Indeed!

I have problems with selection the right LOD stages for the terrain (something with distance calculation and stuff...)
but i could do a 2D terrain system for sure with this technique (but i don't know what's about collision)

Hi! Im back and happy
You guys have been helping me a real lot with this! And i thank you all!
Some questions still remain in my head:
-Where to start?
-About the cube to sphere conversion, love the idea! But wait, using a cube model would't seem right. That leaves me with terrains. So, how do i rotate 6 terrains in order to produce a flawless cube? I've been trying some "homemade" algorithms, but none have worked..
-I love math, so, please, give me mathematic algorithms! I've red a lot both on perlin noise and on procedural planets, but it's mostly just blogs of people telling what they've done. I don't want just to know what they've done, i want to know HOW they've done it. What algorithms, what procedures, i want to actually know they path they used in the programming language, so that i myself can get a clue on what i should do get the same result in Lite-C
So, once again, help me

thanks

Really loved it if someone could continue answering to this topic, pleeeeeease

create a cube out of 6 "models"

each of those models gets transformed via shader to build a sphere (just normalize each vector to 1)

each of those models is made out of a quad tree which subdivides the surface of the model into more polygons if necessary (just a distance check for camera will do it in most of the cases)

i've made a quick test with a plane mdl which is in the range of [-1,0.5,-1] to [1, 0.5, 1]

you need to set the scale and position for each entity representing a part of quadatree manual, but it works pretty good if work with XNA tutorials as a template

Ok, so i work with GameStudio and im not doing this out of actual models, but out terrains. So, to build a cube i'll need 6 faces, that is, 6 different terrains. 1 will remain normal, the other 5 will have to rotated to get the cubic shape. This is one of my first obstacles, because transforming a cube to a sphere, the part about subdivision, all of those things i can manage to get there with a little research, i've actually red a lot of things by now. But this is my first "pain in the ass" obstacle, and i can't get past this because, in GameStudio it's said that terrains cant be moved and/or rotated, this really screws this whole project for me because it wouldn't make sense to use an actual model. Yes, those can be moved and rotated, but it just doesn't make sense to use a sphere model as an earth-like terrain
Are you understanding my problem? It's this one problem that just won't let me keep going with this project, i don't know how people do it, something is missing in my way of thinking

please don't use terrains!
you can't rotate them (because of LOD reasons and terrain chunk clipping)

just use models, in your case they aren't a bit faster than normal models

Thanks! Finally i know where and how to start
but to make an earth-like model i'd have to create a HUGE cube, to later transform it into a sphere (or could i make a sphere model right away?)
Will it still be possible to use subdivion algorithms with 3d models? I got the ideia that in GameStudio those would only work with terrains
But in the end, to simulate an actual universe where you could get lots of detail and actual terrain variations when close to each planet, wouldn't be possible with gamestudio's terrains because these can't moved and/or rotated, therefore my universe would be completely static
you have helped me alot! thank you!

the problem with a sphere model is that you cannot do math that easy as with a cube.
just use a shader which normalizes each vertex position before modifying it with matView and matProjection.
Then you will get a sphere. The subdivision is made with models you don't subdivide the models. Gamestudio also cannot subdivide terrains

When i have enough time i'll try some stuff with GameStudio and eventually will post the results here
Procedural Planets are really hard but the motivation is greater than the dificulty
thank you!