This is one of a series of threads that I'm posting.  I worked on an image all weekend and encountered several problems with it.  In fact, I posted it in my gallery and the first comment I got on it fully confirmed my opinion that it sucked so I took it down.  I thought it would have some value to write what happened and post it for two reasons.

1.)  My experience might be interesting for some to see what kinds of things happened during the development of the image.
2.)  I would like feedback on what I could try to get it better.

I think these issues are big enough to have a separate thread for each so I have broken this up into separate threads with images and text to support each issue.

I am happy for any and all feedback!

Thanks!

Smoke
What worked.
Nothing
What didn't Work.
Everything.
What I tried.
I wanted to simulate the ring shaped dust cloud from the jet wash off the rocket blast.
What I ended up with was a ring using a weirded shaped sphere and a surface replicator and then set the blur option.
I also tried to do this with every cloud I own or have . So I tried this with volumetric clouds, 4D clouds, regular clouds and primivol clouds.  Nothing worked.  I wanted a dense puffy variant ring shaped cloud bank.  I did not try a particle generator using spheres.  That's a maybe option I can consider.  I also tried two rings with two different sized objects replicated on it.  So so results (still crummy really).
I also tried a few fog options, zip on results.
What I learned.
Clouds suck for this type of application.
The volumetric clouds that C provides would have been really cool but they just fade into zip when you make them small.  I think they're really designed to be applied in a big space (the volumetricness wasn't something I could figure out how to control).  This was a major disappointment given that I think Volumetric clouds are generally exactly what I wanted but alas, I couldn't get a teeny one (at least I couldn't figure it out).
The primivol clouds were also a reasonable option except that when you stack a gob of them together, they cancel each other out and the only thing that renders is just the parts that don't overlap (not good).
The blur option has a lot of possibilities IF it had better control of the falloff.  Also, post render effects are always a big problem as they interfere with each other when something is behind another something so it's just not a reasonable deal that works very well.  It's clumsy and messy.  I guess I could have used some kind of splat on a particle generator.  That maybe has possibilities.
Let's face it, what I want is a true volumetric puffy dense ring shaped cloud bank.  I don't know how to do it with what's available in C today.  I look at other packages and they seem to pull this off with fluid types of models.  Hmmmmmm.

 Yep, fluid solvers would be nice. The common trick for this type of work is using sprites and then using point clouds the Replicator or Particle Generator to create the cloud mass. (until the recent spat of plug-in fluid solvers for other 3D apps, sprites was just about the only way it was done)

I don' t know if you've seen my tut on this but here's the link: Particle Smoke Trails

Actually I have seen your smoke trail tutorial and I thought it was excellent.  I had  considered using that method but just didn't have time.  I would have to create the splat which probably isn't too hard but not easy either (nothing is easy the first time!).

The other thing I was thinking of was just to use a regulat particle generator with spheres and play with the variation in sizes.  There are still options to explore.

I'm bummed that the volumetric clouds don't work when they're small (only when they're big).  Kind of weird.  I don't know how they're coded to work but I'm quite curious now.

The biggest part of smoke/clouds that makes it really hard is that dang subsurface scattering effect.  Volumetrics are nice and it's something C could use.  For some reason, I thought Bryce has volumetric textures but I'm not sure.