Explaining the Science of Goniophotometry

This guide explains the science of goniophotometry, including the IES motion types (A, B & C), what motion type to use when testing different types of lighting product, the importance of making measurements in the photometric far-field, and the pros and cons of configuring your system with either a photometer, a colorimeter or a spectroradiometer.

What is Goniophotometry?

Photometry is the science of measuring the brightness of light sources as the human vision system would rank them. The relationship between wavelength in the 380-780nm visible light band and the perceived brightness of the light at each wavelength is defined by the CIE spectral luminous efficiency function for photopic vision, which for daylight levels of illumination is known as the “photopic” response. Goniophotometry is the measurement of the amount of light shining in a specified direction. The word “gonio” derives from the Greek, meaning angle.

One of the reasons why goniophotometry is so important is for the purpose of creating “photometric data files”. These are machine-readable data files containing a table of the luminous intensity values for the light source as a function of angle, recorded over a 2pi steradian hemisphere or a complete 4pi steradian sphere. Photometric files are normally compiled into either of two industry-standard formats, IES (.ies) or EULUMDAT (.ldt). The photometric data for a light source can then be imported into lighting design software to simulate the level of illumination and beam pattern on a surface, helping to create lighting schemes without recourse to time consuming and expensive physical models.

What is a Goniophotometer?

A goniophotometer is a device that combines a photometer and a goniometric motion stage. The photometer is the photoelectric light meter that mimics the photopic spectral sensitivity of the human eye, turning incident illumination into an electric current. Some photometers double up as colorimeters, allowing for the measurement of both the amount and the colour (chromaticity, correlated colour temperature) of the light source. Sometimes, a spectroradiometer is used either instead of – or as a supplementary photodetector to – the photometer or colorimeter. For a further discussion on the pros and cons of photometers or colorimeters versus spectroradiometers, see below.

We should at this point define what we mean by “the amount of light”. For any kind of lamp or luminaire, the photometric property of interest in goniophotometry is the luminous intensity. Luminous intensity is the proportion of the total luminous flux emitted by the lamp, shining in a specified direction, per unit of solid angle, as illustrated in Figure 1. The units used are lumens for luminous flux and steradians for solid angle, but for convenience, we refer to the lumen per steradian as the more familiar unit called the candela (cd). In photometry, luminance (cd/m²) is what you measure from a display or sign, whereas luminous intensity (cd) is that property of interest from a lamp or luminaire.

Goniophotometer Design & IES Motion

Turning next to the goniometer part of the equation, this is the device that (usually) rotates and tilts the light source under test about a fixed photometer position. For a complete description of the output of a lamp or luminaire, we must measure the luminous intensity over the range of angles that the light source radiates into. For a downlighter-type luminaire, this means that we must measure the light output over a hemisphere (2pi steradians of solid angle), whereas with other types of light source (for example, incandescent lamps), we must measure over a complete sphere (4pi steradians).

In the case of most commercial goniophotometers, the light meter (photometer, colorimeter and/or spectroradiometer) is placed at a fixed position and receives light from the device under test as it is rotated and tilted by the goniometer. Alternative designs do exist and these include goniometer stages that rotate the lamp about its azimuthal axis while the photodetector rotates in an arc about the lamp, referred to as “moving detector” goniophotometers. A variation on that theme is the “moving mirror” type of goniophotometer, whereby the lamp is again rotated about its azimuthal axis, and a mirror rotates in an arc about the lamp, reflecting the light to a fixed photodetector placed at a distance away from the goniometer. Yet another goniophotometer design rotates the lamp about its azimuthal axis, and the light is received onto an array of photodetectors arranged in arc about the lamp.

What can be critical for some lamp technologies is whether the light source changes its orientation with respect to gravity (its burning position) during a scan. Incandescent and fluorescent lamps generally emit the same amount of light regardless of their orientation with respect to gravity. Conversely, some types of discharge lamps (notably metal halide), suffer from a significant change in light output depending upon whether they are held in a parallel or perpendicular direction with respect to gravity. For these types of lamp, you must firstly mount them such that they are operated in the intended (design) orientation, and secondly ensure that the burning direction doesn’t change with respect to gravity during a goniometric measurement.

LED-based solid state lighting (SSL) is also generally immune to the orientation in which it is operated. However, this assumes that the product enjoys adequate heat sinking. In some cases, a small difference in light output can be observed depending upon the orientation of the light source with respect to gravity, although nothing like as much as with metal halide discharge lamps.

