PDA

View Full Version : WIP - Batai



Master TMO
05-15-2010, 02:13 PM
I'm an almost complete newb at this world-building business, so I thought I'd post this map as I worked on it. If I do anything silly/stupid, folks can correct, and if I do happen to do something new and good, I can answer questions.

As a concept, there's nothing special about Batai the world, just an Earth-like. I'm working on it for a friend's MMORPG. He would probably be happy with this default map, but I'm going to use it as a learning experience. Here's the default map.

I'll add posts to this thread as I make progress.

arsheesh
05-15-2010, 02:16 PM
Hello Master TMo, and welcome to the guild. Hm, I'm not seeing a map?

Master TMO
05-15-2010, 02:20 PM
I attached the wrong one, so you must have looked while I was swapping maps out. ;)

Master TMO
05-15-2010, 02:27 PM
I exported it into Photoshop to make a contour map, so I could determine how the mountain ranges would affect the wind patterns, and where to place deserts and rainforests. I built a wind pattern map in Powerpoint since I'm no meteorologist and couldn't handle trying to figure all that out in my head. ;)

Changing the rainfall to fit is my next task. Right in the middle of the map is an area that goes below sea level and is completely surrounded by mountains. It's also just south of the 30 degree latitude, which is the lat where deserts tend to form. The mountains would block all the rainfall brought in from the ocean to the east, so it's going to be a Death Valley-like area only massively larger. The eastern side of that mountain range is going to be totally soaked with all the rain that should have been dropped in that valley.

tilt
05-15-2010, 03:14 PM
wow... I usually just put my finger in the air and go... hmmm... deserts would be cool ... here.. ;)
nice that you do all the hard work first :)

Steel General
05-15-2010, 03:53 PM
This is interesting so far, and I am looking forward to watching it develop.

Like tilt, I just more or less spin in a circle and plunk my finger down and say "Here's the desert" - all that other stuff is to much work for me.

jwbjerk
05-15-2010, 04:58 PM
Looks pretty good.

I don't know what method you are using to figure weather, but you might be interesting in this guide (http://www.cartographersguild.com/showthread.php?9937-Wind-and-Ocean-Currents&p=107675&viewfull=1#post107675)... it's worked pretty well for me.

Maybe you are more comfortable in powerpoint, but i did my figuring high/low pressure areas, etc. on transparent layers in photoshop. It's nice to have the info right there to turn on/off when you need to see it.

Master TMO
05-15-2010, 07:44 PM
Downloaded the Guide, thanks for the reference! After several web searches, I found this page that I used as my reference guide: http://www.ux1.eiu.edu/~cfjps/1400/circulation.html

And I would have loved to use PS to develop the high- and low-pressure areas, but I couldn't do the wind direction arrows quick and easy there, so I used Powerpoint, which *loves* making arrows. ;) Probably just a problem of me not knowing how to build them in PS, and I was too lazy to look up that on top of the climate research I was doing.

Oh, I was going to ask: Are there any good custom climate shader images for use in Fractal Terrains available? I haven't found any yet on a few quick casts, but I'll keep looking.

Climate before and after maps coming up soon...

Master TMO
05-15-2010, 08:20 PM
These are using the default Fractal Terrain climate colors, so they're not very 'realistic' colors. I've also added in the terrain shading to help you see what is where.

Steps I took:
* Selected the area that I'm unimaginatively calling the 'Great Desert', feathered the selection and globally set the rainfall across it to 2 in/year.
* Added Global Noise to the Rainfall, to add some randomization to it. The only visible effect of this is that the boundary between chapparal (brown) and desert (yellow) is fuzzier than before.
* Selected the 'little desert' to the northeast of the Great Desert and did a similar process, but it has more rainfall (base 5 in/year).

* Then I selected a band along the equator (about +/- 10 or 15 degrees), made it wavy around the terrain obstacles, heavily feathered it, then added about 20 inches of rainfall a year.
* Did the same thing again with a slightly smaller band and pumped up the temperature about 10-15 degrees.


Hrrmmm... Looking at it, I think I need to reduce the rainfall south of the Great Desert more. It's got a lot of rain there from being on the equator, but shouldn't get quite so much. It's not often you see a desert and a tropical rainforest right next to each other. ;)

My next step is probably to fill in the lakes where the elevation goes below sea level.

