Setting ambient accuracy (-aa) to 0

Dear all,

I intend to model the amount of solar radiation falling onto different parts of a building facade. To do this, RTRACE is used. An array of measurement points at 1-unit interval is assigned on the facade.

To model it properly, I am currently deciding on the ambient settings. My understanding is that, by setting -aa to 0, there will be no interpolation of irradiance values, and so all the backward ray-tracing will be conducted exactly at the measurement points. But I also read an advice aganist turning -aa to 0 for the purpose of evaluating irradiance value in a space (see the thread below).

Could anyone advise whether it is appropriate to set -aa to 0 in this situation? Corrections on any misconception of mine are also appreciated. Thanks!

Cheers,
Kam

Gregory J. Ward gregoryjward at gmail.com
Thu Dec 28 22:04:38 CET 2006

Previous message: [Radiance-general] ambient accuracy parameter
Next message: [Radiance-general] Modeling reflective glazing and glazing with ceramic frit
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Just wanted to chime in on this thread. The -aa 0 setting is
appropriate when you have a lot of small geometry that isn't safe to
leave out of the ambient calculation (using the -ae or -aE options),
or for special applications such as mkillum or rtcontrib. Turning
off the ambient cache is NOT a good idea when what you care about is
individual ray accuracy, such as when you are using rtrace to
evaluate radiance or irradiance values in a space. It can be done
with -aa 0, but you have to pay close attention to your other
parameter settings (such as -ad, -as, -lw, and -lr) and alter them
appropriately. The -ar setting has no effect with -aa 0. Ditto for -
aw.

-Greg

Hi Kam,

If you use -ab 1, there is no significant penalty for setting -aa 0 in a point-by-point rtrace calculation for a small number of positions. With -ab 2 and greater, setting -aa 0 may take a very long time, or require reducing the -ad and -as parameters, which will negatively impact the accuracy of your result.

-Greg

ยทยทยท

From: Kam Shing Leung <[email protected]>
Date: December 3, 2008 2:38:34 AM PST

Dear all,

I intend to model the amount of solar radiation falling onto different parts of a building facade. To do this, RTRACE is used. An array of measurement points at 1-unit interval is assigned on the facade.

To model it properly, I am currently deciding on the ambient settings. My understanding is that, by setting -aa to 0, there will be no interpolation of irradiance values, and so all the backward ray-tracing will be conducted exactly at the measurement points. But I also read an advice aganist turning -aa to 0 for the purpose of evaluating irradiance value in a space (see the thread below).

Could anyone advise whether it is appropriate to set -aa to 0 in this situation? Corrections on any misconception of mine are also appreciated. Thanks!

Cheers,
Kam

--------
Gregory J. Ward gregoryjward at gmail.com
Thu Dec 28 22:04:38 CET 2006

Just wanted to chime in on this thread. The -aa 0 setting is appropriate when you have a lot of small geometry that isn't safe to leave out of the ambient calculation (using the -ae or -aE options), or for special applications such as mkillum or rtcontrib. Turning off the ambient cache is NOT a good idea when what you care about is individual ray accuracy, such as when you are using rtrace to evaluate radiance or irradiance values in a space. It can be done with -aa 0, but you have to pay close attention to your other parameter settings (such as -ad, -as, -lw, and -lr) and alter them appropriately. The -ar setting has no effect with -aa 0. Ditto for - aw. -Greg