Jump to content

  • Log In with Google      Sign In   
  • Create Account


Member Since 21 Jan 2009
Offline Last Active Feb 25 2014 07:16 PM

Posts I've Made

In Topic: Keeping track of torations arround arbitary points

16 February 2009 - 01:04 AM

Well, it seems that i need your help again about this matter. Everything works smooth as , up to a point but then it breaks. I use the angles that i extract from the rotation matrix along with the translations to move to that location and then glRotate so i can finally draw some things... and everything is good up to a point when i can see that the object i am drawing faces on a weird angles and not towards where it is supposed to be...

I checked my code and is similar to what i have found over the internet and people say that it is indeed working but when i do arbitary rotations to the 3d space it seems that things brake at a point and if i keep moving arround they get fixed. I tried to understand when things break and why but i could not find anything..

I am open for different suggestions if you have any on how to draw on the place where the camera is :)
I can post my code if you think that it helps but i am just using something similar to http://www.gamedev.net/community/forums/topic.asp?topic_id=319213 ths one to extract the angles.

Thank you one more time, your help is priceless :D

In Topic: Keeping track of torations arround arbitary points

13 February 2009 - 01:01 AM

If you extract the angles from the rotation matrix there are ambiguities which can lead to object facing the wrong way if you draw something with that extracted angles.

Is that ambiguity resolved if you use quaternions ? Cause from what i see there is not much of a difference cause in both ways you need to get the atan which actually will give two solution for each angle.

Any help ?

In Topic: Keeping track of torations arround arbitary points

09 February 2009 - 12:52 PM

Thank you for your answer. Let me provide some more info then and see where it goes :)
The interface is not something special i think. It is a 3d world representation in order to simulate a "virtual" test field for photogrammetric calibrations. So the user can move and rotate freely at the test field. I dont want to keep track of all his movements and rotations but at the specific points that the user chooses i want to know the location of that specific point as it is the exterior orientation of the camera. And to know that what i actually need is the 3 tranformations and 3 rotations from the origin of the coordinate system.

I dont really know what else to say. I hope i provided some more info but if you have any specific question i am more than happy to answer. At the moment i am using an actor frame to manage the moving arround the "world".

PS. I dont have to use Euler angles but i need to know the Euler angles as an output lets say (its a simulation so what i test onwards is how accurate i find the orientation and location so i got to have something to compare it with :) ) and also cause i want to draw something at that location which can only be done with glPushMatrix() transform and rotate with that position and orientation in space for every "selected location". I am open to suggestions though :) sure there is a better way to do it.

What is more accuarate quaternions or Euler? i know i can extract the euler angles from the rotation matrix but there can be different combinations that satisfy that rotation. Since i have never worked with quaternions i am asking for you experience over the matter!

Thank you again !

[Edited by - Xerion on February 9, 2009 11:52:42 PM]

In Topic: How can i create gradient circles ?

26 January 2009 - 05:55 PM

Either i am missing something or its my fault that i forgot to mention that i am using OpenGL.

Thanks for the idea though i think i can still use something similar in opengl and create gluDisks with the same center but have a hole in the middle so each one gets drawn after the other with different colours.... but this will require lot of resources!