Master TMO
05-16-2010, 10:57 PM
Argh... Doing lakes and rivers is proving difficult. Especially that one in the far north - if I just run the Fill Basins command, that entire section up to the sea is completely filled in and flattened. So I'm trying to cut a channel through the mountains and it's very slow going. I didn't mind it flattening in the tropical zone (makes for a better Amazon river with river paths filling in the whole area) or the southern small lake. But that northern one is just too big to convincingly be a flat plain.

Also, I don't know if it's a function of the fractal generator I chose or not, but if I run the 'Fill Basins as Lakes' command, it just floods the entire place with lakes. ;) Basins all over the place. Same issue if I fill with land, flat plains everywhere. So, lot's of manual corrections.

Master TMO
05-17-2010, 01:12 AM
Well, after working on basins all day long, I think I found a good method for working on them. The problem has been finding the actual low point of the edge of the basin, so that lowering the altitude there actually affects things. Filling basins with land was not easy to spot those areas. Filling with water is better, and also easier to work with, since you can erase land while the water is still there and still reset the water later back to sea level.

Master TMO
05-19-2010, 12:23 AM
Found an even better method, yay! :)

Turn on Prescale Offset Editing in your World Settings dialog, run a selection around the area, feather it, then Global Raise the Prescale Offset a small amount. Then shrink the selection and do it again. Vary the amount you shrink the selection by and the amount you raise the Prescale Offset. It allows you to raise the elevation without losing the fractal map bumps. I'll post a map update tomorrow, I think.

waldronate
05-19-2010, 01:43 AM
http://www.ridgenet.net/~jslayton/CGTutorial/ has some useful techniques for working with FT if you haven't seen it.

Master TMO
05-19-2010, 10:57 AM
Yes, I have that tutorial, although I wasn't following it explicitly. I had actually forgotten that he did so much filling and smoothing in the middle. I may take a copy of the original Batai and just follow his steps one-by-one and see what results.

Master TMO
05-19-2010, 11:31 AM
Small technical problem - for whatever reason (probably because I'm on Windows 7, even though I've tried a different compatibility settings), the direct painting tools are not usable. I can't change the size or brush strengths. Prescale Offset painting goes straight to a 1 value rather than the 0.01 it should be. The Global Raising and lowering within a selection still works for me though.

waldronate
05-19-2010, 11:39 PM
It's a known bug in the last release. Contact ProFantasy tech support and they can point you toward a beta version without that problem.

Master TMO
05-21-2010, 12:43 PM
Alright, starting over from scratch again, with the Beta version installed. Still have a few technical issues we're working out with it, but once those are resolved we'll hopefully be good to go. ;)

Following the tutorial, I brushed up a couple of the smaller mountin chains, but nothing too much. The map I chose does a reasonable job of having mountains near the coasts, although the biggest ones are still in the center of the landmasses. Oh well. This is one of those areas where experience counts - the more of these you do, and the more topo maps you see, the better an idea of what is 'realistic' you have. As a newb to this, experience is still lacking.

I have filled in most of the below-sea-level lakes, and am just trying to find any small ones still out there before continuing. To help with that, I set it to high-contrast mode, with the seas colored black, and the land white. Then I just scan the white areas for black spots and those are my lakes that need to be fixed.

25002

Once those are all filled in, I will follow the steps in the base tutorial listed above with basins. Note that the below sea-level basin in the 'Great Desert' has been filled in too. I figured for my first one I should stick to the tutorial, and wait until I've got more practice before breaking from established tradition. ;) Besides, if I really want it, I can put it back in later manually.

Master TMO
05-23-2010, 08:31 PM
Random experiment time! :D Just playing around to see what would happen - Starting over with the base map, I did the following steps:

* Lowered the waterlevel to -1000 ft.
* Filled the basins
* Raised the waterlevel back to 0 ft.
* Smoothed the Prescale Offset layer by 1
* Smoothed the Offset layer by 1
* Selected bands around the Horse Latitudes (30 degrees above and below the Equator)
* Feathered the selection by 30 pixels
* Lowered the rainfall by 20 inches, and increased the temp by 10 degrees
* Selected a band around the Equator
* Feathered the selection by 30 pixels
* Increased the rainfall by 20 inches
* Incised the waterflow using the settings in the basic tutorial
* Found the rivers.
* Exported the map

The goal was just to see how the bands would affect the climates being generated. Ideally, I should tweak the bands to take terrain into account. Figuring out how to do that will be one of the next things I do.

