Timeline for OpenGL ES object rotation around z axis
Current License: CC BY-SA 3.0
23 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| S Mar 16, 2014 at 15:33 | history | bounty ended | CommunityBot | ||
| S Mar 16, 2014 at 15:33 | history | notice removed | CommunityBot | ||
| Mar 13, 2014 at 12:44 | comment | added | Itzik984 | @RomanReiner yeah i know. i was trying to come up with a solution to my original problem, it seemed to be working, until it turned out it doesn't... | |
| Mar 13, 2014 at 12:21 | comment | added | Roman Reiner | The projection matrix should not be transformed. That's what the view matrix is for. | |
| Mar 13, 2014 at 11:48 | comment | added | Itzik984 | @RomanReiner Yeah i know that i do have problems with the rotations, the z axis is rotated with the projection, and the x y axis's are done on the model. i was hoping to figure out a way to combine them both, it seems like i will have to take a different approach in order to get it right... | |
| Mar 13, 2014 at 8:39 | comment | added | Roman Reiner | Imagine looking at a clock from its back. It will still go clockwise from its own perspective but for you it will go counter-clockwise. | |
| Mar 13, 2014 at 8:39 | comment | added | Roman Reiner | It seems like you have your view and your model matrices mixed up. When I look at your videos I realized that whenever the model does not rotate the way you expect it is because it rotates around its own coordinate system: Dragging right-left always rotates the model around its x-Axis (wherever that is pointing at the time), dragging up-down always rotates the model around its y-Axis and the same for the z-Axis. For example in the first video: The reason the z-Rotation is correct in one case and (appears) inverted in the other case is that you look at the model from its back. | |
| Mar 10, 2014 at 16:40 | answer | added | codemonkey | timeline score: 1 | |
| Mar 10, 2014 at 14:41 | comment | added | Raxvan | @Itzik984 why are you trying to rotate around x,y,or z axis ? , it would be better to rotate the object around a vector and the vector you calculate based on the camera position | |
| Mar 9, 2014 at 1:14 | comment | added | concept3d | Show us how do you construct the rotation matrix, are you using euler angles? If so then the code is behaving correctly according to the implementation, and you need to take a different approach. | |
| Mar 8, 2014 at 16:56 | history | edited | Itzik984 | CC BY-SA 3.0 |
added 4 characters in body
|
| Mar 8, 2014 at 14:12 | history | edited | Itzik984 | CC BY-SA 3.0 |
edited body
|
| Mar 8, 2014 at 14:09 | comment | added | Itzik984 | Just added some edits and bounty to this question :) just thought to update you and see if you have an answer that pops in your head... | |
| S Mar 8, 2014 at 14:05 | history | bounty started | Itzik984 | ||
| S Mar 8, 2014 at 14:05 | history | notice added | Itzik984 | Authoritative reference needed | |
| Mar 8, 2014 at 14:03 | history | edited | Itzik984 | CC BY-SA 3.0 |
added 570 characters in body
|
| Mar 6, 2014 at 22:41 | comment | added | Fault | Further considering your problem: when you are rotating, you're rotating not only the vertices of the model, but the axes of the model as well. So a rotation of 90 degrees about the x axis would result in the z axis pointing straight up or down. You should be using the world axes for the rotation, not the model's axes, to avoid this. Now where in your code the problem is occurring, I cannot say... | |
| Mar 6, 2014 at 14:35 | comment | added | Itzik984 | And BTW, i would really love to hear if theres an option solving the second problem without moving the camera, because that could be an opening to a whole variety of other bugs :) | |
| Mar 6, 2014 at 14:26 | comment | added | Itzik984 | Thanks a lot for the hints Fault! The first problem was solved, and yeah it was pretty easy. about the second one, i'm having some trouble figuring it out, do you have examples or tutorials about how to do that? currently i'm moving the ׳projection׳ matrix with the rotations i had before (the _currentRotation CC3 vector). But as expected, the results are not good and the x & y rotations are not 'focusing' on the object itself, its just being panned and moved the screen - quite weirdly i must say. i another video is needed, i will upload it. Thanks once again! | |
| Mar 5, 2014 at 4:34 | history | tweeted | twitter.com/#!/StackGameDev/status/441069124421423104 | ||
| Mar 5, 2014 at 2:25 | history | edited | Itzik984 | CC BY-SA 3.0 |
added 4 characters in body
|
| Mar 5, 2014 at 1:47 | comment | added | Fault | +1 for using videos to illustrate the problem. Regarding the first issue, it's an easy fix: when z axis is positive, rotate in one direction, when negative, rotate in the other direction. The second issue stems from the fact that axial rotations aren't commutative. A rotation about the x axis, then the z axis, isn't the same as a rotation about the z axis and then the x axis. It's often easier to actually rotate the camera about the model (with polar coordinates), than keep the camera stationary and rotate the model. (I'd give an answer, if I wasn't short on time). | |
| Mar 4, 2014 at 21:54 | history | asked | Itzik984 | CC BY-SA 3.0 |