Graphics Reference
In-Depth Information
subtended by the small sphere is therefore
r 2
π π
m (Ω) = 4
(26.48)
4
π
R 2
= π
r 2
R 2
.
(26.49)
Substituting this in Equations 26.47 and 26.43, we get
r 2
R 2
R 2 L π
4
π
, so
(26.50)
Φ
L =
r 2 ) .
(26.51)
4
π
(
π
P
We've made two approximations in this computation: that the dot product is
always 1, and that the area occluded by the emitting sphere is
Q
r 2 . If you instead
evaluate the integral exactly, you'll see that the two approximations exactly cancel
each other.
π
R
H
r
We'll now consider a similar example and analysis. This time we have a
small disk-shaped emitter of radius r , that emits light only on one side (see
Figure 26.28).
We enclose it in a hemisphere H of radius R , and first compute the power
density at the North Pole P just as before; once again the power density is
Figure 26.28: A Lambertian emit-
ting disk that radiates on one side
only.
r 2
R 2
D P = Lm (Ω) = L π
.
(26.52)
At a point like Q that's off-axis by the angle
φ
, the Tilting principle applies, and the
power density arriving at Q is only cos
times that arriving at P . Thus, the total
power arriving at all points of the hemisphere (which must be the total emitted
power Φ )is
φ
Φ=
H
) L π
r 2
R 2
(cos
φ
(26.53)
R 2
= L π r 2
cos
φ
(26.54)
H
by pulling the constant out of the integral. Further simplifying,
R 2
S +
r 2
R 2
Φ= L π
cos
φ
(26.55)
r 2
= L
π
cos
φ
(26.56)
S +
because the area of H is R 2 times that of S + . Finally, by the Average height prin-
ciple, we get
Φ= L π r 2
π
(26.57)
so that
Φ
L =
r 2 ) .
(26.58)
π
(
π
 
Search WWH ::




Custom Search