Master TMO
05-23-2010, 08:45 PM
Heh. A few amusing things about this version of the map. First off, there is a fairly major river running through the desert from out of the mountains. In the south there is a desert on top of a mountain.

On the plus side, the climate is nicely varied and follows the terrain well.

alizarine
05-23-2010, 08:59 PM
I like watching the experimentation - and that you're going to all the trouble to work out the proper placement for everything ^.^ Lots of hard work, but it really does pay off as long as you're not like me (I skipped several steps and almost messed some geographic areas ...). Looking forward to more of this! My personal thought about the overall map is that it's kinda odd all the land masses connect to each other (just a personal thought, really). Personal thought aside, there, this is looking good. I'll have to keep an eye on this.

- Alizarine

Master TMO
05-23-2010, 09:18 PM
I've thought a little bit about how the tectonics plates are laid out on this map. It looks to me like a Pangaea-style super continent (sp?) in the process of splitting up. The southern landmass is swinging south and east away from the main continent with a counter-clockwise twist, opening up that sea. Eventually it will split off completely. I'm not sure what's happening with the northwestern landmass, since it's connected to the main one over the north pole. Maybe it's splitting off, maybe not. That peninsula right in the center of the map is probably splitting off too, only remaining connected by a narrow neck of land.

Hard to say for sure, though. I haven't done any geology since college, and that was only an entry-level course. :)

I should plot that out more though, as that will tell me where the volcanoes are. That southern mountain range is probably volcanic. Actually, that north-western sea has mountains almost all the way around it. Perhaps its in the process of closing up, which would make all those ranges volcanic also. That small range of mountains on the western edge of the peninsula is probably also mildly volcanic, as it doesn't appear to be too large yet.

rdanhenry
05-23-2010, 09:23 PM
A major river starting in the mountains and flowing through a desert? Hey, you've got your world's Nile.

Master TMO
05-23-2010, 09:29 PM
A major river starting in the mountains and flowing through a desert? Hey, you've got your world's Nile.

Good point. I've lived in the central US most of my life. The Missouri and Mississippi are the rivers I'm most familiar with. But both the Nile and the Colorado go through desert areas. Thanks for pointing that out.

alizarine
05-23-2010, 10:20 PM
Wow - tectonic plate shift? Man ... I tried thinking about that, but didn't spend much time on it for my own map. I did find a great e-book somewhere about tectonic shift and using it for created maps. It's called "A Magical Society: Guide to Mapping" from WWW.EXP.CITYMAX.COM - it does cost some money (like $15.00 I think?) but it's an absolutely fantastic resource. It outlines a lot of great info about axial tilt and seasons, tectonic plates, mountain creation and placement, and a whole lot of other stuff. There's some "fluff" (there's a story about some guy playing god and creating a world, which is where the creation info comes in) but the info presented is really helpful. Just if you want some reference. This is going really well, by the way tho'.

- Alizarine

Diamond
05-24-2010, 01:59 AM
That link doesn't seem to work, alizarine...

Steel General
05-24-2010, 09:34 AM
There appears to be several places it can be downloaded for free;

Here (http://e23.sjgames.com/item.html?id=XRPFREE2), here (http://www.yourgamesnow.com/index.php?main_page=product_info&products_id=5) & here (http://www.rpgnow.com/product_info.php?products_id=2444) are a few.

Master TMO
05-24-2010, 10:58 AM
Good stuff here. I've just scanned it so far, but it's already pointed out a few problems with my tectonic thinking above. Thanks for the links, all.

Master TMO
05-26-2010, 12:54 AM
Another random notation/suggestion: When using Fractal Terrains, and you've selected the map you're going to modify, export world views of Altitude, Rainfall and Temperature into jpegs and stack them into layers in Photoshop. It let's you see what you have to work with without having to wait for redraws every time you change views.

