Page 1 of 2

Middle Earth DEM Project

PostPosted: Mon Nov 21, 2005 10:47 pm
by JavaJones
Just wanted to let people know about a big project I'm involved with which is currently steering in the direction of leaning heavily on L3DT - the Middle Earth DEM Project (ME-DEM for short).

ME-DEM is an extremely ambitious, large-scale project aimed at nothing less than the complete modeling of the world of Middle Earth, created by J.R.R. Tolkein and most famously known from his Lord of the Rings trilogy. The "DEM" - Digital Elevation Model" part is the current focus, in which existing hand-drawn maps of the world of Middle Earth will be translated into actual heightfields. After that we'll be focusing on plant life and realistic texturing/geological mapping, using information from the various Middle Earth literature as well as extrapolation from Earth-similar areas, since Middle Earth turns out to be sort of a reimagining of an Earth history. Eventually we also plan to incorporate cave maps and 3D models of all Middle Earth towns, castles, monuments, etc.

As you can see it is a gargantuan project, spanning at least several years, if not decades. It is as much for the challenge as the end product for most involved at present. The seriously active members are few at the moment, but we always welcome new interested parties, whether you are content just to observe, or actually wish to contribute in some way. There are all kinds of areas for people to give input and assistance! If you have any interest at all in the idea, even if you just want to use the end result down the line for projects of your own, I encourage you to come by and check it out.

The Middle Earth DEM Project

- Oshyan

PostPosted: Tue Nov 22, 2005 10:29 pm
by Hinklemister
This sounds like a great project that would be really cool to see. Unfortunatly I don't have the time or resources to help any more than just to cheer on the sidelines. Good luck with this project.

PostPosted: Tue Nov 22, 2005 10:43 pm
by monks
Bah- It'll never fly !!!!! :lol:

bigup for ME-DEM. Coolest project ever ! 8)

monks
Middle Earth drone no. 234235

New gallery images uploaded

PostPosted: Mon Apr 17, 2006 2:20 pm
by monks
Hi all- I've uploaded a few images to the gallery taken from the ME-DEM article which is here:

http://me-dem.ashundar.com/content/view/46/33/


I used Aaron's Terragen Skybox tutorial - very nice!

I've got a skybox going spare, if any one wants it - DeathTwister?

- looking forward to the playing with the new L3DT goodies now :D

monks

PostPosted: Wed Apr 19, 2006 8:36 am
by Aaron
Hi Monks,

Well done on the guide, and I love the screenshots.

Sheesh, I'm glad L3DT isn't a heightfield editor (stricto sensu), as otherwise I'd feel obliged to try to match Wilbur's tesselation tool. It looks very slick.

Cheers,
Aaron.

PostPosted: Wed Apr 19, 2006 10:06 pm
by monks
Thanks Aaron:

hehe- obviously you've not tried it then -it still has a few improvements to be made: ;) I wouldn't call it slick. Though given time, it will be.

You know, I was thinking, you've got a reasonable handle on this physics...(slightly). you are very well placed to do this applied math and simulation. Do you have any plans in that regard: in terms of simulation of glaciers for instance. Do you think it could be done? Tectonics: there's another one: what do you think? If a dedicated physics processor became mainstream- that would be interesting wouldn't it?

monks

PostPosted: Wed Apr 19, 2006 10:53 pm
by Q-dad
Any idea when Wilbur v1.54 will be available for download...? Only v1.46 is available at the Wilbur web page now...

PostPosted: Thu Apr 20, 2006 7:22 am
by Aaron
Hi Monks,

monks wrote: You know, I was thinking, you've got a reasonable handle on this physics...(slightly). you are very well placed to do this applied math and simulation. Do you have any plans in that regard: in terms of simulation of glaciers for instance. Do you think it could be done?


I should start by saying that my field in Physics is electron optics and analytical spectroscopy (read: zapping rocks with death-rays); geophyisics and numerical modelling aren't my thing. Doing any 'proper' sort of modelling of glaciers/tectonics would require some serious finite-element modelling, which I haven't done in years (I don't even remember which text-book to look for). However, FEM is just a fast and efficient method for solving sets of differential equations, and there are other iterative methods that are easier to implement (though slower, less accurate, generally less informative, etc). The water-table modelling calc is an example of this; it should be done by FEM, but I cheated and used a simpler (though far from correct) iterative routine.

