Cheetah3D 6.0 is out - What's next?

Cheetah3D 6.0 is out - What's next?

Hi,
since Cheetah3D 6.0 is out now I want to give a small preview of what I plan for the v6.x updates.

Cheetah3D 6.0 already has a very wide feature set but there are often the small things missing which can be hugh timesavers. Luckily these small things can often be implemented easily.

With v6.1 and v6.2 I want to focus on the rigging and character animation tool. Since the character animation tool saw no real update since 4.0 this is more than overdue. Here especially things like a better IK solver, spline IK, support for oriented joints throughout the whole tool chain, splint joint, parent object, mirror joint, etc. come into my mind.

One later updates I want to focus on the modeling, rendering, UV mapping and various UI features. So you can expect improvements throughout the whole toolchain in v6.x. The only thing that probably won't come are major new features which take me half a year for just one new tool.:wink:

Cheetah3D 6.x should become something like Snow Leopard was for Mac OS X. I want to focus on improving and extending the existing tools. So there will be many small and medium size features which lift Cheetah3D on the next level. :smile:

Bye
Martin
 
Sounds great. It's generally the point releases that excite me the most because of the workflow improvements.
 
Thanks Martin. Your communication is very appreciated. I do hope we'll see some improvements to rendering. I realize that a new shader (SSS) might be beyond the scope of your plans, but I hope you'll consider it. In general it is good to hear the long term goals of Cheetah and I look forward to the improvements ahead.
 
Thanks Martin. Your communication is very appreciated. I do hope we'll see some improvements to rendering. I realize that a new shader (SSS) might be beyond the scope of your plans, but I hope you'll consider it. In general it is good to hear the long term goals of Cheetah and I look forward to the improvements ahead.

I can't promise SSS yet. Since integrating SSS properly with global illumination it is so small job.

Rendering features like better interior lightning, photon mapping, light portals, better irradiance caching, higher render resolution, sub-polygon displacement mapping and motion blur have a higher priority on my renderer todo list.

Bye
Martin
 
Hi,
since Cheetah3D 6.0 is out now I want to give a small preview of what I plan for the v6.x updates.

Cheetah3D 6.0 already has a very wide feature set but there are often the small things missing which can be hugh timesavers. Luckily these small things can often be implemented easily.

With v6.1 and v6.2 I want to focus on the rigging and character animation tool. Since the character animation tool saw no real update since 4.0 this is more than overdue. Here especially things like a better IK solver, spline IK, support for oriented joints throughout the whole tool chain, splint joint, parent object, mirror joint, etc. come into my mind.

I want to focus on improving and extending the existing tools. So there will be many small and medium size features which lift Cheetah3D on the next level. :smile:

Hi Martin, that sounds good. I'm very curious to see how the joint features will be by 6.2.

When I animate a character, I rely largely on poses that I can adjust much easier with mostly just a skeleton/joint set up. The challenge I have is making adjustments when the character pivots on one foot, and when I want to have his/her hips lean at an angle while still pivoting on that foot. I'll look into the more complex solutions, but I usually end up just moving the character's joints in free space with rotation of the Lead Joint to make the adjustment, then set the pose after bringing the foot back to where it was before the adjustment. It's a bit clumsy, but it's the fastest way to set up a character, as though they were a piece of clay to adjust instead of a complex marionette.

I'd love to work with animation like it was claymation/stop-motion, but I know you mentioned other animation software doesn't offer the ability to rotate the chain of joints above or below at will on any joint.

Thanks again for an awesome program in Cheetah 3D, Martin. Keep up the great work! :icon_thumbup:
 
Last edited:
Since I tend to render translucent surfaces that need subsurface light scattering, SSS is a bit more important to me. But I realize that you need to implement the features that the majority of your customers use and that you can't cater to the minority. I'd be happy with a fake SSS effect that might require less modification to the shader system and would result in flexibility for perhaps other effects. It would just require some modifications to how transparency works.

See: http://www.peranders.com/w/index.php?title=Fake_Sub-Surface_Scattering
 
Wow, that is a interesting paper. Thanks for letting me know. But for the first I have to work on non-stretching bones.:redface:

Bye
Martin

Despite the many wishes that come up (typicly more around every new release :redface: ) I'm allways more than happy whith what you've given us an what we can expect! Thanks a lot for showing interest in our suggestions!

I like the fake sss idea tough ;) would be cool if this could be acheeved in cheetah easilly...
 
Last edited:
It looks to me that this fake SSS trick (a) is actually very close to real SSS but doesn't do the lighting calculations in screen space in a pre-render pass to save on computation (i.e. it's conceptually similar to actual SSS, just less efficient) and (b) requires some features which C3D currently lacks in its render pipeline. The "absorption" business is something we simply don't have (and which would be damn useful). Also the way he's blocking light from getting out would seem to require being able to distinguish inner from outer surfaces (which, again, would be damn useful — hmm maybe that we can do via normal calculations).

Given Cheetah 3D's current limitations, this trick is quite similar to this faux SSS trick (except instead of having an absorption setting I used inner geometry to block light transmission beyond a certain point).
 
It looks to me that this fake SSS trick (a) is actually very close to real SSS but doesn't do the lighting calculations in screen space in a pre-render pass to save on computation (i.e. it's conceptually similar to actual SSS, just less efficient) and (b) requires some features which C3D currently lacks in its render pipeline. The "absorption" business is something we simply don't have (and which would be damn useful). Also the way he's blocking light from getting out would seem to require being able to distinguish inner from outer surfaces (which, again, would be damn useful — hmm maybe that we can do via normal calculations).

Given Cheetah 3D's current limitations, this trick is quite similar to this faux SSS trick (except instead of having an absorption setting I used inner geometry to block light transmission beyond a certain point).

I feared it wasn't going to be easy :smile: thanks for explaining!
i also like your faux sss solution!
 
Ah well - ZBrush BPR gets better in rapid steps, soon it will be able to replace all uses of external apps for me.
 
Further info on the faux SSS technique linked above. C3D's shader system cannot, at this time, distinguish between back faces and front faces (as I feared) ... i•n is calculated as though all faces are front-faces. (I don't know how hard this might be to fix, or if we could get an i•n channel that allowed us to differentiate them for special effects.)

If this worked as expected then the node setup attached would let us render the inside (back) of the test sphere.

If we did have such an ability, among other things we could use the "create a shell for an object, flip it, and make it black" technique for adding outlines when toon-shading. I don't know about blocking light passing through objects — the way this works in C4D may simply reflect a property of C4D's render pipeline. The way the blurred transmission trick works depends on light ignoring back faces for some reasons and not others.
 

Attachments

  • Screen Shot 2012-05-02 at 9.20.57 AM.PNG
    Screen Shot 2012-05-02 at 9.20.57 AM.PNG
    44 KB · Views: 529
Great!

Hi Martin,

Thanks for the sneak peeks at what's ahead. It really is exciting to know what's coming down the road.

Any word on the more "robust" boolean operations you had mentioned previously?

Thanks,

Rick
 
Back
Top