and so on. CC3 can take that vector input and allow for further processing such as fractalization of vectors, adding text, adding symbols, and so on. CC3 includes a set of raster operations in its effects for basic operations such as blur, glow, and a whole host of others (the upcoming CC3+ enhances the speed of these operation as well as adding some new ones).

What you have with the masks above is a set of rasters, and they are easily amenable to operations with raster tools like Wilbur.

One algorithm for generating a raster fractal is scale-and-add. Start with a relatively small raster like 128x128 where every sample is 0. Add Gaussian noise of a particular amplitude (say, 12. Now resample that raster to twice its original resolution (256x256) and add noise of half the original amplitude (64). Keep doing this process over and over (for 512x512 size and 32 amplitude noise, then 1024x1024 and 16 amplitude noise) until you get an image that is your desired size. The output of this process is approximately a fractional Brownian motion (fBm) fractal. The character of the underlying interpolation function will determine to some extent the character of the final result, especially if the scaling isn't a simple power of 2. A nice and smooth function like a bicubic will typically give smoother hills than a linear or nearest-neighbor interpolation.

The process that I described in a previous post is quite similar to the above process, except that it starts from your desired shapes and, in addition to adding noise, enforces real-world constraints like river connectivity at each stage. The processing uses non-local operators like basin filling and precipiton erosion to move information laterally across the raster.

One way to enforce local rules is, as you guess, through the use of selection masks. The basic idea is that a selection is a grayscale raster whose intensity (0 through 255) controls how much of an operation is applied (0=no new effect, 255=full new effect). Selecting the area for your flatter area, feathering that selection (which smooths out the edges of the selection to avoid sharp transitions), and applying an operator like exponent greater than 1 or scaling less than 1 will typically get something like your desired effect. A blur might also work. The implementation of erode is probably not going to get what you're after to smooth out a smallish area.

I never got around to implementing a median filter for Wilbur (it's on my list of things to do). The Morphological Dilate filter is equivalent to a maximum operation on a 3x3 pixel block, and the Morphological Erode filter is equivalent to a minimum operation on a 3x3 pixel block. The radius on the morphological tools is just the number of times that the erode or dilate operator is applied. A per-pixel min or max operator against a constant can be found on Filter>>Fill>>Set Value. A per-pixel negated min and max operator against a pair of constants can be found on Filter>>Height Clip.

The color gradient tools in Wilbur are hidden behind buttons that are labeled "color list". The gradients in Wilbur are pretty difficult to work with, I will admit. They have been on my list of things to deal with, but there are a whole lot of other, more interesting toys for me to work on. The help file parts fall into the same category. My next huge project in the works for Wilbur is to get actions working (basically a scriptable macro system). It's got a long way to go and I'm chronically short on free time as well as sleep, but the work really simplified the overall code base.

It's possible that you saved a PNG surface with a .mdr extension. Wilbur has another little gotcha in that you have to select the file type from the drop list, not just enter it as an extension on the file name (can't use the extension trick because there are multiple file types that have the same extension, such as PNG surface and PNG texture). If you rename the .mdr file to .png and then load it into Wilbur, you should be able to use Filter>>Fill>>Set Value with a value of 0.01526 (1000/65525) and operation of Multiply to scale it back to your original range. Then use File>>Save As and make sure to select the "mdr surface" type from the file dropdown.

[Deadshade]

One algorithm for generating a raster fractal is scale-and-add. Start with a relatively small raster like 128x128 where every sample is 0. Add Gaussian noise of a particular amplitude (say, 12. Now resample that raster to twice its original resolution (256x256) and add noise of half the original amplitude (64). Keep doing this process over and over (for 512x512 size and 32 amplitude noise, then 1024x1024 and 16 amplitude noise) until you get an image that is your desired size. The output of this process is approximately a fractional Brownian motion (fBm) fractal. The character of the underlying interpolation function will determine to some extent the character of the final result, especially if the scaling isn't a simple power of 2. A nice and smooth function like a bicubic will typically give smoother hills than a linear or nearest-neighbor interpolation.

The process that I described in a previous post is quite similar to the above process, except that it starts from your desired shapes and, in addition to adding noise, enforces real-world constraints like river connectivity at each stage. The processing uses non-local operators like basin filling and precipiton erosion to move information laterally across the raster.

This is cute ! It reminded me of the iteration producing the Cantor set. When you described the process many posts above, I was still in a stage where I was hardly able to put a 2x1 map in FT so that it was Chinese to me.
I redid it now and was pleasantly surprised that I not only could follow but that I was more or less understanding what you were doing and why. Of course I have no idea why you gave the values you did to the flow exponent or why you used the pre blur at one stage but I can't understand everything at the first trial.
Thinking about the Cantor set helped me to understand better.

