User-Defined Sources Manual

A point source is placed in the scene via a special geometric description. There are two ways to define the location and orientation (pointing direction for a shaped source) of the source:

The geometric description of the point course is assigned a material that describes the radiometric properties of the course. An appropriate entry must be made in the Material Database file (.mat). A source material description is different than a conventional surface or volume material. This description consists of the following components:

Point sources, by definition, subtend a solid angle of zero from any view direction. This poses a problem for directly viewing a source under a ray tracing modality since the pixel solid angle (or portion thereof) that the ray represents will never "see" the source and there is no explicit surface to calculate the incident flux upon.

To approximate the expected effect of a point source, we add in a contribution from each source that falls within the solid angle of a pixel ray. That contribution is based on treating the source as if the distribution integrated by the solid angle is a Dirac delta function, such that the integrated radiance into the ray solid angle is equal to the magnitude of the intensity into the direction of the ray origin. Since the radiance required by DIRSIG as a solution for that ray is distributed over its solid angle, we weight the source integrated radiance by the inverse of that angle.

Transmission between the source and the sensor is also taken into account, and the rest of the solid angle is calculated as usual by sampling the scene (since the solid angle of the source is still zero, we don’t have to worry about the source blocking a portion of the scene).

To compute the total integrated radiant power from a DIRSIG radiant intensity (INT) file, the file must be integrated over the surface of a sphere (the file nominally represents the radiation going in all directions) and all wavelengths:

\( \mathrm{Total~Radiant~Power} = \oint \int\limits_{\lambda_1}^{\lambda_2} I( \lambda ) ~d\lambda ~d\Omega ~\mathrm{Watts} \)

Since real world sources are primarily employed to aid in human vision, most source descriptions are in terms of luminous flux (which incorporates the response of the human visual system) rather than radiant flux. The luminosity function (F) describes the (average) spectral sensitivity of human visual system to spectral radiance as a function of wavelength:

To get the total photopic power or brightness, the total radiant power integral must incorporate the luminosity function:

\( \mathrm{Brightness} = \mathrm{Total~Photopic~Power} = \oint \int\limits_{\lambda_1}^{\lambda_2} F( \lambda ) ~I( \lambda ) ~d\lambda ~d\Omega ~\mathrm{lumens} \)

Most sources only radiate a portion of their energy at wavelengths overlapping the human visual system. For example, most broad-band sources (incandescent) have a significant amount of energy emitted in the infrared regions (most sources of this type convert only about 10% of their input energy to visible light). The terms luminous efficacy and luminous efficiency are often employed to describe how much of the spectral energy generated by a source is in the response region of the human visual system. The luminous efficacy is the ratio of the total photopic power (brightness) and the total radiant power:

\( \mathrm{Luminous~Efficacy} = \mathrm{\frac{Total~Photopic~Power}{Total~Radiant~Power}} = \frac{ \oint \int_{\lambda_1}^{\lambda_2} F( \lambda ) ~I( \lambda ) ~d\lambda ~d\Omega} { \oint \int_{\lambda_1}^{\lambda_2} I( \lambda ) ~d\lambda ~d\Omega} ~\mathrm{lumens/Watt} \)

The luminous efficiency is the ratio of the luminous efficacy to the maximum luminous efficacy of 683.002 lumens/Watt.

\( \mathrm{Luminous~Efficiency} = \mathrm{\frac{Luminous~Efficacy}{683.002}} \)

Neither the luminous efficacy nor the luminous efficiency values explicitly convey information about how much of the electrical energy input to the source was converted to radiant energy. In some cases it is assumed that the any electrical power that is dissipated as heat (for example) is captured in the total radiant intensity.

The table below contains luminous efficacy and brightness values for common sources: ^[1]

Low pressure sodium (LPS) and high pressure sodium (HPS) lamps are commonly used in urban lighting applications for streetlights. An LPS bulb can have an electrical wattage of 10 to 150 Watts and an efficacy of 100 to 150. Unlike the broad spectrum generated from most incandescent bulbs, LPS bulbs generally have a narrow emission spectrum centered at 589.3 nm and are therefore recognized by a pure orange color. In contrast, an HPS bulb uses high pressure gas (to induce pressure broadening in the sodium emission) and other metals (including mercury) to create a broader spectral coverage than LPS bulbs. These bulbs are usually perceived as having a pinkish orange color. An HPS bulb has a typical electrical wattage of 50 to 1000 Watts and an efficacy of 100 to 150 (higher values for higher wattage bulbs).

Consider the 40w_2700k.int file included in the Sources1 demo, which has the following characteristics:

The luminous efficacy and brightness corresponds well to a "40 Watt tungsten incandescent" source in the table above.

By default the contribution of user-defined is always considered. However, in some cases the user may wish to disable these contributions. For example, to turn off all the lights in a scene at night or to ignore all the lights in a scene during the day. A single, central option is available to disable all user-defined source contributions (via the secondarysources.disable entry) and the default value for this option is false.

When scenes are spatially large and/or there are a large number of sources, computing the contribution from every source can be can lead to excessive run-times. Therefore a contribution rejection mechanism is used to ignore the contributions from sources that are expected to not make significant contributions to the radiance reflected by a surface.

Starting in DIRSIG 4.6, a sorting mechanism was introduced to setup an internal grid-like map that allows the model to only consider sources within a given area that will contribute significantly. This threshold drives which sources are considered important as a function of position in the scene. The image sequence below was created using the Sources1 demo, and shows how the threshold can impact the results. In this example, the source is a modest 40 watt bulb that is 5 meters from an 18% reflector. The default threshold (1.0e-05) considers the reflected radiance from this source to be insignificant, and ignores that sources (left image). The center image overrides the threshold with a value of 1.0e-06, which considers the contribution significant near the source, but it does not meet the threshold within the internal grid sorting as we get father away. The right image employs a threshold of 1.0e-07, which is low enough that the source is considered significant everywhere within the field-of-view.

This threshold can be manipulated in the options file via the secondarysources.threshold entry. The default value for this option is 1.0e-05.

Direct viewing of a source (discuss previously) is controlled in the .options file via the secondarysources.viewdirect entry. The default value for this option is true (sources are directly viewable).

The following example shows these variables being manipulated in a DIRSIG Options Editor (see the Dictionary tab) or directly in the .options file:

The following example shows the relevant sources options being manipulated in a DIRSIG Options Editor (see the Dictionary tab).