-
1. Re: CS^ timeline bug when rotating a sphere
Curt Y Nov 8, 2012 8:36 AM (in response to cyberwasp)Could be a video driver problem also. What GPU, and do you have latest driver (latest does not always mean best however)?
-
2. Re: CS^ timeline bug when rotating a sphere
cyberwasp Nov 8, 2012 9:03 AM (in response to Curt Y)Curt. I created Jupiter first and succesfully roated it in timeline. I even exported it to an mp4. I then created mars a few days later and was unable to recreate the rotation using timeline. The only update that happened was the last one for cs6. Am still using the same video driver since I created jupiter.
Several other people on the NAPP forum have tried my example with the same results and it's why I'm trying to get adobe to see it. Video card is nvidia Gforce gt 545 3gbyte
-
3. Re: CS^ timeline bug when rotating a sphere
conroy Nov 8, 2012 8:04 PM (in response to cyberwasp)I guess the problem is the way Photoshop represents rotation of mesh and extrusion objects. Rotation of the Environment, View, Scene and lights are recorded as the specified negative or positive number of degrees, but the specified rotation of a mesh or extrusion object is transformed to values between -180 and 180 degrees and then recorded.
For example, rotations of the Scene about any axis by:
-720, -300, -180, -45, 0, 30, 90, 180, 360, 480, 720, 3600 degrees
will be recorded as:
-720, -300, -180, -45, 0, 30, 90, 180, 360, 480, 720, 3600 degrees
In contrast, all rotations of a mesh or extrusion seem to be recorded as being between -180 and 180 about the X and Z axes, and between -90 and 90 about the Y axis.
Say you set the initial keyframe with no rotation of the planet mesh, and the next keyframe with a 3/4 revolution of the planet about its Y axis. The animation will be rendered as a -1/4 revolution between the keyframes.
If you try to set that second keyframe with a full revolution of the planet mesh, the animation will contain no revolution between the keyframes.
A workaround is to set the initial keyframe at 0 revolution of the planet, then keyframes at 120, 240 and 0 degrees again. You'll have to use the on-screen manipulator widget to rotate the planet by 120 degrees at a time because directly entering 120 and 240 in the Properties panel will result in 60 and -60. The values 120 and 240 won't actually be recorded as such, but effective equivalents will be recorded if you use the widget.
If there is only a single object (planet) in the animation then you could workaround rotation complications by rotating the Scene instead. An initial keyframe with 0 degrees rotation of the Scene followed by a keyframe with 360 degrees rotation of the Scene will be rendered as a full revolution between these keyframes. If that second keyframe is specified with 3600 degrees, the animation will contain 10 revolutions of the Scene between these keyframes.
You should report this to Adobe. I would, but they don't usually respond to my problem reports.
Edit: made changes to first workaround paragraph.
-
4. Re: CS^ timeline bug when rotating a sphere
cyberwasp Nov 9, 2012 10:59 AM (in response to conroy)Conroy.After 5 different techs I found one that actually used Photoshop ande was willing to listen. We both created a plain sphere and rotated it manually and using timeline. He agreeed there was a program flaw and is taking it to the next level. Whether they patch it or wait for the next release, here's hoping.
-
5. Re: CS^ timeline bug when rotating a sphere
conroy Nov 9, 2012 11:40 AM (in response to cyberwasp)A terrible problem on OS X was introduced by the 13.0.1 update. If you click the mouse in any application's window, to make use of the computer while Ps is (slooooooooooooooowly) rendering, the render is halted. It makes the computer virtually useless for the duration of a Ps render.
It must be a couple of months since I reported that!
-
6. Re: CS^ timeline bug when rotating a sphere
cyberwasp Nov 9, 2012 12:38 PM (in response to conroy)Yeah that would be annoying. I see a lot of people complaining about render times but feel that would be machine dependent. Most of my renders go quickly, under five minutes. But than I'm using a 3.2ghz six core machine with 12 gig and 3gig video card. if I knew I'd be doing 3d I would have gotten at least 24g ram. Maybe I'll get brave and add some next year. but right now it's quick enough for me.
If I were you, I'd report the bug again. the more noise you make the more gets done.
-
7. Re: CS^ timeline bug when rotating a sphere
conroy Nov 9, 2012 2:02 PM (in response to cyberwasp)Cool!!! I certainly hope you'll post a link to the result.