Here is what it gave :

There were a few things that I did not do according the tutorial :
- I didn't load the selection 0 all the time. I selected the continents by height instead.
- I omitted the routine with inverting to the sea because I saw no "bad" pixels in the sea.
- I pressed the continents Under water because some parts of the map had merged in the end phase. There are still a few places which stayed merged.
The problems to solve
- the northern continent will be covered by ice and snow so that it shouldn't be so rugged (only the mountains should because this is the highest spot on the planet)
- the transition from the smooth coast to the hills is too brutal.
- the hills are too "bumpy" - I guess too much noise at some stage
- the coast is still not fractal enough
- there is some general problem with the 2 big lakes (SW and N). They should have an outlet towards the sea (otherwise they would overflow anyway) and there should not be high ground between them and the sea - this shows especially for the SW one. I thought perhaps to force an outlet by lowering the ground between a point on the lake and the sea - an artificial canyon - but was not sure if it would not mess with the incise function. I also expected that there would be big rivers going to the lakes but there are none.
- and of course I'll have to tackle the coloring part ... btw is there even a very basic totorial/documentation explaining how to work with the Wilbur shaders ?
I find that this is the most difficult part. The whole "Texture" tab is pretty much impenetrable.

One way to enforce local rules is, as you guess, through the use of selection masks. The basic idea is that a selection is a grayscale raster whose intensity (0 through 255) controls how much of an operation is applied (0=no new effect, 255=full new effect). Selecting the area for your flatter area, feathering that selection (which smooths out the edges of the selection to avoid sharp transitions), and applying an operator like exponent greater than 1 or scaling less than 1 will typically get something like your desired effect. A blur might also work. The implementation of erode is probably not going to get what you're after to smooth out a smallish area.

I think I understand . This is typically what I'll need for the northern ice continent.

It's possible that you saved a PNG surface with a .mdr extension. Wilbur has another little gotcha in that you have to select the file type from the drop list, not just enter it as an extension on the file name (can't use the extension trick because there are multiple file types that have the same extension, such as PNG surface and PNG texture). If you rename the .mdr file to .png and then load it into Wilbur, you should be able to use Filter>>Fill>>Set Value with a value of 0.01526 (1000/65525) and operation of Multiply to scale it back to your original range. Then use File>>Save As and make sure to select the "mdr surface" type from the file dropdown.

Yes this might be the case. But when I save PNG surface and reload the same, the heights are wildly different like in the example I gave. I checked that when I save a mdr and load the same mdr, the heights are conserved.

EDIT : What is the maximum resolution of Wilbur ? I wanted to try something big and when I selected 10 000 x 5 000, I get a gray cadre around the image so that the image itself is something like 9000x4500. Or am I doing something wrong ?

EDIT 2 Your tutorial always ends with a resolution 4096x2048 whatever the initial resolution of a map is . At what step and how I recover the initial resolution (say 8000x4000) ?

[waldronate]

A PNG Surface file is stored as a 16-bit grayscale value. The original input data range (from minimum altitude to maximum altitude) is scaled to the range of 0 to 65535 (the full range for a 16-bit integer value). An MDR file, on the other hand, saves the internal floating-point raster at the full floating-point precision). All of those options shown for MDR don't really do anything and the defaults are all that you should use.

One way to get more detail in your coastline is to load your original selection, invert it to select the sea, and then push all of the sea below 0 (say, -1 or so). Then load up the original selection again and run a few rounds of precipiton erosion with the selection in place. This effect will shrink your landmasses a little because the edges are flowing off of the selection out to sea, but it will bump up the detail in the coastline. One unfortunate artifact of precipiton erosion is the axis-aligned square shape of its deltas (where sediment is moved out to sea). It looks like you kept the sea at 0 during the operation and didn't keep the initial selection loaded because you have little deltas covering your coastlines, which is what you get without some way to prevent all of the mass moving downhill from piling up.


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.

The Wilbur shader is a pretty ugly beast. It's actually 5 shaders and a page to direct how those colors are blended together. Intensity controls the lighting value. Altitude is the hypsometric shading (color by altitude) with a separate set of things on the page for land and sea. Latitude applies a color list based on the top and bottom surface values. Slope applies a color list based the local elevation (vertical) angle. Facing applies a color list based on the local azimuth (horizontal) angle. Blending specifies how the colors specified on the various pages are blended (facing, latitude, altitude, slope, and texture are a linear RGB blend; the Intensity result is multiplied by that blended color to get the final result). Texture on the Blending page is just a picture that goes into the blending color. If a coefficient on the Blending page is 0, then there will be no contribution from that color. 1 would be full contribution and the coefficients should all sum to 0 (the system will normalize the coefficients to 1 for your convenience). If you want to turn off the shading, you can set Ambient on the Intensity page to 1 or select "Height Code" from the General page.

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.

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.

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.