Also, I've heard how FT doesn't take wind or water currents into account when doing the climates, but the boundaries I'm seeing here, while hardly perfect, look to me like they're taking at least some of that into account. Although I'm now working on the most recent beta version (they had to update it to work on my system), so one of the updates along the line may have involved improving those algorithms.

I created the Altitude shade map this way because FT lets you load Selection maps from B/W bmp files. If necessary, I can select a chunk of the map and save it to a BMP and use it for a feathered selection mask.

The attached maps are: Altitude, Rainfall, Temperature.

PS: I apologize for not checking first; should I be uploading smaller versions of these files? Loading 3 of them up at once made me realize I could be using more space than I should.

Edit: Updated the attachments to the new version of the world.

waldronate
05-26-2010, 06:20 AM
FT doesn't take into account heat or moisture transfer due to air or water movement. It uses a static model based on axial tilt, latitude, altitude, and a random field in its temperature and rainfall computations.

The design goal for terrain, temperature, rainfall, and climate in FT was "here's something to start from and some tools to use; it's up to you to generate a plausible result."

Master TMO
06-06-2010, 02:47 AM
Spent the day adjusting rainfall on Batai. I may be too detail oriented, working the base data to produce the climate I want in an area, rather than just painting in the climate directly. As I get practice, I'll get better at manipulating things right the first time, instead of painting, undoing and painting some more. Next step would be to modify temperatures.

Here is the adjusted rainfall and climate maps. To point out what I've done, the desert is the most obvious bit, but I've also increased the rainfall both NW and E of it, on the coastlines, where the prevailing winds would be bringing in water from the ocean and dumping it.

Gidde
06-06-2010, 02:51 AM
Keep in mind as you build your world that with deserts at the equator that means your planet has next to 0 axial tilt, which means you won't have much seasonal climate change at all. If you have more of a tilt, you have more seasonal climates, but your arid area splits and moves to the degree of tilt (hence Earth with an axial tilt of around 23 degrees has arid belts at the Tropics, which are 23 degrees north and south of the equator).

Master TMO
06-06-2010, 02:54 AM
The desert has it's origins at the Horse latitudes, with hot dry air going both SW and NE from the 30 degree line and tall mountain ranges are completely blocking rain from getting there to relieve it. That's why the coastlines got more rain dumped on them - the rain that *should* have gone to the desert is getting dropped early or shunted around. At least, that was my thinking.

Gidde
06-06-2010, 04:04 AM
Ah cool, just wanted to make sure you had considered it. Zoomed out it looked like it was centered on the equator. Looking good, regardless :)

jwbjerk
06-06-2010, 03:10 PM
Spent the day adjusting rainfall on Batai. I may be too detail oriented, working the base data to produce the climate I want in an area, rather than just painting in the climate directly.
That's the way i'm doing it. Weather you are being too picky probably depends on weather this level of detail is fun for you or frustrating.


Here is the adjusted rainfall and climate maps.
I can figure out some of this, but not all of the colors you used for climates are self-evident. A legend would be nice.

Is grey tundra/permafrost? If these mountains are supposed to be roughly like earth's, i think you've been too aggressive in bringing colder climates near the equator with the mountains. It's true that as you ascend a mountains climate changes in a similar way to moving toward a pole, but at least on earth, most of the square miles in "mountains" is much, much lower than the peaks. Yes, the top of the tropical Andes are much colder than at sea-level, but you don't see a big blob of "tundra" there looking at a global Koppen map.

For the rainfall, green is low and red is high?



Keep in mind as you build your world that with deserts at the equator that means your planet has next to 0 axial tilt, which means you won't have much seasonal climate change at all. If you have more of a tilt, you have more seasonal climates, but your arid area splits and moves to the degree of tilt (hence Earth with an axial tilt of around 23 degrees has arid belts at the Tropics, which are 23 degrees north and south of the equator).

I'm not claiming that latitude has nothing to do with desert formation, but it's certainly not the only factor. If you look at earth you can find deserts on the equator (Ethiopia), and well-watered lands right on the tropics of cancer and capricorn, (China, Florida, North-eastern Argentina).

