This particle_system thing has memory problems, I think. Now, I'm not a
programmer, but it seems to me that as time goes by while I'm doing
animations something piles up somewhere and messes things up. If I stop it,
and restart it, things work fine. This doesn't seem right. Here's the code.
The problem occured on frame 16 of 99; paricles flying all over the place.
When I continued from 16, it worked fine.

// Persistence of Vision Ray Tracer Scene Description File
#version unofficial MegaPov 0.5;
#include "colors.inc"
//Start of particle pattern demo file
camera {
 location <-3, 2.5, -8>*3
 angle 35
 look_at < 0, 2, 0>
}
light_source {< 20, 75, 50>*5 color White*0.5}
light_source {< 0, 2.5, -8>*5 color White*0.3}
light_source {< -75, 150,-50> color White}
#declare Floor =
union {
 box {< -100, 0,-100>, < 100, 0, 100>
  texture {
   pigment {color rgb < 1, 1, 0.8>}
//   pigment {checker color Black transmit 0.5, color White transmit 0.5}
  }
 }
 box {<-2, 0,-2>, < 2, 4, 2>
  texture {
   pigment {color Green}
  }
 }
// sphere {<-1.5, 3.5, 1>, 2
//  texture {
//   pigment {color Blue}
//  }
// }
}
object {Floor}
particle_system {
 wind < 0, 0, 0>
 gravity < 0,-0.00981, 0>
 iterations 1000+int(400*clock)*10
 time 500*clock
// time 40*clock + 30
 jitter 0
 drag 0
 environment object {Floor}
 threshold 1.5

 particle_emitter {
  location < 0, 4.1, 0>
//  object {sphere {<-5.5, 1, 0>, 1}}
  radius 0.3
  lifespan 10000
  rate 100
  elasticity 0.5
  direction < 0, .2, 0>
//  direction < 1, 1, 0>
  angle 12, 16
 }
 texture {
  pigment {color rgb < 0.7, 0.75, 1>}
 }
}

Post a reply to this message

In article <3971fe09@news.povray.org>, "Tony[B]" 
<ben### [at] panamac-comnet> wrote:

> This particle_system thing has memory problems, I think. Now, I'm not 
> a programmer, but it seems to me that as time goes by while I'm doing 
> animations something piles up somewhere and messes things up. If I 
> stop it, and restart it, things work fine. This doesn't seem right. 

A memory leak somewhere? How does it mess things up? What exactly do you 
do when you restart it? I mean, do you quit the program and start it up 
again, or do you just stop the render?


> Here's the code.
> The problem occured on frame 16 of 99; paricles flying all over the 
> place. When I continued from 16, it worked fine.

What do you mean by "particles flying all over the place"? Were there 
more particles than there should have been, or was there the proper 
number of particles, just scattered all over?

I will try out your code later, though I don't see any obvious problems 
with it.

-- 
Christopher James Huff - Personal e-mail: chr### [at] maccom
TAG(Technical Assistance Group) e-mail: chr### [at] tagpovrayorg
Personal Web page: http://homepage.mac.com/chrishuff/
TAG Web page: http://tag.povray.org/

Post a reply to this message

I have this same problem. I always assumed it was the AMD chip I'm using,
versus Intel. That is to say, I get this with any build of MegaPatch. I
never see it with the official Pov, but most of the animations I do use MP
syntax. I also get it following repeated renderings of simple scenes using
radiosity (render, make changes, render, make changes, lather-rinse-repeat).
The computer simply stops rendering, makes glacial progress on a scene it
just rendered quickly, or crashes to the desktop altogether.

Bummer man! (reboot) <sigh>

Grim

Post a reply to this message

I've had similar problems.  Here's a snippet from an e-mail where I was
discussing this with another user.

<-----Begin----->
About 1 out of every 5 or 6 renders, the file will parse properly and the
render window will open, but then it just stops.  Breaking the render prints an
error about "no IDATs written to file."  I guess it's a png initialization
problem.

I've also had it do screwy things where the render is very slow and the
particles seem like they are doubled.

Also changing apparently unrelated code (e.g. camera angle statement) can cause
an error in "Minverse".

In all cases, simply trying the render again without changing anything seems to
work ok.

I also discovered that specifying a pigment pattern for temperature in the
particle_system block will cause a system crash unless you specify a color_map
to go with it.

Mass seems backwards to me.  Making the particles heavier seems to make them
more buoyant.

<-----End----->
Mike Wilson

"Tony[B]" wrote:

> This particle_system thing has memory problems, I think. Now, I'm not a
> programmer, but it seems to me that as time goes by while I'm doing
> animations something piles up somewhere and messes things up. If I stop it,
> and restart it, things work fine. This doesn't seem right. ...

