Sky component only

radiance-general
1

Thomas,

I used mkillum yesterday and bingo!!! got some better results, that I am trying to validate.
One question though, I just need to be certain that there is no ambient light in my results.

so to check, I am using the following for mkillum and then rtrace:

mkillum -I -h -dp 2046 -ar 32 -ms 0.063 -ds .2 -dt .05 -dc .75 -dr 3 -sj 1 -st .01 -ab 0 -aa .1 -ad 512 -as 256 -av 0 0 0 -lr 12 -lw .0005 -af scene2.amb scene2.oct < window.rad > mkwin2.rad

rtrace -I -h -dp 2046 -ar 32 -ms 0.063 -ds .2 -dt .05 -dc .75 -dr 3 -sj 1 -st .01 -ae Zero_Ref -ab 0 -aa .1 -ad 512 -as 256 -av 0 0 0 -lr 12 -lw .0005 -af scene2.amb mkscene2.oct < mk1.pts > existing3.dat

I assume that by setting -ab 0 and -av 0 0 0 there should be no ambient calculation, and indeed, lookamb just returns the values from rtrace, rather than a data file for value, position etc.. I believe this is a good thing.

Your thoughts would be greatly apprecicated however.

many thanks

nick

Thomas,
thanks for that. I had tried it in the past but rtrace returns
values of zero!!
I was confused about this, until I remember that for a uniform, or overcast sky (the types I am testing), the sky is a glow source, which I believe is not tested in the direct calculation.
which means I am back to the drawing board.

You could probably replace your windows with illums
using mkillum and a higher -ab value. For your final
rtrace calculations you can use -ab 0 again and only
get the "direct" contribution of the window illum.

I have to admit that I don't know what causes your
problems and therefore this approche may fail as well.

Thomas

···

On 22.09.2005, at 13:19, nick devlin wrote:

2

Thomas,

I used mkillum yesterday and bingo!!! got some better results, that I am trying to validate.
One question though, I just need to be certain that there is no ambient light in my results.

so to check, I am using the following for mkillum and then rtrace:

mkillum -I -h -dp 2046 -ar 32 -ms 0.063 -ds .2 -dt .05 -dc .75 -dr 3 -sj 1 -st .01 -ab 0 -aa .1 -ad 512 -as 256 -av 0 0 0 -lr 12 -lw .0005 -af scene2.amb scene2.oct < window.rad > mkwin2.rad

First I thought you should use ab > 0 for mkillum but this would add
the indirect contribution of nearby buildings, which is probably
not what you want.

I'd reduce the value of -dp (you don't have any virtual sources if
you're only interested in the sky) and increase the value of -ad
to get a very good sampling of the hemisphere (4096 or even more).

What's -ms for? Are you calculating paticipating media?

rtrace -I -h -dp 2046 -ar 32 -ms 0.063 -ds .2 -dt .05 -dc .75 -dr 3 -sj 1 -st .01 -ae Zero_Ref -ab 0 -aa .1 -ad 512 -as 256 -av 0 0 0 -lr 12 -lw .0005 -af scene2.amb mkscene2.oct < mk1.pts > existing3.dat

I wouldn't use the same ambient file for the mkillum and
the rtrace process. One (mkillum) contains the outside,
the other only the inside of your room geometry.

I don't think any of -sj, -st, -lr, -lw have much influence
on the calculation given your sky-only-setup.

I hope someone with more experience with mkillum and
rtrace can give you better hints. I'm just speculating
here.

Thomas

···

On 23.09.2005, at 12:29, nick devlin wrote: