OctaneRender™ Standalone v1.0 RC 3 (obsolete)

[Refracty]

looks great roubal

[abstrax]

Thanks for the new version and for the mac .dmg, unfortunately all the rc os x versions are extremely unstable, it's very difficult to do the benchmarks. here are a couple of crash reports, i hope that you could find a solution
ciao beppe

The crash reports unfortunately don't point to any particular problem. We don't have a Mac Pro here, so all our testing is happening either on a MacBook Pros or on a Mac mini and there I couldn't reproduce the issues. I guess you tried different CUDA driver versions? What is you hardware configuration and is there a way to reproduce the issues?

Is there anybody else who has these problems on Mac OS?

Thanks,
Marcus

[FooZe]

It's looking good! Unfortunately the turntable animation is still broken. At least it doesn't work as it used to. I can select a camera target to rotate around but it still skews the rotation. It is impossible to get the rotation to be limited so it goes around the up axis. Also the camera target position changes as the scene is animated which doesn't seem right to me?

Jason

Hi Jason,

I'm looking into this but i can't seem to duplicate your problem (or at least not in a way which is problematic).
I do notice the camera target change just slightly on starting the animation. IE: x,y,z of 0,0,10 will change to 0.0000008,0.0000112,10.0000486. However this was the case with 2.58e also.
I'm sure the dev's will know more precisely but i'm fairly confident this will be a by product of the calculations involved in finding the camera path/locations for the frames.

Perhaps you have the camera's "up" parameter slightly off vertical?
Can you give an example or something i can follow to help reproduce your problem here? Maybe post a simple scene that causes your problem?

Cheers
Chris.

[bepeg4d]

thanks marcus for looking into it, i'll try to find a way for reproducing the issue
ciao beppe

[grimm]

Thanks for looking at this Chris. I must have had the camera's up off like you suggested, because it's working fine now. I could have sworn I checked it on the last version. So I don't know what I did. Thanks again,

Jason

Hi Jason,

I'm looking into this but i can't seem to duplicate your problem (or at least not in a way which is problematic).
I do notice the camera target change just slightly on starting the animation. IE: x,y,z of 0,0,10 will change to 0.0000008,0.0000112,10.0000486. However this was the case with 2.58e also.
I'm sure the dev's will know more precisely but i'm fairly confident this will be a by product of the calculations involved in finding the camera path/locations for the frames.

Perhaps you have the camera's "up" parameter slightly off vertical?
Can you give an example or something i can follow to help reproduce your problem here? Maybe post a simple scene that causes your problem?

Cheers
Chris.

[bepeg4d]

ok, effectively i have found a newest os x nvidia driver and now all seems much more stable:
http://www.nvidia.com/object/macosx-295 ... river.html
ciao beppe

[karanis]

Looks greater... And really faster, and colors (especially glossy ones) are better ...

Just got quite a bit of "hot pixels" on same scene`s wrt prv. versions (I hate "hot pixels" and always try to setup to not to have them) but I`m not 100% sure that every parameter was the same, I`ll check it out deeper and write here if I realize something important. I think it was my fault now. (or maybe heterogeneous usage of 5xx and 6xx and FP calculation diffs ).

And about the turntable animation I must say, it was the only disappointing feature for me in Octane. I resolved my problems after giving up to use it . I am not thinking about "where the mistake is" anymore because I think it`s a concept that must be re-handled, rewritten. As a normal user, as the name says, I was expecting a turntable. But in POV, it is a table that I`m moving around, which effects everything. Many poor renderers solved the problem by selecting a model to rotate in selected axis (in specified degrees/frames). It would be a very easy solution and a big addon for our georgeous renderer. That would also require sticky procedural textures to local corrdinates of the subject model, bla blabla. Anyways, dear Grimm, if you can make a test like unit vector setup (all translations are 1 and camera angle is 45 degrees down, targeted to center of your object) I`m expecting everything to work out fine.

Finally, congratulations and thanks guys... Very good job.

[nildoe]

Very good news about rc3 congrats otoy team, i have a question regarding the improvement on Keplers.. how is that compared to the 5xxx series fermi based cards now? are they the same? faster? still slower?

Nildo

[MattRM]

Nice news !!!

Matt

ps : Could you send me and give me more answer about *.ocs file writing process and file description please ?

[madcoo]

Great news !
Thanks for this new release, I'll give it a go ASAP.

Cheers guys