An area that mountains prevent from getting moisture is going to be a desert no matter what it's latitude.

altasilvapuer
06-06-2010, 07:55 PM
TMO, how is that you are still a grey pip? BONK! ;)

I love your attention to detail here, and I'm eager to see where this goes from here. You have me getting very antsy to get back into my map, as soon as I can get the scratch disks large enough that Photoshop can actually OPEN my map. Fail, ASP; fail.

-asp

Master TMO
06-07-2010, 02:41 PM
@jwbjerk - Sorry for the climate colors - they're the default colors used by Fractal Terrains, and I believe they were picked more so that different terrains would be distinctly obvious. The final version will use a more intuitive color scheme. And I'll look at the temperature of the central mountains - I hadn't actually adjusted those yet. That's the way FT set it. Adjusting inland temps shall be my next task, as I think I've finished the coastlines now. I might not have thought of it if you hadn't pointed it out to me though. Thanks. I'll have to look at temps for Central Asia, see what they range. I believe that's the closest real-world equivalent to that area, or maybe the northern edge of the Himalayas. I've added a legend to the climate map post. I was going to put it here, but navigating between two different posts on two different pages probably would have been rather difficult. ;) The legend is at a smaller scale and didn't insert very well into the actual map, so I just left it separate.

@Altas - Thanks much for the Bonk and the encouragement! :)

Master TMO
06-07-2010, 03:47 PM
Oh, and yes. Low rainfall is green and high is red. Again, default colors from FT.

jwbjerk
06-07-2010, 04:18 PM
@jwbjerk - Sorry for the climate colors - they're the default colors used by Fractal Terrains, and I believe they were picked more so that different terrains would be distinctly obvious. The final version will use a more intuitive color scheme.

I've added a legend to the climate map post. I was going to put it here, but navigating between two different posts on two different pages probably would have been rather difficult. ;) The legend is at a smaller scale and didn't insert very well into the actual map, so I just left it separate.
Thanks, i haven't used FT. Perhaps the colors are all second-nature to you FT users. Though i'm a little suspicious of what FT seems to produce. I don't know how it can tell the difference from climactic data between "tropical deciduous forest" "tropical evergreen forest". That's not a distinction i've seen on a real climate map before.



And I'll look at the temperature of the central mountains - I hadn't actually adjusted those yet. That's the way FT set it. Adjusting inland temps shall be my next task, as I think I've finished the coastlines now. I might not have thought of it if you hadn't pointed it out to me though. Thanks. I'll have to look at temps for Central Asia, see what they range. I believe that's the closest real-world equivalent to that area, or maybe the northern edge of the Himalayas.
Now that i know what the climates are, i can observe, overall your world seems to be much colder than Earth. Which is fine if that's your intention. Tundra in many places comes almost 2/3rds of the way from the poles, while on earth it barley extends 1/3rd of the way from the poles. This makes your boreal forest clumps between the tropics more understandable, though still not something i would expect in such large quantities.

There's three possibilities i can think of:
1) FT has exaggerated the elevation beyond earth norms, and thus the temperature makes sense
2) These mountains are much flatter than earth mountains (as if you filled in most of the space between peaks), and thus peak temperatures are widespread, and show up on a global map
3) FT doesn't calculate temperature to elevation well.


I look forward to seeing how you progress.

Master TMO
06-07-2010, 04:41 PM
Heh. Perhaps instant email notification of posts on this thread isn't such a good idea. ;) I keep interrupting work to come back here.

Elevation may be a part of it, as I have run the 'Fill Basins' command a few times. The assumption being that basins would eventually fill with either earth or water. And the southern half of the Great Desert actually went below sealevel. That was one BIG honkin' basin. ;) And it got filled in up to the level of the surrounding mountains.

The southern desert elevation seems to be in the 7000's, the central/north central section is in the 8-9k's, and the very north is 3k, but also quickly changes into other non-desert climate types. So most of the desert is way up there in elevation. Higher than Denver. However, one of my thoughts was to reintroduce some of that southern basin. It's not as if there will be much rain to fill it up with water or earth. But if I do, that will be one of the last tasks I do, as some steps will require me to fill basins in.

