Re: cigarette smoke

["Ajit Menon" <xsiajit(at)gmail.com>]

This may seem like blasphemy but have u tried using something as simple to use as After Effects' plugin particular?
You can move XSI camera data into After effects using mindthink.de
's camera export script. 

The 2d particle plugin supports 3d depth... and if u use the particles in conjunction with Auxilliary emission, and play around with with Turbulence and spion u can get some amazingly cool wispy cigarette smoke. You can link the xyz data into the particular emitter's coords by an _expression_...

It doesn't work so well if the camera is super close to the smoke as you will need a million or more particles to keep it smooth.... but it's worth a try.

On 4/25/07, Schoenberger <XSI(at)digidragon.de> wrote:

  |> I sincerely hope they implement some "slave particles" or "virtual
  |> particles" or whatever you want to call it for the
  |> upcoming particle system.
  |> unless you can do gazillions of real particles without a
  |> big penalty.
  |> But somehow I dont think it's ever gonna be enough.

Yes, I already tried that.
I made a small thin particle simulation and no matter how many particles I add, it is always a bit noisy.
So there need to be a different way for it. Perhaps if you increase the size and use metaballs as volume instead.
The next thing I tried was to normalize/clamp the densities, alread helped a bit.

(at)Kris: That is something you can try too.
On the misc tab, try to change the output to "normalized with falloff".
And about the sample scene I made. The distort noise is good to add some structure (looks very nice in a still), but not very good
for smoke animation. Perhaps you can add a volume_texcoord as coord input for the fractal and move the noise in particle direction
(usally to the top).
Or _expression_-link the position of the coord to another implicit sphere. (The one in the scene controls the intensity of the
distortion)


Holger Schönberger
technical director
The day has 24 hours, if that does not suffice, I will take the night



  |> -----Original Message-----
  |> From: owner-xsi(at)Softimage.COM
  |> [mailto:owner-xsi(at)Softimage.COM] On Behalf Of peter boeykens
  |> Sent: Tuesday, April 24, 2007 10:09 PM
  |> To: XSI(at)Softimage.COM
  |> Subject: Re: cigarette smoke
  |>
  |>
  |> getting rid of the lots-of-seperate points look, and
  |> trying to get a smooth
  |> and continuous result is one of those things where you
  |> spend a lot of time
  |> with particles.
  |> I sincerely hope they implement some "slave particles" or "virtual
  |> particles" or whatever you want to call it for the
  |> upcoming particle system.
  |> unless you can do gazillions of real particles without a
  |> big penalty.
  |> But somehow I dont think it's ever gonna be enough.
  |>
  |>
  |>
  |> ----- Original Message -----
  |> From: "Nguyen Tran" <nguyen(at)topixfx.com>
  |> To: <XSI(at)Softimage.COM >
  |> Sent: Tuesday, April 24, 2007 9:46 PM
  |> Subject: Re: cigarette smoke
  |>
  |>
  |> > I've had decent success with the genietail shader under
  |> 32bit XSI5.11:
  |> > http://www.binaryiris.com/main/?id=GenieTail
  |> >
  |> > My shader settings:
  |> > step= low value like 10-20,  maxSampleParticles 1,  (low
  |> is less lumpy)
  |> > resolution factor=2, blendRadius=5000, maxMatches=100
  |> >
  |> > emissionrate = around 4000
  |> >
  |> > You'll have to mess around with it to find a good balance of
  |> > quality/speed.
  |> > Nguyen.
  |> >
  |> >
  |> > Kris Rivel wrote:
  |> >> I was trying to search the archive on Softimage's site
  |> but as usual...its
  |> >> broken.  Anyway, any new approaches to doing wispy
  |> cigarette smoke
  |> >> blowing around?  I know you can get some nice smoke
  |> effects with Holger's
  |> >> volume shaders but I'm having a hard time getting those
  |> nice thin
  |> >> streams.  Small particles seem to just make it lumpy.
  |> My old approach
  |> >> years ago to this was a crap load of particles, low
  |> alpha value and some
  |> >> "burn" to eliminate the sprites and give it a smooth
  |> concentrated look
  |> >> when the particles get dense enough.  This worked
  |> sometimes but only on a
  |> >> small scale and could crash any machine it was on.  Any
  |> new approaches or
  |> >> thoughts to doing something like this?  I need to make
  |> quite a bit of it
  |> >> coming of a cigarette and blowing around the room.
  |> >>
  |> >> Kris
  |> >> ---
  |> >> Unsubscribe? Mail Majordomo(at)Softimage.COM with the
  |> following text in
  |> >> body:
  |> >> unsubscribe xsi
  |> >>
  |> >
  |> > --
  |> > This message has been scanned for viruses and
  |> > dangerous content by MailScanner, and is
  |> > believed to be clean.
  |> >
  |> > ---
  |> > Unsubscribe? Mail Majordomo(at)Softimage.COM with the
  |> following text in body:
  |> > unsubscribe xsi
  |>
  |> ---
  |> Unsubscribe? Mail Majordomo(at)Softimage.COM with the
  |> following text in body:
  |> unsubscribe xsi
  |>


---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi