FMS landscapes from the beginning

[TreeHugger]

[michaellown]

I am trying to build the example program from the manual as my first landscape to practice on, copied everything word for word substituting rocks with grass (trying for a large flat field that I can build my skills on).

The faulty line of code seems to be: SCENE 500 200 20

Does this mean I have done something wrong with the contour map (what program should I be using to make it?)

Michael

[KEW]

There is nothing wrong with this statement. Send a message containing your .scn file up to the end of the "QUADRANT" statements and we will try to pinpoint the error.
KEW

[TreeHugger]

[KEW]

Hi,

Yes the ground size would need adjusting for 4X4 , It would leave a blank border on two sides. He is using 5X5. The culprit was the name of the contour bitmap file in the previous line. It contained a couple of spaces which upset the program.

Thanks for your comments about the resolution of the height map. I did some experiments with this some time ago and found that the number approximately represents the length of the slope between adjacent contours in meters. The minimum I could get to work consistently was 10, which gives a very steep slope between contours which are adjacent but have a large difference in elevation. I was able to use this very effectively in one landscape, to make a deep lake containing 3D tree reflections which were correct from any point of view.

Regards

[TreeHugger]

[KEW]

Here is a screen-shot of my flying field called Clear Lake. The view is taken from the beach, looking across the lake. The dock and retrieval boat are in the foreground:



Here is another shot looking away from the lake, showing the cabin and pic-nic table:



KEW

[KEW]

[TreeHugger]

I just ran some simliar tests and can confirm your results. The last figure in the scene statement is indeed the grid width in meters. It's an absolute value and doesn't change with scene width.

But the values depend upon another - The bigger the scene is, the coarser the grid must be set or I would receive an error message upon loading: "Access violation at address 00481217 in module FMS.exe". You just can't have a big scenery and a fine ground resolution. That's what I meant by saying a smaller ground would allow for a lower number and a finer grid.

In my experience, the factor between size and grid is about 25, as in these scene statements:
500 ... 20
250 ... 10
200 ... 8
100 ... 4

Unfortunately this relation is not constant across different computers and may depend upon other factors I didn't test. That's why I lost interest in scenery making, without a technical documentation it's just fishing in the mud. The whole idea of setting a grid width strikes me as unneccessary, because the grid resolution is already fully determined by the size of the heightmap in pixels.

Regards

P.S.: Nice screenshots, thanks for sharing!

[KEW]

Thanks for the information on the relationship between scene size and the grid width. I have not tried on other computers yet but when I do, I will be wary.
The grid size also relates to the way the slope zig-zags along a NE to SW or NW to SE contour. I found that the length of the zigs and zags is equal to the grid size.
I have solved a lot of scenery problems by keeping them fairly small. A 5 x 5 layout allows me to use a hi-res square for the center (1024 x 1024) and lower ones for the surroundings (256 x 256). That way I can conserve memory for 3D graphics and still have a very detailed scene.
Thanks for you comments on the screen shots. I was quite flattered when a friend asked me if they were photographs or paintings! I enjoy the challenge of the limitations of FMS. They make me much more creative in pushing these limits.

Regards,

KEW

[KEW]

Here are a two more of my flying sites:

This is the clubhouse at Mile Wide Park. I have finished putting glass in the windows and have the club sign up now:


This auditorium was made entirely from edited pictures taken from the internet:


KEW