Archived

This topic is now archived and is closed to further replies.

WebsiteWill

Reverse Engineering Motion

Recommended Posts

I am in need of the ability to create a model of a solar system in a geocentric model. Heliocentric is easy to model as everything naturally rotates around a static (relatively) object. However, I need this perspective from the Earth''s point of view. As if the Earth was stationary (no rotations or revolutions) and everything else rotated and revolved around us. I am having a really hard time both mimicing this movement in 3D Studio and with coming up with the functions that model this movement. Anyone have any experience with this? Theoretically, if I base everything off of the Earth then it would be a matter that the sun would revolve around the Earth once every 24 hours (to model a single Earth rotation) but also take into account the extra revolution every 365 days along with the tilt of the Earth''s axis. Pretty confusing stuff. Help is much appreciated. Thanks, Webby

Share this post


Link to post
Share on other sites
If you''re trying to do this on 3Dstudio, the only way I think you can do this is to build an heliocentric model (everything rotates arround the Sun) and then attach a camera to the Earth, so when you render, the camera will always appear positioned relative to Earth.

Share this post


Link to post
Share on other sites
I was using 3D Studio as my basis. I figur once I get my motions correct there then it wouldn''t be such a big deal to implement then in code. The end result needs to be a program similar to a VERY scaled down version of Celestia that only takes into account the equivalent of Sun, Mercury, Venus, Earth, Moon, Mars. But the Earth in the simulation needs to remain stationary and a camera mounted on the Earth will still see the movements of everything else in space. Kind of the say we see things now ever though we know the Earth is actually moving. For this model the Earth cannon move.

Webby

Share this post


Link to post
Share on other sites
I would do the helicentric model, and in your code afix the camera to the earth. This will always make your earth 0,0,0 relative to the POV. ie: Earth doesn''t move.

Good luck,
Will

Share this post


Link to post
Share on other sites
That''s the point though. I don''t need the illusion of it not moving, I need it not to move. The Earth has to stay in a fixed position and everything else has to mimic the movement around it. Fixing a camera to the Earth was my first idea as well and it was quickly blown out of the water...

Webby

Share this post


Link to post
Share on other sites
Guest Anonymous Poster

Run your model to move everything around the sun, then as the last part of the model, translate and rotate the scene by the earth''s current location and rotation. This should be mathematically equivalent to fixing the camera to the earth, but you will do it in model space.

Share this post


Link to post
Share on other sites
But, what is the difference in both terms? I mean, if you see everything move around, how do you know its not you thats moving? lets say, we are in a space ship... Whenever we see outside the window, everything is moving, the earth, the moon, etc... so... how would it be different if you say the space ship was moving, and not everything else?
What I think you want to do is to make as if the earth was in front of the camera, then you could do the camera fixing(i know this has been said loads of times) to the earth. Like, have the camera''s position at 0,1,1 and have the world and camera''s view at 0,0,0.

Share this post


Link to post
Share on other sites
Something like that could work. I''ll have to ask some more questions from the person requesting this. Though I am pretty sure that what he wants to be able to do is zoom out like you can do in Celestia and watch things move around the Earth. So just fixing a camera to the Earth''s surface probably won''t get the job done.

I think I am getting some where with 3D Studio.

For instance. In our solar system Sol is center and the Earth revolves around it while at the same time rotating on it''s axis which is tilted 23.5 degrees away from the sun. This is what gives us our seasons. I am currently running a simulation where the Earth is static. I am animating Sol around the Earth once every 24 frames (to represent a day). Sol is animated in a circular path. I am animating the tilt of the path from
-11.75x, 11.75y to 11.75x, -11.75y over the course of 8640 frames. This effectively represents a whole Earth year (360 days not 365 to make the math simpler).
The effects appear to give the same results as if the movement was Sol centered as far as daily light cycles are concerned. Now I think that since our moon revolves around us naturalls then I can just have it animate around Earth and it will work nicely. Mercury, Venus and Mars all rotate around Sol so I an thinking that I can do that as well. If it all looks good I am going to attempt to write up the application to view it with a free roaming camera and allow the adjustment of speed. Then I will know for sure if the seasons are proper or not. This will also allow me to make the correct alignments based on planetary positions at a single point in time.

Hope it works out. So far it''s looking like it just might.

Webby

Share this post


Link to post
Share on other sites
Actually, just realized that this method is not working.

What should be occurring is that I should probably keep the Sol orbit level to the Earth and simply move it''s value up and down with respect to the Earth''s equator. Say for instance that the Earth model is 360 units tall and Sol''s orbit is centered at 180 units (it revolves perfectly around the equator of Earth). Then over the next 180 days I animate Sol''s path up 11.75 units followed by down 23.5 units over the following 360 days. And repeat.

Dunno, it looks right, but then it doesn''t!!!

Webby

Share this post


Link to post
Share on other sites
quote:

Something like that could work. I''ll have to ask some more questions from the person requesting this. Though I am pretty sure that what he wants to be able to do is zoom out like you can do in Celestia and watch things move around the Earth. So just fixing a camera to the Earth''s surface probably won''t get the job done.



Sure it would. If you need to transform the camera, transform it relative to the earth and you''ll be fine.

I''m going to throw another two cents on the side for a sun-centric model. Try modelling some of the planets (or even better something like the moons of jupiter) before you worry about perfecting the sun - they are much more difficult to get right, since they appear to change speeds and reverse direction relative to the earth.

If you''re set on this, though, the phrase to google is ''Ptolemaic planetary model.''

Share this post


Link to post
Share on other sites