Illuminance calculation on a virtual surface?

Hi-

I have been using a script with rtrace to calculate the illuminance
values on a grid of points representing the work surface in my modeled
room. But now I would like to generate a false-color image of the
illuminance values on a plane .8m above the floor. I can easily do
the floor (or walls, etc.) with a parallel projection view, but I
don't see how to get the workplane illuminance values without creating
an opaque surface at the correct height which would change the light
distribution. Does anyone know of a way to specify a "virtual" surface
or other technique to do this?

Thanks,
Jim Estes

Hello Jim,

You can simply slightly change your script by specifying rtrace option -fac that will directly generate a Radiance picture on its output. Of course, you will surely have to increase the density of the calculation points piped to rtrace in order to have a picture with reasonable size.

Raphael Compagnon

···

At 27.06.2004 23:02, you wrote:

Hi-

I have been using a script with rtrace to calculate the illuminance
values on a grid of points representing the work surface in my modeled
room. But now I would like to generate a false-color image of the
illuminance values on a plane .8m above the floor. I can easily do
the floor (or walls, etc.) with a parallel projection view, but I
don't see how to get the workplane illuminance values without creating
an opaque surface at the correct height which would change the light
distribution. Does anyone know of a way to specify a "virtual" surface
or other technique to do this?

Thanks,
Jim Estes

Raphael Compagnon wrote:

Hello Jim,

You can simply slightly change your script by specifying rtrace option -fac that will directly generate a Radiance picture on its output. Of course, you will surely have to increase the density of the calculation points piped to rtrace in order to have a picture with reasonable size.

How about making the plane out of a trans material with zero reflectance and 100% transmission? Martin (Moeck), didn't you mention that technique once? Or am I confused again? =8-)

···

----

      Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

No good. Trans gets ignored by the rpict -i option. Just use pfilt to scale the illuminance map if it's not big enough. (John M's suggestion -- he's sitting next to me in Leicester.)

-Greg

···

From: Rob Guglielmetti <[email protected]>
Date: June 28, 2004 3:12:13 PM GMT+01:00

Raphael Compagnon wrote:

Hello Jim,
You can simply slightly change your script by specifying rtrace option -fac that will directly generate a Radiance picture on its output. Of course, you will surely have to increase the density of the calculation points piped to rtrace in order to have a picture with reasonable size.

How about making the plane out of a trans material with zero reflectance and 100% transmission? Martin (Moeck), didn't you mention that technique once? Or am I confused again? =8-)

----

     Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

Hi Greg, Hi John,

Greg Ward wrote:

No good. Trans gets ignored by the rpict -i option. Just use pfilt to scale the illuminance map if it's not big enough. (John M's suggestion -- he's sitting next to me in Leicester.)

Oh yeah? =8-p

vwrays -x XRES -y YRES -vf viewfile -fd | rtrace -h -fd -opn octree \

rtrace -fdc -I render_options -x XRES -y YRES octree > illum_picture.pic

Greg, does this look familiar? A little ditty from a year or so ago? You sent me this tip when I asked you how to get illuminance on a building's curtain wall. The first rtrace computes the intersection point (which is fast) and the second rtrace does the illuminance calculation. Cool, yes? Yes.

Now, given this little tip, could he use trans as I described?

···

----

      Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

Rob Guglielmetti wrote:

Hi Greg, Hi John,

Greg Ward wrote:

No good. Trans gets ignored by the rpict -i option. Just use pfilt to scale the illuminance map if it's not big enough. (John M's suggestion -- he's sitting next to me in Leicester.)

Thanks, that should get me there. Also thanks to Raphael for pointing out
the obvious that I missed in the rtrace man page for computing an image
(using the "-fac -x and -y" options, so far it is generating bad picture
files, but I'm on the right track).

vwrays -x XRES -y YRES -vf viewfile -fd | rtrace -h -fd -opn octree \
> rtrace -fdc -I render_options -x XRES -y YRES octree > illum_picture.pic

Greg, does this look familiar? A little ditty from a year or so ago? You sent me this tip when I asked you how to get illuminance on a building's curtain wall. The first rtrace computes the intersection point (which is fast) and the second rtrace does the illuminance calculation. Cool, yes? Yes.

Now, given this little tip, could he use trans as I described?

Thanks. That looks promising. My next step was to create a finer grid for
input to rtrace anyway. If this technique works not only would it do that
but hopefully (maybe) render the image much faster also. It might take a
little while but if I come up with anything interesting I'll report back.

Jim Estes

OK, Rob -- you got me. Your little trick would work. I don't think he
really needs it, though, since he already has a method for generating
the grid points for rtrace. He can simply increase the resolution of
the grid he already has (as Raphael suggested) and/or upsample the
result using pfilt. Or your method. All will work.

-Greg

Quoting Rob Guglielmetti <[email protected]>:

···

Hi Greg, Hi John,

Greg Ward wrote:

> No good. Trans gets ignored by the rpict -i option. Just use pfilt to
> scale the illuminance map if it's not big enough. (John M's suggestion
> -- he's sitting next to me in Leicester.)

Oh yeah? =8-p

vwrays -x XRES -y YRES -vf viewfile -fd | rtrace -h -fd -opn octree \
> rtrace -fdc -I render_options -x XRES -y YRES octree > illum_picture.pic

Greg, does this look familiar? A little ditty from a year or so ago?
You sent me this tip when I asked you how to get illuminance on a
building's curtain wall. The first rtrace computes the intersection
point (which is fast) and the second rtrace does the illuminance
calculation. Cool, yes? Yes.

Now, given this little tip, could he use trans as I described?