irradiance vs. radiance / luminance vs. illuminance

Hi again,

I've been puzzling over an extremely bright ground plane in one of my
images. I think I've tracked it down the the use of rpict's -i switch
for calculating irradiance rather than radiance. Can anyone help
clarify whether I'm on the right track?

If I want a luminance image, i.e. with levels corresponding to what one
might actually see (and can compare cd/m2 values at different points) I
omit the -i switch (and calculate radiance). If I include the -i switch
(and calculate irradiance), then I obtain illuminance (lm/m2 levels).
In both cases the radiometric units and turned into photometric units
(using Radiance's 179lm/W conversion factor) by the image viewer
software (I'm using the Ecotect RadianceIV viewer).

Radiance is W/m2.sr while irradiance is W/m2. Luminance is lm/m2.sr
(cd/m2) while illuminance is lm/m2 (lux). I think this means radiance
and luminance are properties of a (solid angle) source while irradiance
and luminance are properties of a (flat) surface?

So my extremely bright ground plane is actually showing its lux level
(the amount of light falling on it, rather than the amount of light it's
reflecting). If I want to illustrate the relative brightness of
different surfaces I guess I need to go back and calculate luminance
rather than illuminance.

I did notice that if I clicked on the bit of 'ground' beyond the ground
plane I get a much smaller value (in fact the same as the value in the
luminance image - because I'm clicking on the ground as defined in the
sky file - a glow source). Why is this? Does the concept of
illuminance on a direct light source not make sense?

Sorry for the long winded email equivalent of thinking out loud, and I
appreciate that my understanding of photometry is pretty sketchy.

Nick

Hi Nick,

You generally seem to know what's going on, so let me just verify and clarify a few points:

If I want a luminance image, i.e. with levels corresponding to what one
might actually see (and can compare cd/m2 values at different points) I
omit the -i switch (and calculate radiance). If I include the -i switch
(and calculate irradiance), then I obtain illuminance (lm/m2 levels).
In both cases the radiometric units and turned into photometric units
(using Radiance's 179lm/W conversion factor) by the image viewer
software (I'm using the Ecotect RadianceIV viewer).

Radiance is W/m2.sr while irradiance is W/m2. Luminance is lm/m2.sr
(cd/m2) while illuminance is lm/m2 (lux). I think this means radiance
and luminance are properties of a (solid angle) source while irradiance
and luminance are properties of a (flat) surface?

That is correct. Bear in mind that certain surfaces do not report irradiance, such as glass, which you see through as usual with the -i option.

So my extremely bright ground plane is actually showing its lux level
(the amount of light falling on it, rather than the amount of light it's
reflecting). If I want to illustrate the relative brightness of
different surfaces I guess I need to go back and calculate luminance
rather than illuminance.

Right.

I did notice that if I clicked on the bit of 'ground' beyond the ground
plane I get a much smaller value (in fact the same as the value in the
luminance image - because I'm clicking on the ground as defined in the
sky file - a glow source). Why is this? Does the concept of
illuminance on a direct light source not make sense?

Right again -- sources don't report light landing on them (irradiance), since they are considered to be emitters, and thus aren't affected by the -i option.

-Greg

Perhaps 0.25 is a bit too high for asphalt but reasonable for concrete?
BRE [1] list a table of surface reflectances with asphalt = 0.07 and concrete = 0.2 ~ 0.3. Anyone know of a more detailed reference for surface reflectances?
  

I think in this case, taking a grey chart, going out and finding the
value by comparision would work well. The problem is that there are so
many different types of concrete (granularity, additions, surface
finishing) plus outside factors (wet / dry etc) that you cannot get one
color / reflectance for the material.

BTW if I look at the road in front of the office I am working in right
now, I also wonder if the reflectance of the real-world asfalt is not
much higher. The 0.07 may be "pure" asfalt (so almost black as bitumen)
while the road surface here contains sand and stones to make it rough
and dureable. So if you walk around with your grey chart, just make one
more comparision :wink:

One problem with outdoors surfaces is that the values measured in lab
can be far apart from real world. Erosion, dust, humidity, vegetation...
many factors affect what we see outside, besides the pure "material". So
I guess in a proper simulation, one should always try to check these
reflectances, at least if they have big influence on the result.

CU Lars.

Hi all,

A question related to Greg previous mail:

Radiance is W/m2.sr while irradiance is W/m2. Luminance is lm/m2.sr

(cd/m2) while illuminance is lm/m2 (lux). I think this means radiance
and luminance are properties of a (solid angle) source while irradiance
and luminance are properties of a (flat) surface?