Post a reply to this message

In article <3972a489@news.povray.org>, "GrimDude" 
<vos### [at] yahoocom> wrote:

> I have this same problem. I always assumed it was the AMD chip I'm 
> using, versus Intel. That is to say, I get this with any build of 
> MegaPatch. I never see it with the official Pov, but most of the 
> animations I do use MP syntax. I also get it following repeated 
> renderings of simple scenes using radiosity (render, make changes, 
> render, make changes, lather-rinse-repeat).
> The computer simply stops rendering, makes glacial progress on a scene it
> just rendered quickly, or crashes to the desktop altogether.

I have had trouble with the "glacial slowness" you mentioned...it occurs 
with the plain MegaPOV too? I usually have to quit and start up 
MegaPOVPlus again. Sometimes it is *very* tenacious, and just won't 
leave. Gremlins? :-)
I think it may be a problem with bounding...maybe some variable doesn't 
get initialized and starts up with random values. I got a similar 
slowdown by completely turning off bounding, but that may be 
cooincidence.

-- 
Christopher James Huff - Personal e-mail: chr### [at] maccom
TAG(Technical Assistance Group) e-mail: chr### [at] tagpovrayorg
Personal Web page: http://homepage.mac.com/chrishuff/
TAG Web page: http://tag.povray.org/

Post a reply to this message

In article <39730597.FC0ED7E0@iastate.edu>, Mike Wilson 
<maw### [at] iastateedu> wrote:

> I've also had it do screwy things where the render is very slow and the
> particles seem like they are doubled.

What do you mean by "doubled"? Are there twice as many particles as 
there should be? Are they twice the size they should be?


> Also changing apparently unrelated code (e.g. camera angle statement) 
> can cause an error in "Minverse".

Interesting...are you using particle_stretch? I think I need to add some 
error checking to it, because it won't work with particles that aren't 
moving...but particles usually are moving.


> In all cases, simply trying the render again without changing 
> anything seems to work ok.

Has it ever happened in an animation?
I have encountered problems with the slow rendering, but never when the 
first render went fine. It seems like if one render goes correctly, all 
of them will until you quit the program...


> I also discovered that specifying a pigment pattern for temperature 
> in the particle_system block will cause a system crash unless you 
> specify a color_map to go with it.

Sounds like a possible problem with the pigment pattern...though I will 
check the temperature pigment, just in case.


> Mass seems backwards to me.  Making the particles heavier seems to 
> make them more buoyant.

That is because the gravitational force doesn't take mass into 
account...I will fix that. Gravitational acceleration should be 
constant, but currently, gravitational force is constant. Mass should 
behave properly with other forces, though, and gravity works fine on 
particles with a mass of 1.

-- 
Christopher James Huff - Personal e-mail: chr### [at] maccom
TAG(Technical Assistance Group) e-mail: chr### [at] tagpovrayorg
Personal Web page: http://homepage.mac.com/chrishuff/
TAG Web page: http://tag.povray.org/

Post a reply to this message

Chris:

Thanks for providing this information.

Here are some more details from my end:

The particles seem doubled in size, or more specifically, perhaps, stretched
in the direction of the initial velocity.

I'm not using particle_stretch.  Whenever I start playing around with
something new, I usually comment out everything I can and just change one or
two things at a time and I hadn't tried particle_stretch yet.

I haven't tried to do an animation with your patch yet.

If it makes any difference, I'm running the Windows version and have had the
same things show up under Win95 and NT 4.0.

Best,

Mike Wilson

Chris Huff wrote:

> In article <39730597.FC0ED7E0@iastate.edu>, Mike Wilson
> <maw### [at] iastateedu> wrote:
>
> > I've also had it do screwy things where the render is very slow and the
> > particles seem like they are doubled.
>
> What do you mean by "doubled"? Are there twice as many particles as
> there should be? Are they twice the size they should be?
>
> > Also changing apparently unrelated code (e.g. camera angle statement)
> > can cause an error in "Minverse".
>
> Interesting...are you using particle_stretch? I think I need to add some
> error checking to it, because it won't work with particles that aren't
> moving...but particles usually are moving.
>
> > In all cases, simply trying the render again without changing
> > anything seems to work ok.
>
> Has it ever happened in an animation?
> I have encountered problems with the slow rendering, but never when the
> first render went fine. It seems like if one render goes correctly, all
> of them will until you quit the program...
>
> > I also discovered that specifying a pigment pattern for temperature
> > in the particle_system block will cause a system crash unless you
> > specify a color_map to go with it.
>
> Sounds like a possible problem with the pigment pattern...though I will
> check the temperature pigment, just in case.
>
> > Mass seems backwards to me.  Making the particles heavier seems to
> > make them more buoyant.
>
> That is because the gravitational force doesn't take mass into
> account...I will fix that. Gravitational acceleration should be
> constant, but currently, gravitational force is constant. Mass should
> behave properly with other forces, though, and gravity works fine on
> particles with a mass of 1.
>
> --
> Christopher James Huff - Personal e-mail: chr### [at] maccom
> TAG(Technical Assistance Group) e-mail: chr### [at] tagpovrayorg
> Personal Web page: http://homepage.mac.com/chrishuff/
> TAG Web page: http://tag.povray.org/

Post a reply to this message