The Illuminating Engineering Society of North America (IESNA, often abbreviated to just “IES”) published a standard in 2001 called LM-75-01 (since updated in 2019 to LM-75-19) which defines three generic types of goniometer motion. These are referred to as type A, B and C and these are illustrated schematically in Figures 2 and 3.

The motion of type A and B goniophotometers is very similar, and for both cases the device under test is rotated ±90° about orthogonal horizontal and  vertical axes. We refer to the corresponding type A or B coordinate system as horizontal-vertical (H-V or X-Y). A goniophotometer that employs type C motion rotates the device under test about an azimuthal axis (the nadir angle which is usually aligned along the polar axis of the coordinate system), while the other axis of motion is termed the elevation (or inclination) axis. The type C spherical coordinate system is referred to as θ-ψ, with θ (theta) being the elevation axis and ψ (psi) being the azimuthal axis.

In a type A or B goniophotometer, the orientation (burning position) of device under test tilts with respect to gravity during a scan. With a type C goniophotometer, the device under test maintains a constant orientation with respect to gravity during the measurement. To avoid errors arising from a lamp or luminaire tilting with respect to gravity, international  lighting standards such as IES LM-79-18, EN 13032-4 and CIE S025 require that the sample is measured using a type C goniophotometer. In addition, a correction factor should be determined and applied to the goniometer readings if the sample is mounted on the goniophotometer in an orientation that is different to its design orientation.

In general, type A/B motion goniophotometers are used when measuring directional lighting products. Typical applications include the testing of vehicle lighting (e.g. automotive headlamp and marker/side lamps) as well as other types of transportation and avionics lighting and signalling devices (e.g. VMS traffic signs). Type C goniophotometers are usually chosen when measuring lamps, luminaires and architectural lighting products. SSL Resource have designed certain of their type C goniophotometers to be able to be converted by the user to type B motion (and vice versa) with the purchase of an accessory kit. This flexibility allows one goniophotometer to accommodate almost any kind of light source, both directional vehicle lighting as well as architectural luminaires.

The moving detector or moving mirror goniophotometers employs type C motion in which the device under test is held in a vertical orientation, either shining down (“base-up) or shining up (“base down”). The lamp’s orientation with respect to gravity does not vary during a measurement. The first axis of rotation is about the vertical azimuthal axis, ψ (psi), over a range of 0-360°. The second axis of rotation is the elevation or inclination, θ (theta), and is applied to either the mirror or the detector which then rotates about the light source, over a range of 0-180°.

Moving mirror or moving detector type goniophotometers can sometimes be relatively bulky, so a popular variation on the theme is the type C horizontal motion design. In a type C horizontal goniophotometer, the device under test is held in a horizontal orientation and rotates about a horizontal azimuthal axis, while the sample’s second axis of motion (elevation or inclination) is about an orthogonal vertical axis. As with the type C vertical configuration, the device under test does not tilt with respect to gravity during a measurement. However, if the sample is not held in its intended or design orientation (for example, a ceiling fixture that would normally shine vertically down but is mounted in a vertical plane and shines horizontally), a correction factor must be determined and applied to themmgoniophotometer readings. SSL Resource offers an accessory device called the Burn Position Corrector (BPC) that automatically determines a correction factor and applies it to the directional luminous intensity readings.

Photometric Distance

A very important factor to understand when performing goniophotometric measurements of lamps and luminaires is the limiting photometric distance for the device under test. Also referred to as the far-field distance, this is the minimum separation between emitter and receiver at which the lamp behaves as a point source. In the far-field, the beam from the lamp propagates with constant luminous intensity in that direction. At the same time, in the far-field, the illuminance follows an inverse squared relationship.

The inverse squared rule states that for a lamp in the far-field, the luminous intensity is a constant at all distances, whereas the illuminance decreases with the square of the increase in separation between source and receiver. This is shown in Figure 4 with a lamp that delivers an illuminance of 400 lux at a defined separation (x), while at double that distance (2x), the illuminance drops to 400/22 = 100 lux.

The importance of making far-field measurements in goniophotometry can be understood as follows. The output of a lamp or luminaire is measured using an illuminance photometer in units of lux. Provided that the measurement has been properly performed in the far-field, luminous intensity values can then be accurately calculated using the inverse squared rule, as the product of  illuminance and the square of the separation between lamp and photometer. The IES or LDT photometric file is then compiled using these computed intensity values. This is convenient because luminous intensity is invariant in the far-field. When the photometric data file is then utilised in lighting design software, the illuminance from the lamp at any distance (in the far-field) can then be computed by the software by application of the inverse squared rule in reverse, illuminance being the luminous intensity divided by the square of the distance.

