I was fixing my wind algorithm over the weekend, and decided to try some smoke.
Instead of the usual media filled spheres, or cunnngly textured spheres, I
decided to try and use _lots_ of small spheres, just to see how it looked.

There are about 1800 particles in any one frame (except the first few, of
course), being blown about by a very turbulent wind. The texture is simply rgb
0.5 with a transmit component based on particle age.

Doing smoke this way is slow and not terribly effective; I think many more
particles are needed to pull the effect off. See what you think.

MJL
--
prism{0,.1,30#local I=1;#while(I<30)#local B=asc(substr(// Mark James Lewin
"#K?U_u`V[RG>3<9DGPL.0EObkcPF'",I,1))-33;<div(B,10)-4mod(B,10)+5*div(I,21)-
6>#local I=I+1;#end,-4pigment{rgb 9}rotate-x*90translate 15*z}//POV-Ray 3.5

Post a reply to this message

Attachments:
Download 'smoke_3.mpg' (358 KB)

I like it - it looks like a jet of liquid diffusing into another liquid, so
it may turn out to be useful.
maybe i'm missing something somewhere, but I fail to see how it's slower
than using media

jim

Post a reply to this message

Looks very similar to steam coming out of a cooling tower.

Post a reply to this message

Nice! At the bottom you can really see the spheres, but the effect at the
top is much more realistic.
"Mark James Lewin" <m_j### [at] yahoocomau> wrote in message
news:3C95105D.BA4A1E76@yahoo.com.au...
> I was fixing my wind algorithm over the weekend, and decided to try some
smoke.
> Instead of the usual media filled spheres, or cunnngly textured spheres, I
> decided to try and use _lots_ of small spheres, just to see how it looked.
>
> There are about 1800 particles in any one frame (except the first few, of
> course), being blown about by a very turbulent wind. The texture is simply
rgb
> 0.5 with a transmit component based on particle age.
>
> Doing smoke this way is slow and not terribly effective; I think many more
> particles are needed to pull the effect off. See what you think.
>
> MJL
> --
> prism{0,.1,30#local I=1;#while(I<30)#local B=asc(substr(// Mark James
Lewin
>
"#K?U_u`V[RG>3<9DGPL.0EObkcPF'",I,1))-33;<div(B,10)-4mod(B,10)+5*div(I,21)-
> 6>#local I=I+1;#end,-4pigment{rgb 9}rotate-x*90translate 15*z}//POV-Ray
3.5
>
>

Post a reply to this message

I liked it VERY much! I wonder if the effect would be enhanced if the
spheres had some kind of slight transparency, and using more of them.

Congratulations, you made me loop through your animation dozens of times!!!

Fernando.

"Mark James Lewin" <m_j### [at] yahoocomau> wrote in message
news:3C95105D.BA4A1E76@yahoo.com.au...
> I was fixing my wind algorithm over the weekend, and decided to try some
smoke.
> Instead of the usual media filled spheres, or cunnngly textured spheres, I
> decided to try and use _lots_ of small spheres, just to see how it looked.
>
> There are about 1800 particles in any one frame (except the first few, of
> course), being blown about by a very turbulent wind. The texture is simply
rgb
> 0.5 with a transmit component based on particle age.
>
> Doing smoke this way is slow and not terribly effective; I think many more
> particles are needed to pull the effect off. See what you think.
>
> MJL
> --
> prism{0,.1,30#local I=1;#while(I<30)#local B=asc(substr(// Mark James
Lewin
>
"#K?U_u`V[RG>3<9DGPL.0EObkcPF'",I,1))-33;<div(B,10)-4mod(B,10)+5*div(I,21)-
> 6>#local I=I+1;#end,-4pigment{rgb 9}rotate-x*90translate 15*z}//POV-Ray
3.5
>
>

Post a reply to this message

jimbobjim wrote:

> I like it - it looks like a jet of liquid diffusing into another liquid, so
> it may turn out to be useful.

Thanks. You've given me an idea...

>
> maybe i'm missing something somewhere, but I fail to see how it's slower
> than using media

Reading data for 2000+ particles from file, calculating their next positions, and
writing them back to file is not a fast process. There comes a point where the
time needed for parsing thousands of particles becomes greater than rendering a
few dozen media filled spheres. And unfortunately, this type of smoke sim needs
_really_ high particle counts.

MJL

--
text{ttf"timrom.ttf"concat(#local O=1;#while(O<7)chr(val(substr(concat(#local Q=
1;#while(Q<7)str(asc(substr("???<?>",Q,1))-56,0,0),#local Q=Q+1;#end""),O,2))),#
local O=O+2;#end"").1,0pigment{rgb 9}translate-<1,.3,-2>} // MJL

Post a reply to this message

Bill DeWitt wrote:

>     Looks very similar to steam coming out of a cooling tower.

I guess it does. Maybe a texture change will enhance this effect. Thanks for the
comments!

MJL

--
text{ttf"timrom.ttf"concat(#local O=1;#while(O<7)chr(val(substr(concat(#local Q=
1;#while(Q<7)str(asc(substr("???<?>",Q,1))-56,0,0),#local Q=Q+1;#end""),O,2))),#
local O=O+2;#end"").1,0pigment{rgb 9}translate-<1,.3,-2>} // MJL

Post a reply to this message

Mahalis wrote:

> Nice! At the bottom you can really see the spheres, but the effect at the
> top is much more realistic.

Thanks! As with most particle simulations, more particles is almost always
desired.