So, getting back to the question(s): Yes; I think that a pretty-good hack could be made for modelling glaciers without stupendous difficulty, but no; I have no plans to try this just yet. I'll think about it.

monks wrote:Tectonics: there's another one: what do you think?


I haven't thought much about tectonics either. Casual consideration suggests that, again, it would be possible to hack together a reasonable-looking approximation, but anything more than that is a task for the big-iron. As before, I’ll think about it, but no promises.

monks wrote:If a dedicated physics processor became mainstream- that would be interesting wouldn't it?


I must admit I'm not very excited about the idea just yet. Firstly, these physics processors are no better at modelling physics than a typewriter is at writing great literature. They’re a tool, and all they do is crunch floating-point operations quickly. To actually utilise one of these things in a program will be just like writing multithreaded programs for dual/multi-core chips; very hard work.

More interesting, I think, is the idea of reverse-multithreading being developed by AMD. This would make multiple physical cores work like one really fast virtual core, which will make any calculation run faster without making programmers shed tears of blood. I image it’s not a very easy thing for them to implement, however, so we might have a bit of a wait on our hands.

Cheers,
Aaron.

PostPosted: Thu Apr 20, 2006 9:05 am
by monks
Q-dad, the current Wilbur download is the latest 1.54. Joe just hasn't updated the info yet on the page: :)

monks

PostPosted: Thu Apr 20, 2006 9:03 pm
by Q-dad
monks wrote:Q-dad, the current Wilbur download is the latest 1.54. Joe just hasn't updated the info yet on the page: :)

Thx :)

PostPosted: Sun Apr 23, 2006 8:03 am
by JavaJones
I hadn't heard about this "reverse multithreading" thing Joe. But glad to hear AMD is working on it. I like the underdog. ;-) Do you have any more info on it?

- Oshyan

PostPosted: Mon Apr 24, 2006 9:51 am
by monks
Much as I thought, it's the dev time, etc. Still, it's always interesting to ponder these realms: especially the glacier, as that is more at our scale, and potentially quite useful. Hey, speaking of death-rays, how about a lil Design-pencil tool by the same name? :D



I must admit I'm not very excited about the idea just yet. Firstly, these physics processors are no better at modelling physics than a typewriter is at writing great literature. They’re a tool, and all they do is crunch floating-point operations quickly. To actually utilise one of these things in a program will be just like writing multithreaded programs for dual/multi-core chips; very hard work.



Of course, you'd have to code for it.

More interesting, I think, is the idea of reverse-multithreading being developed by AMD. This would make multiple physical cores work like one really fast virtual core, which will make any calculation run faster without making programmers shed tears of blood. I image it’s not a very easy thing for them to implement, however, so we might have a bit of a wait on our hands.


Yeah, saw that mentioned the other week. That's quite promising, so basically AMD do much of the work for you: ie, you don't have to code for multiple cores?

monks

PostPosted: Sun Aug 06, 2006 6:24 pm
by monks
For anyone who's interested, there's a new article up on ME-DEM here:

http://www.me-dem.org/content/view/52/33/

It's the first part of our Workflow and covers various terrain modelling apps, including of course L3DT. :)

Cheers,
monks

PostPosted: Tue Nov 07, 2006 10:59 am
by monks
For anyone who's interested, there's a new article up on ME-DEM here:

http://www.me-dem.org/content/view/53/33/

It's the second part of our Workflow and covers watercourse extraction from dems :)

Cheers,
monks

PostPosted: Tue Nov 07, 2006 3:01 pm
by DeathTwister
Wow

Can I link you under Links and Partners in the Texturesource.org website? and have you guys return the Links to TS? That would be sweet, sweet , sweet. You guys are all so awesome.

Yup yup, I will go check out the new article /smiles

DeathTwister :twisted: