Hey everyone, first post, so this is my intro as well.
Anyway, excuse me if this has been discussed but i’ll just post it anyway. Now that 1.0 is out, what would be awesome to see in the next big release would be the ability to set a Sea level and include a host of water dynamics filters that affect the coastal areas. There are lots of erosion possibilities. It would certainly be awesome to have beaches and coastal cliffs created realistically. With a mask input you could determine where soft unstable earth is to create cliffs or whatever. And obviosly, erosion below Sea level would be drastically different. Pretty please?
Well, actually you “kind of” can do that already… You just need to think of it in a more open way…
I said “kind of”, because the effect is not as realistic as we want, but the basics of what one needs are there already.
Lets decompose your thoughts! “Water level”… what is that?.. It’s a boundary height that limits two parts of terrain. Like you said, there is the part above water, and there is the part below water. So assuming you have both parts coming out of a same base heigtmap (so one actually is related to the other) you need to join them with a chooser! Use a height selector to pick up both parts: select everything below water level, pass that as input C into the chooser, and pass the two main submarine and normal heightfields to the other inputs.
As for the types of lands that you find above and below water… well, that is a different story, but you can always think that below water there is usally more deposition thatn above water…
This obviously doesn’t solve the problem of the cliffs, and water dynamics you mention… But I find it more flexible if there is a set of devices, or macros, that have an under water behavior, and another set that operates on dry land; and if you mix everything in the end …
Then after a Coastal Erosion device appears, it’s just a matter of using it…
The problem of doing a higly specific device that tries to solve a lot of problems in a single one is that a bit of the flexibility power is lost. I am advertizing this certain care about keeping flexibility because the days when you whish for a device and have it shortly after, may be getting closer to the present (i’m hinting at the PDK ;)).
This is something that I’ve considered before, and even have done some preliminary R&D into a beach erosion filter. It is a complicated interaction zone though, so it’s taking a lot of thought to produce something that produces worthy results.
So my answer is that “It’s being researched” But obviously I don’t know what the fruits of that research, if any, will be.
Personally, realistic coastlines is one of the thinks that I think is often missing from CG terrain, due to the way we make it (ie intersect the land with a plane of water and say ‘here it is’). A seawater erosion plugin would go a long way to adding that realism.
Hello Fil and Remnant. Thanks for the replies, i’m glad to hear and not in the least surprised that you’ve already given thought to this Remnant.
Fil, i have not used a chooser, i’ll mess around with that some more, thank you. What i’ve done so far is after getting a basic island shape, i used the curve and a masked terrace to flatten out some beached. Then with a height selecter i did a masked gaussian blur on the terrain below the sea level. In my case, i was making this island for a Battlefield 2 map and so i dont need to go into detail under water because you cant swim down there and it gets fogged out. I also painted a mask on steep coastal areas of the island and terraced those areas for some cliffs.
A macro would be nice indeed, perhaps just something as simple as creating beaches would be good enough. That took time to get right because it was a lot tougher to get beaches (after erosion) where steep terrain meets the supposed sea level, while not going overboard or whatever with the rest.
I still have so many devices to explore though, plus this is the first time i’ve used WM in a while so i’ve forgotten some tricks
I’ll have to post screenshots of the map eventually.
(first post too )
If sea is to be taken into account in world machine, another idea that could be just as important is to make the erosion actually generate lakes across the landscape. At the moment when the fluvial erosion reaches the bottom of a pit in the ground it stops, or gets cut off. maybe world machine could have a system where it calculates a rising pool that reaches the highest point it can and then overflows, continuing the erosion. And silt would pour into the lake and make a flatter area for the lake to be placed (in another program).
Thinking about it though , it would probably be more complicated than that because lakes only develop when the land will not absorb water, so you would have to somehow calculate the water table. Or maybe even the ability of the ground material to absorb water (the difference between rock and soil) if you see what i mean.
proper water systems in erosion is definitely something I’ve looked at. I’ve created a few prototypes of full erosion with lakes and streams, both from my own models and using some old code that Doc Mojo had once created. At the time I reached a dead end in terms of getting high quality output, but that doesn’t necessarily mean that that dead end would be reached now. Like every feature, it’s a matter of “are the benefits large enough to be worth the invested time”. I think so, but finding the time to invest can be difficult.
Matt Fairclough of was talking to me about this very subject sometime back about simulating Continental Scale Erosion from Glaciers and other Fluvial sources literally shaping the continent with complex drainage networks but he said it was very diffucult stuff to do but they are looking into it… When the first time the public got to see the first glimpse of Terragen 2 i just had to ask the diffucult questions… but hey The River Plugin for Mojoworld can do it so why not Terragen 2… i am sure someone will write a killer Plugin for that very thing once the PDK is realised…
But obviously I don’t know what the fruits of that research, if any, will be.