MJL

--
text{ttf"timrom.ttf"concat(#local O=1;#while(O<7)chr(val(substr(concat(#local Q=
1;#while(Q<7)str(asc(substr("???<?>",Q,1))-56,0,0),#local Q=Q+1;#end""),O,2))),#
local O=O+2;#end"").1,0pigment{rgb 9}translate-<1,.3,-2>} // MJL

Post a reply to this message

"Fernando González del Cueto" wrote:

> I liked it VERY much! I wonder if the effect would be enhanced if the
> spheres had some kind of slight transparency, and using more of them.

Thanks! The spheres get more transperent as they age, but I think they needed
more from the outset. The need for more particles is directly proportional to the
need for a faster computer, and unfortunately, 1800 particles tests my patience
as it is.

MJL

--
text{ttf"timrom.ttf"concat(#local O=1;#while(O<7)chr(val(substr(concat(#local Q=
1;#while(Q<7)str(asc(substr("???<?>",Q,1))-56,0,0),#local Q=Q+1;#end""),O,2))),#
local O=O+2;#end"").1,0pigment{rgb 9}translate-<1,.3,-2>} // MJL

Post a reply to this message

> 1800 particles tests my patience

I don't think that sounds like a whole lot.. That shouldn't take more than a
second to parse / recalculate..?

Regards,
Hugo

Post a reply to this message

Port the script to C or Perl, and you'll surprised how fast your computer
can calculate the animation data.

--
Apache
http://geitenkaas.dns2go.com/experiments/
apa### [at] yahoocom

Post a reply to this message

"Mark James Lewin" <m_j### [at] yahoocomau> wrote
> Reading data for 2000+ particles from file, calculating their next
positions, and
> writing them back to file is not a fast process. There comes a point where
the
> time needed for parsing thousands of particles becomes greater than
rendering a
> few dozen media filled spheres. And unfortunately, this type of smoke sim
needs
> _really_ high particle counts.
>
yeah - after I posted I started writing my own version and it's kinda-sorta
slow and doesn't look as good as yours
jim

Post a reply to this message

Apache wrote:

> Port the script to C or Perl, and you'll surprised how fast your computer
> can calculate the animation data.

Unfortunately, I don'y know either of these languages :-( Perhaps one day...

MJL

--
prism{0,.1,30#local I=1;#while(I<30)#local B=asc(substr(// Mark James Lewin
"#K?U_u`V[RG>3<9DGPL.0EObkcPF'",I,1))-33;<div(B,10)-4mod(B,10)+5*div(I,21)-
6>#local I=I+1;#end,-4pigment{rgb 9}rotate-x*90translate 15*z}//POV-Ray 3.5

Post a reply to this message

Hugo wrote:

> > 1800 particles tests my patience
>
> I don't think that sounds like a whole lot.. That shouldn't take more than a
> second to parse / recalculate..?
>
> Regards,
> Hugo

I am using pov script here. I would use a faster language if I knew how to
program in it. Ideally I would make my own patch for pov, but this is well beyond
my capabilities.

MJL


--
prism{0,.1,30#local I=1;#while(I<30)#local B=asc(substr(// Mark James Lewin
"#K?U_u`V[RG>3<9DGPL.0EObkcPF'",I,1))-33;<div(B,10)-4mod(B,10)+5*div(I,21)-
6>#local I=I+1;#end,-4pigment{rgb 9}rotate-x*90translate 15*z}//POV-Ray 3.5

Post a reply to this message

You need more particles <G>

Seriously, if you want me to take a crack at a particle plug in for Moray, I
might be able to convert it to C++.

Also, I wonder if this could be used for 'heat shimmer'. I was trying to
duplicate something a while ago with a hot exhaust that doesn't have smoke;
I was using spheres with some IOR set to them.  Do you think this would
work?

                  == John ==

"Mark James Lewin" <m_j### [at] yahoocomau> wrote in message
news:3C95105D.BA4A1E76@yahoo.com.au...
> I was fixing my wind algorithm over the weekend, and decided to try some
smoke.
> Instead of the usual media filled spheres, or cunnngly textured spheres, I
> decided to try and use _lots_ of small spheres, just to see how it looked.
>
> There are about 1800 particles in any one frame (except the first few, of
> course), being blown about by a very turbulent wind. The texture is simply
rgb
> 0.5 with a transmit component based on particle age.
>
> Doing smoke this way is slow and not terribly effective; I think many more
> particles are needed to pull the effect off. See what you think.
>
> MJL
> --
> prism{0,.1,30#local I=1;#while(I<30)#local B=asc(substr(// Mark James
Lewin
>
"#K?U_u`V[RG>3<9DGPL.0EObkcPF'",I,1))-33;<div(B,10)-4mod(B,10)+5*div(I,21)-
> 6>#local I=I+1;#end,-4pigment{rgb 9}rotate-x*90translate 15*z}//POV-Ray
3.5
>
>

Post a reply to this message

"John D. Gwinner" wrote:

> Seriously, if you want me to take a crack at a particle plug in for Moray, I
> might be able to convert it to C++.

I don't use Moray, but thanks for the offer. At this stage though, I'm not
satisfied with the behaviour of some aspects of my system.

> Also, I wonder if this could be used for 'heat shimmer'. I was trying to
> duplicate something a while ago with a hot exhaust that doesn't have smoke;
> I was using spheres with some IOR set to them.  Do you think this would
> work?

Possibly. It might be slow. If you only wanted the 'shadow' of the heat, you
could try no_image spheres. It'll be faster.

MJL

--
prism{0,.1,30#local I=1;#while(I<30)#local B=asc(substr(// Mark James Lewin
"#K?U_u`V[RG>3<9DGPL.0EObkcPF'",I,1))-33;<div(B,10)-4mod(B,10)+5*div(I,21)-
6>#local I=I+1;#end,-4pigment{rgb 9}rotate-x*90translate 15*z}//POV-Ray 3.5

Post a reply to this message