27 Nov 2015, 17:28
(This post was last modified: 30 Nov 2015, 16:23 by Flávio San.)
I think we're having trouble with names.
Some API methods perform the rotation carrying the other axes together.
Some methods do not.
The methods that has the description "by euler angles" carries in rotation the other axes together.
The methods that has the description "axis angle" do not carry in rotation the other axes together.
The problem is that if you rotate the cube with "by euler angles", you move the other axes along with the cube rotation. For example! Imagine the cube spinning a horizontal rotation, to the left or to the right with the method "by euler angles". You must realize it by the Y axis because it is the axis that we will spin horizontally for our camera.
Then the cube rotate horizontally carrying the X and Z axis, because that's how the method "by euler angle" perform rotation, carrying the other axes. So, for example, if you stop the cube 1 turn angle (HALF_PI) rotation Y axis, using the method "by euler angle", the X axis becomes to the position of the Z axis in the camera we are using.
If you rotate the X axis, for our same camera, we have a rotation identical to the Z axis. That's the problem of rotating using the methods "by euler angles". We carry the other axes.
What I'm talking about is that missing a method that can I perform, in the same tick, the rotation of two or more angles without carry any of the angles at their spatial position.
Some API methods perform the rotation carrying the other axes together.
Some methods do not.
The methods that has the description "by euler angles" carries in rotation the other axes together.
The methods that has the description "axis angle" do not carry in rotation the other axes together.
The problem is that if you rotate the cube with "by euler angles", you move the other axes along with the cube rotation. For example! Imagine the cube spinning a horizontal rotation, to the left or to the right with the method "by euler angles". You must realize it by the Y axis because it is the axis that we will spin horizontally for our camera.
Then the cube rotate horizontally carrying the X and Z axis, because that's how the method "by euler angle" perform rotation, carrying the other axes. So, for example, if you stop the cube 1 turn angle (HALF_PI) rotation Y axis, using the method "by euler angle", the X axis becomes to the position of the Z axis in the camera we are using.
If you rotate the X axis, for our same camera, we have a rotation identical to the Z axis. That's the problem of rotating using the methods "by euler angles". We carry the other axes.
What I'm talking about is that missing a method that can I perform, in the same tick, the rotation of two or more angles without carry any of the angles at their spatial position.
(25 Nov 2015, 20:48)Ketschak Wrote: Could you describe in detail what you are trying to achieve?
✠ nnDnn ✠ The peace of Christ! ✠ nnDnn ✠