Right now I'm working on getting good rivers in there using FTs river command. It's a bit tricky making sure elevations are all good, as it is possible if you're not careful to have rivers flow uphill (FT runs the rivers until they hit the sea - if the river winds up in a basin without hitting the sea it will just climb right back out the other side and keep going).

Master TMO
06-08-2010, 12:39 PM
Hmm - given what I'm told 'bonk' is slang for over in Britain, I'm guessing any Brit members of this forum are laughing themselves silly over the line '@Altas - Thanks much for the Bonk'.

:D

Master TMO
06-11-2010, 11:18 PM
Just a check, to see if anyone watching this thread is knowledgable with the Incise Flow command in FTPro - I am wanting to use it to create wider riverbead channels across the many flat areas of my map, but it really only seems to show any visible changes to the sloped areas. Any suggestions on what settings to use to get a visible channel across the areas filled in with the Basin Fill command?

Thanks in advance.

waldronate
06-12-2010, 12:59 AM
http://www.ridgenet.net/~jslayton/CGTutorial/ shows that the Amount setting on the Incise Flow tool will make flat-bottomed channels but they will end up at 0 altitude; the Effect Blend setting will push that up by an amount intermediary between the valley bottom and original landscape when varied between 1 and 0. You have to interpret the visual results a little, but that's what seems to happen.

Master TMO
06-12-2010, 06:38 PM
I followed that tutorial's steps in the original map, and it carved out the mountains just fine. But what I was looking to do was put in riverbeds on the flat plain areas. It is carving gullies in the sloped areas, but I'm not seeing any effects at all in the flat plains. I've got those two flat areas right next to each other, the southern Great Desert and the Rainforest just south of it, and the mountains surrounding both are etched when I run it, but there is no apparent change to the plains themselves.

waldronate
06-12-2010, 07:32 PM
Do you have a picture? Incise Flow works with aggregate flow, meaning that if either the flow isn't connected or if there is insufficient flow then there won't be an effect. http://www.ridgenet.net/~jslayton/FunWithWilburVol1/index.html shows how the effect should work across relatively flat areas (the code for incise flow is the same in both Wilbur and FT).

If the flow isn't connected (which is quite likely if there hasn't been a basin fill operation recently) then there will be odd "pits" at the end of rivers.

Master TMO
06-13-2010, 01:54 AM
I will collect and post some pics for you. I'll make sure to do a Basin fill and smooth of 2 before running it, too.

Master TMO
06-13-2010, 02:02 AM
Oh, and while I'm here, I'll post the super-pretty output of the world. I don't think there's a lot left to do to it other than rivers. It uses a2area's climate image map. I have Waldronrate's climate image map also, but it didn't seem to mesh quite as well in the arctic regions. There were long strands of black rock running in patterns throughout it. Probably a quick randomization pass on temp/rainfall would have fixed it.

Master TMO
06-13-2010, 08:25 PM
Ok, here are the pictures. They are, in order:
1: Base image of the Rainforest. I picked it because it won't have the possible excuse of not enough rainfall.
2: After a Basin Fill
3: After a Global Smooth 2
4 & 5: After Incise Flow with settings 1, 0.667, 1, 1, with and without rivers. I picked those settings because I knew they'd be strong enough to have *some* effect.

You can see the carving along the sloped edges, but there is no visible change to the flatlands. Any ideas if I'm doing something wrong?

PS - Since the rivers typically run from NW to SE, I changed the sun to the NE to maximise the chances of highlighting a trench edge with shadow.

Thanks in advance for any help!

waldronate
06-14-2010, 06:41 AM
FT's rivers won't cut below sea level. It's a "feature" of the algorithm. It's also a feature of real rivers. If you drop the sea level a bit and do the river cuts, they will do quite a bit to your terrain in those areas.

RobA
06-14-2010, 04:29 PM
It's also a feature of real rivers.

ROFLMAO!! :D

-Rob A>

su_liam
06-14-2010, 08:46 PM
Well, in his defense, the sea level in FT is defined as a raster, so, you know, Death Valley and the low areas around the Dead Sea(apparently some variation of "dead" is a word meaning "way down there") are above sea level. Locally.

