| 20 |
|
sources corresponding to bright, concentrated regions in |
| 21 |
|
the given ambient environment. |
| 22 |
|
Any local geometry is ignored in the input octree, which |
| 23 |
< |
should be derived from a captured light probe. |
| 23 |
> |
should be derived from a captured light probe and modeled |
| 24 |
> |
as a distant hemispherical or spherical glow source. |
| 25 |
|
The output sources may then be combined with this environment |
| 26 |
|
to produce a more efficient scene for rendering, faster and |
| 27 |
|
less prone to sampling artifacts. |
| 30 |
|
.I \-d |
| 31 |
|
option may be used to specify the number of ray samples, which |
| 32 |
|
defaults to 262,000. |
| 33 |
< |
Calculation time is roughly proportional to accuracy, and the |
| 34 |
< |
default is fine enough to find sources the size of the sun |
| 33 |
> |
Calculation time is roughly proportional to this setting, and the |
| 34 |
> |
default is fine enough to resolve sources the size of the sun |
| 35 |
|
(half a degree) or larger. |
| 36 |
|
.PP |
| 37 |
|
The |
| 38 |
|
.I \-t |
| 39 |
|
option may be used to manually set the radiance threshold for |
| 40 |
|
sources, in watts/sr/meter^2. |
| 41 |
< |
The default uses the top 1 percentile of the environment, which |
| 41 |
> |
The default uses the top 2 percentile of the environment, which |
| 42 |
|
is usally a good value. |
| 43 |
|
.PP |
| 44 |
|
The |