That is correct. Bear in mind that certain surfaces do not report
irradiance, such as glass, which you see through as usual with the -i
option.

and

Right again -- sources don't report light landing on them (irradiance),
since they are considered to be emitters, and thus aren't affected by the -i
option.

So if I understand well, when we calculate irradiance image (rpict -i
option) depending of surface materials, for various objects different values
are reported :
*light sources (light,glow) - radiance
*glass - radiance?
*trans - ?
*illum - radiance?
*other materials - irradiance

Please fill the blanks :slight_smile:

Thanks,
Marija

···

On Fri, Sep 5, 2008 at 7:13 PM, Greg Ward <[email protected]> wrote:

Hi Marija,

The complete list of materials that are completelly ignored when -i is set:

trans
trans2
dielectric
interface
mist
glass

Surfaces made of these materials pass as if they aren't there. Light sources, illum's, etc. yield radiance rather than irradiance. This includes surfaces modified by the following materials:

light
illum
glow
spotlight

I hope this clarifies matters.
-Greg

···

From: "Marija Velickovic" <[email protected]>
Date: September 8, 2008 5:17:02 AM PDT

Hi all,

A question related to Greg previous mail:

On Fri, Sep 5, 2008 at 7:13 PM, Greg Ward <[email protected]> > wrote:
Radiance is W/m2.sr while irradiance is W/m2. Luminance is lm/m2.sr
(cd/m2) while illuminance is lm/m2 (lux). I think this means radiance
and luminance are properties of a (solid angle) source while irradiance
and luminance are properties of a (flat) surface?

That is correct. Bear in mind that certain surfaces do not report irradiance, such as glass, which you see through as usual with the -i option.

and

Right again -- sources don't report light landing on them (irradiance), since they are considered to be emitters, and thus aren't affected by the -i option.

So if I understand well, when we calculate irradiance image (rpict -i option) depending of surface materials, for various objects different values are reported :
*light sources (light,glow) - radiance
*glass - radiance?
*trans - ?
*illum - radiance?
*other materials - irradiance

Please fill the blanks :slight_smile:

Thanks,
Marija

Hi Greg,

Greg Ward wrote:

The complete list of materials that are completelly ignored when -i is set:

trans
trans2
dielectric
interface
mist
glass

Surfaces made of these materials pass as if they aren't there.

What does this exactly mean when values on the image are traced in ximage?
For example if I have a window made of some glass material or some curtain
on it made of trans material. On irradiance image when I try to trace value
of some pixel on the window surface, what do I get? Is it irradiance of the
objects behind the window (outside ground, buildings etc)? Does
window/curtain transmittance has an influence on reported value?

What happens for surfaces made of transparent BRTfunc materials (like double
glazing)? Are they ignored too, or their irradiance value is reported?

Thanks again,
Marija

Hi Marija,

Surfaces made of these materials pass as if they aren't there.

What does this exactly mean when values on the image are traced in ximage?
For example if I have a window made of some glass material or some curtain on it made of trans material. On irradiance image when I try to trace value of some pixel on the window surface, what do I get? Is it irradiance of the objects behind the window (outside ground, buildings etc)? Does window/curtain transmittance has an influence on reported value?

If you query a pixel on a window, the window acts as if it's not there, so it's transmittance has no effect.

What happens for surfaces made of transparent BRTfunc materials (like double glazing)? Are they ignored too, or their irradiance value is reported?

The BRTDfunc isn't one of the materials on the list, so it shows up with an irradiance value like other materials. I'm sorry if this is confusing, but you can usually tell simply by looking at the image which materials are transparent and which aren't. If you want all materials to be treated the same and irradiance to be computed for everything, you can use the following in place of the rpict -i option:

vwrays [view options] -ff -x xres -y yres -pa 0 | rtrace -h -ff -opN octree | rtrace -ffc -I -x xres -y yres > irrad.pic

The first trace comes up with the intersection points and surface normals for the second trace, which computes irradiance at those points. You need to be sure that your view has no distant light sources in it, since those don't have intersection points or surface normals.

-Greg

vwrays [view options] -ff -x xres -y yres -pa 0 | rtrace -h -ff -opN octree | rtrace -ffc -I -x xres -y yres > irrad.pic

Obviously, I forgot the octree on the second rtrace command. Also, you need to specify the appropriate x and y resolutions for square pixels, or use the more complicated procedure with the vwrays -d option specified on the man page:

  http://radsite.lbl.gov/radiance/man_html/vwrays.1.html

-Greg