Master TMO
06-14-2010, 11:18 PM
I'll try lowering the sea level there and see what it does. But the elevation of the area ranges from 2k feet in the NW to 1000 in the SE. It's not really that close to sea level. And it also didn't affect the Great Desert, which is over a mile above sea level (but rainfall might have affected it - I don't know if Incise Flow uses potential rainfall or actual, I assume actual).

waldronate
06-15-2010, 02:25 AM
What's your editing resolution? If it's fairly low then the slope on the canyons might be shallow enough that you can't really see much.

Master TMO
06-15-2010, 09:58 PM
4096 - not very low. I'm wondering if maybe this beta version I'm working with is only grooving down original terrain, and not new? I'll run some experiments and see if it seems like it or not. Might just be a glitch in the new model. I'll let you know what I find. Thanks for the help so far!

waldronate
06-16-2010, 12:51 AM
Things should be visible at that detail. The process FT uses is: evaluate the surface, fill basins, find flow amounts, subtract flow amounts (with specified exponent and results clipped to 0) from the offset channel. The next screen refresh will recalculate the world with the changes.

It's possible that you're using a large blur and/or a large Amount, which can result in large areas of the higher flow (effectively farther along the rivers) being forced to 0 and so having no visible effect. What settings are you using? Try an amount of 1, flow exponent of 0.3, effect blend of 1 and blur of 0. If those settings don't result in visible channels in areas not at sea level then nothing will.

Wilbur is a good tool for learning the incise flow settings. It has faster setup and is a little bit easier to fiddle with.

Master TMO
06-16-2010, 12:54 AM
Flow settings the first time were 1, 0.667, 1, 1. I've also used 1, 0.667, 1, 0. (0 blur)

Master TMO
06-16-2010, 01:39 AM
Hrm. Maybe I'm just expecting it to do something it doesn't do. Attached is a screen cap from a section of a different map, in pristine untouched condition. I found a basin area to test pre-basin fill incise flow vs post. But on the pre-basin fill, it still didn't carve any canyons or riverbeds on the flat areas, only on the slopes. Is that the expected result?

waldronate
06-16-2010, 01:03 PM
Ah, now I see the problem. The basin fill operation is very important because it connects all of the land pixels to the sea pixels. The underlying fractal function is full of pits at all levels. When the flow calculation reaches a pit it stops. The incise flow operation is not a true fluvial simulation in that it does not accumulate water and move sediment. It only incises the existing flow into the underlying terrain.

A true fluvial simulation involves rainfall (adding water at each pixel), flow (moving water and sediment at each pixel), and evaporation (removing water at each pixel). This sort of operation is extremely compute intensive and suffers from a serious scale problem. Water flow simulations start to lose their versimilitude when pixels get much larger than a few meters. A basic fluvial simulation run at a worldwide level just looks wrong. The features it generates are inappropriate for ten kilometer-sized pixels.

Consider the precipiton computation in Wilbur. It is not a complete fluvial simulation, but people seem to like the way it looks. The way it works is that the system picks a random pixel and moves terrain height from that pixel to its lowest neighbor. This process repeats until the "flow" gets to a pit. A new random pixel is then selected and the operation starts again. It's totally unrealistic but works because your eye wants it to work. It's much the same way that basic fBm noise can look like a terrain at first glance.

If I ever find a good simulation that does plausible erosion on low-resolution maps then I'll probably include it in FT. I haven't found one, though. At this point FT can't generate endorheic basins like the American west or the Taklimakan basin that hosts the Gobi desert due to a lack of correct water flow computations in the terrain model as well as lack of air flow computations in the climate model.

Master TMO
06-16-2010, 02:04 PM
Ok, I understand. I have been asking it to do something it's not intended for. No worries! :) Knowing that, I'll mess about, see what I can do, including Wilbur. Sorry to spend so much forum bandwidth on the subject. Back to the map! *plays dramatic action movie music*

Master TMO
06-18-2010, 05:51 PM
I think I'm about ready to finish this one. I just need to clean it up and see how the client wants to see it. I'll try to do a GoogleEarth shot of it over the weekend and post it. There's nothing significantly changed from the Pretty version posted above.