I discovered what was really slowing down my render, it's the part where my
isosurface intersects my media container. Is there a generalized reason for
this? or is it something that might be particular to my code which I am
embarrassed to show?

It slows down by a factor of 100 or so.

Post a reply to this message

In article <38b707b7@news.povray.org>, "Bill DeWitt" 
<the### [at] earthlinknet> wrote:

> I discovered what was really slowing down my render, it's the part 
> where my isosurface intersects my media container. Is there a 
> generalized reason for this? or is it something that might be 
> particular to my code which I am embarrassed to show?

Do you mean your media container is an intersection with an isosurface? 
Or that your media container and the isosurface are occupying the same 
region of space?
Does it occur with the media turned off? What shape is the container? 
What trace level is it going up to(this should be in the rendering 
statistics)?

-- 
Chris Huff
e-mail: chr### [at] yahoocom
Web page: http://chrishuff.dhs.org/

Post a reply to this message

"Chris Huff" <chr### [at] yahoocom> wrote :
>
> Do you mean your media container is an intersection with an isosurface?

    No,

> Or that your media container and the isosurface are occupying the same
> region of space?

    Right, I have an isosurface which is sitting on top of and projecting
down into a media container.

> Does it occur with the media turned off?

    No.

> What shape is the container?

    Box.

> What trace level is it going up to(this should be in the rendering
> statistics)?

    Trace level? Under trace there is a "Quality" listing that says "9"...

Post a reply to this message

In article <38b7e1ea@news.povray.org>, "Bill DeWitt" 
<the### [at] earthlinknet> wrote:

> > Or that your media container and the isosurface are occupying the same
> > region of space?
> 
>     Right, I have an isosurface which is sitting on top of and projecting
> down into a media container.

What does that mean, "projecting down into"?
Do you have a light source on the opposite side of the isosurface, so 
it's shadow falls on the media?


> > Does it occur with the media turned off?
> 
>     No.

Ok, it is probably due to the many times a ray has to be shot at the 
isosurface from the media to compute where the shadow is. But you said 
it only happens in the area where the isosurface overlaps with the media 
container...have you tried other objects which tend to be slow, like 
julia_fractal? Is the isosurface transparent? If so, is it hollow?


> > What trace level is it going up to(this should be in the rendering
> > statistics)?
> 
>     Trace level? Under trace there is a "Quality" listing that says "9"...

No, the the maximum level to which POV traced. The max_trace_level 
keyword sets the upper limit to this, and the maximum level actually 
computed is in the rendering statistics. If there is some total internal 
reflection, that might be using a very large number and eating up 
processor time, but I don't think that is the problem.

-- 
Chris Huff
e-mail: chr### [at] yahoocom
Web page: http://chrishuff.dhs.org/

Post a reply to this message

"Chris Huff" <chr### [at] yahoocom> wrote :
>
> What does that mean, "projecting down into"?

    Well, actually it means what I said the first time, the isosurface
intersects the media container. They overlap. They share some points in
space. Part of the isosurface is above the media container and part of it is
inside the media container. There is a volume which is isosurface, a volume
which is media container and a volume which could be said to be both
isosurface -and- media container.

    So as not to give away exactly what I am doing, although I do appreciate
you giving it some thought, imagine that there is a foot dangling into
water, the foot is my isosurface, the water is the media. That part of the
image in which you can see part of the foot under the water is the slow
part.

> The max_trace_level
>

    Ah, default. No reflection, and only the transparency of the media
container. It's down to about one point per 10 minutes. Fortunately it is
only a dozen points across and 8 -10 lines deep.

Post a reply to this message