If you can't beat em, use them as your back-end

Blender is open source GPL code, so borrowing stuff from it would force C3D to become open source and effectively free, so incorporating "good stuff" from Blender isn't practical. But using it as a back-end renderer is totally viable, and it could be almost as seamless as using Cheetah 3D's internal renderer.

In general, the step from editing to rendering is disjoint in most apps — in many apps the rendering engine is actually an entirely separate app that is only loosely bound to the user interface app. In high end apps, the code is factored so that third-party renderers can be swapped in (e.g. plenty of Maya folk use Maxwell or Renderman to produce final output).

Everyone loves writing rendering engines.

Rendering is a hot topic right now. Indeed it seems it's always been popular. There's intense competition to implement new renderers based on physically based rendering (new theoretical rendering model) and GPU acceleration (new implementation model). Cheetah 3D's renderer is ray-tracing based (older model) and CPU based (old implementation model). It's very good for what it is, but it's missing features many of us consider essential (e.g. volumetrics, motion blur, SSS, render-to-layers, etc. etc. etc.) and it has two feet firmly planted in the past.

Now, if Cheetah 3d had a native plugin interface and was hugely popular, guys like Maxwell, Lux, Renderman, Octane, etc. would simply do the integration work themselves so this would be a non-issue.

The problem for Cheetah 3D — in my opinion — is that it's on the wrong side of history. There are ridiculously good, cheap, fast, high quality renderers out there (Lux is free, Blender — which has three different rendering engines — is free, Octane is cheap (199 euros), and so on). Rendering has become commoditized, so treating your renderer like it's the crown jewel of your 3D app is wrong-headed.

When I first played with Cheetah 3D (I think it was version 3) it was a rudimentary modeling program with a very good renderer. When version 4 came out with rudimentary character animation functionality and subdivision surfaces I jumped at it. Times have changed. It's now a very capable modeling program with a dated renderer.

Economics 101

Joel on Software, in "Strategy Letter V", discusses the fundamental principle of economics w.r.t. selling products — you want your competitors to be expensive and your complements to be cheap or free. Right now, really good renderers are really cheap or free, so positioning yourself as a competitor to them is insane. On the other hand, 3D software with decent usability is insanely expensive or even non-existent.

http://www.joelonsoftware.com/articles/StrategyLetterV.html
 
I already knew how to use Blender, Lux, Yafaray, etc. before buying Cheetah. Heck, Bunkspeed flew me down for a job interview as a trainer / technology demonstrator almost 5 years before I bought Cheetah. I authored much of what you can read about Global Illumination and other rendering topics on English Wikipedia.

So I didn't buy Cheetah because I was sold by its rendering engine. I bought it because it seemed like a well-made, relatively inexpensive native Mac app. It is fun to use and works super well.

It's still nice to know how to use Blender and other 3D apps, but Cheetah just occupies a different part of my brain. :smile:
 
Interesting old thread. I came here looking for a volumetrics thread but found this. You guys read my mind about C3D and Cycles, wouldn't it be great.

It would be nice to hear about plans for falcon tho
 
Back
Top