[Radiance-general] ambient accuracy parameter

[Mark Stock]

Marija,

I use -aa 0 for models which contain massive amounts of detail 
and for which computing and storing enough ambient points would 
quickly run me out of memory. The first scene to require this was

http://mark.technolope.org/image/p36.html

but I have used it on most of my work since then.

I did a little work with -aa 0 before this and found that 
oversampling combined with low -ad could produce good images. I 
made a page showing the results:

http://mark.technolope.org/radmisc/aa0_ps1_test/final.html

I hope this helps you.

Mark
mstock at umich.edu

On Wed, 27 Dec 2006, Marija Cvetkovic wrote:

> Hi Martin,
>
> You are talking about ambient resolution (-ar) parameter. For -aa images  are 
> better with lower settings of -aa, and maximum recommended value is 0.3.
> I know what this parameter means for rendering, but my question is when value 
> -aa 0 should be used and when not?
>
> Marija
>
> Martin Moeck wrote:
>> I use the -aa setting to get a smoother indirect calculation. This is 
>> especially important if the -ad parameter is low, i.e., <=512. A high -aa 
>> parameter, such as 5000 or whatever, creates a dense grid of 
>> indirect/reflected values. You can check that by making two images, one 
>> with a low setting of ad and aa, and one with a low setting of ad and a 
>> high setting of aa. That will show you immediately what the ambient 
>> accuracy parameter is about. The image contains less blotches. 
>> Martin Moeck
>> 
>> 
>> -----Original Message-----
>> From: radiance-general-bounces at radiance-online.org on behalf of Marija 
>> Cvetkovic
>> Sent: Tue 12/26/2006 2:30 AM
>> To: Radiance general
>> Subject: [Radiance-general] ambient accuracy parameter
>>  Hi,
>> 
>> I've been reading some previous posts related to ambient parameters 
>> settings and I have some questions.
>> 
>> When it is recommendable to use -aa 0 value? If I understand this parameter 
>> well, 0 value turns off irradiance caching, so we can use it when caching 
>> isn't needed (mkillum calculation for venetian blinds or rtcontrib).
>> When we set -aa 0, does -ar have influence on minimum sampling distance, 
>> since: min sampling distance = max scene dimension* aa/ar.
>> 
>> When I run a rtrace calculation with -aa 0, calculation time is much 
>> shorter than with -aa 0.1. Why?
>> 
>
>
> _______________________________________________
> Radiance-general mailing list
> Radiance-general at radiance-online.org
> http://www.radiance-online.org/mailman/listinfo/radiance-general
>
>
>