Index of /~mccoyjo/povray/cm

	Name	Last modified	Size	Description
	Parent Directory		-
	box-dis.gif	05-Jul-2006 13:24	5.2M
	box-pm.gif	06-Jul-2006 11:51	3.6M
	box-rt.gif	06-Jul-2006 11:58	3.6M
	povray.pdf	03-Oct-2006 01:05	41K
	sphere-dis.gif	05-Jul-2006 13:04	5.2M
	sphere-pm.gif	05-Jul-2006 13:10	3.6M
	sphere-rt.gif	05-Jul-2006 13:07	3.6M
	torus-dis.gif	06-Jul-2006 12:50	5.2M
	torus-pm.gif	06-Jul-2006 12:41	5.2M
	torus-rt.gif	06-Jul-2006 12:23	5.2M

Notes on running the parapov curriculum modes scripts.

Solutions to the curriculum modules (animated gifs) can be found at:

http://cs.earlham.edu/~mccoyjo/povray/cm/

==Parallel structure of parapov==

Parapov partitions the rendering work in a workpool approach.  A
master node hands tells a slave to render an x pixel by y pixel
portion of the image.  This granularity of the work load can by
modified by adding +PWn and +PHn to the pertanent .ini file where n is
a number of pixels. The default vaules are +PW32 and +PH32.

The default granularity does not scale well over around 4 processes.
Good results have been obtained by making the granularity a quarter of
the final image size.  

When photon mapping, the master process does very little work.  A good
-np value to pass to mpi run is one more than the number of processors
available for rendering.

==Output Notes==

The output of each process is logged and placed in the working
directory of the node local to each process (i.e. parapov_n1.log).

The default output generated is a number of ppm images that are meant
to be merged into an animated gif.  Given a machine that has
ImageMagick's convert utility installed, the following command line
will produce an animated gif from the ppm images:

convert -delay 30 *.ppm animation.gif

For better looking results, more frames of animation can be added in
the appropriate ini file.  Just increase the value of the Final_Frame
parameter.

Resultant image size can be changed with the +W and +H options.

==.pov file syntax for used rendering options==

Raytracing is povray's default mode.  You only need to specify a
camera position, a light source, and scene objects for a basic
ray-traced image.

To enable photon mapping, photon blocks must be added globally and to
each object that is to interact with the photons (including having a
light sources that emit photons).  A basic, usable global photon block
looks like:


global_settings{
	photons{
		count 200000
		autostop 0
		spacing .0005
		jitter 0
	}
}

A light source that emits photons:

light_source{
	<10*clock-5,5*sin(clock*pi),0>, rgb <1,1,1>
	photons{
		reflection on
		refraction on
	}
}

An object that interacts with photons (note that to see caustics with
transparent objects, refraction must be enabled by giving the interior
block an ior value grater than 1):

sphere{< 0, 1.5, 0>, 1.0
	texture{Glass}
	photons{
		target
		reflection on
		refraction on
	}
	interior{
		dispersion 1
		ior 1.5
	}
}

Enabling dispersion (the seperating of photons of different
wavelengths with refracting through media -- aka prism effect of color
spreading) is simple: give set the dispersion value in the interior
block to a value greater than 1.  Good results happen between
dispersion values of 1.1 and 1.4.