The following is from a discussion i created in the GECK wiki. It provides the needed information:

Previous title: "HELP! needed...URGENT"

I want to find out if it will be possible to extract the land form of the "Vault101 Entrance" mountain thing. Including it's rocks and other details.
My goal, somehow, is to try and anolyse the geometry of this cell (-4,-4) and be able to create a contour map which i can then use to create a model in real life so i can use it in my project.
Just wondering if it is possible. If so, how do i go about doing it.

REALLY urgent!

Thank you in advance.

Notes:

• What i am hoping for is that i can somehow extract the cell and its terrain details as a 3D model, in a format that will work in Google SketchUP.
  
• However if there are other ways of doing this, where i can still achieve my goal of creating a contour map OR JUST be able to get the needed info to make a real life model of the cell. (perhaps not the entire cell, but more specifically the immediate area around the cliff entrance to "Vault 101")

Please email me at: ezesonricson@gmail.com (if you need to)

^ The above sums up exactly what problem I have ^

"I have just been in formed that i have a one week extention on my project. So its not as Urgent as before but still quite urgent." ^_^ : March 30 @ 11:38 am CAT (Central African Time)

Below is just what happened in the GECK wiki discussion>

Hi Ezesonricson,

Unfortunately, I don't have an answer to your question. However, I thought I'd let you know that this sort of question is generally much more likely to be answered on the Forums, so you might try asking there if you haven't already. The appropriate forums on The Nexus Forums could also come in handy.

Best of luck,
Cipscis 22:54, 28 March 2011 (UTC)


TEDIOUS SOLUTION DISCOVERED

"Ha ha.. Thanx that would probably be more apropriate. I was in too much of a panic to realise the difference between a wiki and a forum. Though, just by luck I did discover a way of doing this, (in case another lunatic like me comes around here). "
It takes forever but at least its something.

The idea is:

(You need use the orthographic viewpoint and a regular object ( I used the nuka fountain because of its shape and size).)

• By placing the object at regular height intervals and in the same x,y position. U can use the top view to slice the landscape at every object interval. Im not sure what happens but when you use top view in orthographic view, instead of zooming into and centring on the object, it keeps the camera view exactly at the same height but cuts away the landscape at the "zoom height" above the selected object.
  
  You can see now how this can be exploited.
  
  Thats the principle, the hard work starts here.
  
  "I have not been able to figure out how to orientate the "top view" camera so that it can capture the area needed at the best view. (the camera still centres on the object with north pointing up, it just does not zoom into it)"
  
  
• Now that you understand the principle the first step is to find out at which height the object might be so that it first starts to cut the landscape. By using the movement grid, set at a desired movement interval (which will eventually become your contour intervals. I used 32 in this experiment), you move the object downwards bit by bit while using T (topview) each time to see if the landscape is cut.
  
  Once you've established the height at which your object must be in for it to start cuting the landscape you can now move on...
  
  
• The next step is to duplicate the object and move it downwards, while checking the T of the object in the lowest position ever so often to see if it now cuts through the entire area you have chosen (as an indication of when to stop duplicating). Remember to stick to the movement intervals.
  
  
• Now you can start gathering some contour information.
  The following step is just a long repition of:
  
  • Using T on the object in the highest position.
    
  • Capturing the image with print screen (alt printsrcn).
    
  • Pasting the image in a desired image editor.
    
  • Deleting the object you have just used so that the one below can be seen.

This is basically what you do to figure out the contours of a specific site.

You can see how you can get the contours to be quite detailed by adjusting the movement intervals.

Some tips and warnings:

• Fiddle around with the rendering options.. (please save though before doing this) ..so that you can get the best image results.
  
• At some points in the last step you might need to rotate the camera when the object is below the landscape:
  
  • Neva use zoom.. otherwise the consistency of your images will be lost.. unless you can re-adjest the camera into the EXACT same position as before.
    
  • Do not delete the object once you've captured the image and you are now moving onto the next object height. Instead keep the selection on the object and try to get a good view of all the objects below the landscape before deleting it. This will make it easier to select the next object and still keep the view consistent.

Notes:

• It will be great if there is an image capture function in the GECK as the images captured through print screen are low quality and makes it hard to work out whats going on because the image is just too pixelated.
  
• As mentioned above, you can't choose the best orientation of the camera because of the T function always setting north to be up.

^ this is basically what I am doing at the moment and its taking forever... *sigh

Here's one possibility, although it does not work for the ground itself, only the rocks and other static items:

1) Select all the items you wish to include in your model

2) Press ctrl-g. This will cause the GECK to create a 'static collection' - basically a single NIF file containing all of the items you have selected. You give this collection a name and a .nif file with this name will appear in the fallout3/data/meshes/SCOL directory.

You thus then have a model in .nif format of the area in question (although, as I said, missing the actual ground). I don't know if that helps, but you can get Nifskope to export the model as a .obj file which you hopefully can then import into other software.