[Radiance-general] a bug with rtcontrib?

[Greg Ward]

Hi Pavlina,

You should set -bn 146 on your rtcontrib line, so that even if  
nothing comes back in certain bins, the number of output values is  
held constant.

Also note that the -as parameter gets overridden to 0 because it is  
incompatible with the MC technique used by rtcontrib.  Indirect  
calculation settings in general should be different from what you  
would use for rtrace.  Large values for -ad do not necessarily help  
you that much in rtcontrib, but I haven't explored the parameter  
space enough myself to really offer sound advice.

Best,
-Greg

> From: "Pavlina Akritas" <Pavlina.Akritas at arup.com>
> Date: May 22, 2008 2:26:03 AM PDT
>
> Hello everyone,
>
> I am trying to run 'rtcontrib' with the latest Radiance version to
> compute the illuminance contributions according to a Tregenza sky.
>
> The code that I am using looks like this:
>
> rtcontrib -w- -h- -I+ -ad 1000 -as 500 -ab 1 -V -b tbin -o
> results/sky/%s$i.data -m skyglow -b 0 -o results/ground/%s$i.data -m
> groundglow -f tregenza.cal model.oct < grid/grid.grd
>
> The code works fine when the grid used points up (+z). However, I  
> would
> like to find the illuminance on a wall grid, thus the grid points must
> point in either +y, -y, +x or -x. When I run 'rtcontrib' in this case
> the number of columns varies between 438 and 435 (since rtcontrib
> outputs r g b values). Thus, 'rtcontrib' seems to be skipping a patch
> every now and then. I have tried also to increasing the settings  
> but the
> same problem occured.
>
> Is it possible to resolve this problem somehow so that the number of
> columns is consistent all the time (i.e. 438 including the ground)? If
> yes, how?
>
> Thank you,
>
> Pavlina Akritas