This prediction of illuminance would fail should the original goniophotometric measurement have not been performed in the far-field. In the near-field, the photometer is sampling just part of the beam and thus any intensity value computed from near-field illuminance will be significantly lower than the true far-field intensity.

The Photometric Near-Field Versus Far-Field

As explained above, it is important to perform goniophotometric measurements on lamps and luminaires in the far-field, but where exactly is this for any given light source? The concept of near-field versus far-field in photometry is shown in Figure 5.

The “luminaire” in this illustration is comprised of a linear array of three RGB coloured LEDs, the light from which additively mixes to create white light in the far-field. We define a parameter called the luminous aperture (d in Figure 5), which is the maximum dimension of the lit area of the luminaire under test. This could be the diameter of a circular downlighter fitting, the length of a linear or tubular lamp, or the diagonal of a square or rectangular luminaire.

Considering next the separation between the plane of the light source and the surface being illuminated (x in Figure 5), we regard distances of up to x = d as being fully near-field. It is only when x > 10d that the beam of light has (probably) entered the far-field. For further information on this topic, please refer to the CIE publication 70-1987 “The Measurement of Absolute Luminous Intensity Distributions”.

The 10d far-field multiple can be safely applied to any diffused, wide angle lamp or luminaire. However, care must be taken with lamps with narrow beams, or luminaires that are comprised of an array of widely spaced emitters. The far-field distance for these types of light source will be much greater. EN 13032-4 and CIE S025 both recommend using a 15d multiplier for sources with beam angles of <30°. Meanwhile for luminaires comprised of widely spaced, discrete emitters, the working distance should be computed as being ≥15 times the sum of the luminous aperture and the separation between emitters.

In the automotive field, vehicle lighting regulations (e.g. UNECE R20, R98, R112, R122) stipulate that headlamps must be measured at a distance of 25m. Research has shown that measurements of headlamps performed at ~10m for comparison testing correlate closely with the 25m far-field results, but for type-approval testing, the measurements must be performed at the regulation 25m.

Photometer or Spectroradiometer?

In recent years, goniometers in combination with spectroradiometers (“goniospectroradiometers”) have become a popular choice for measuring the output of luminaires. This popularity is often attributed to the ability of a spectroradiometer to perform all of the relevant optical tests in one measurement, including measuring if photometric, colorimetric and spectral parameters. However, compared to traditional photometers or colorimeters, spectroradiometers suffer from a number of critical limitations arising from their relative lack of photometric sensitivity and more restricted dynamic range.

A spectroradiometer is an optical detector that separates the incident light into its component wavelengths, which are imaged onto a linear or 2D photodiode array. Each detector in the array is called a “pixel” and receives one narrow band of wavelengths, thus recording the spectral power distribution of the light source. Relevant photometric and colorimetric parameters are calculated from the product of the spectral power and the CIE photopic or tristimulus colorimetric observer functions.

The main technical arguments against the use of spectroradiometers in goniometry concerns their more limited dynamic range and relative lack of photometric sensitivity. Consider that the silicon photodiode used in a filter photometer has an inherent radiometric range that extends over 9 decades from 1pW to 1mW. In comparison, typical spectroradiometers provide a much more limited dynamic range of just 3000:1, which is 4 decades, and with a signal-to-noise ratio of 300:1. The absolute sensitivity of a spectroradiometer also compares poorly against a filter photometer. Based on a measurement distance of 1m, a filter photometer will be capable of measuring luminous intensity values as low as 0.001 candelas, whereas a spectrometer will struggle to detect much below 0.5 candelas.

The impact of the reduced sensitivity and dynamic range from which spectroradiometers suffer will manifest itself in a number of different ways. First, at any any given measurement distance, the spectroradiometer may receive insufficient signal and therefore the test data may be rendered “noisy” and inaccurate. Moving the spectroradiometer closer to the goniometer will improve the signal level, but that necessitates a time consuming and inconvenient realignment of the whole goniometer system. When testing a high power light source, the spectroradiometer may need to be moved further away again. Second, when testing light sources with complex beam patterns, for example street lights and vehicle headlamps, the measurement distance will be set to sample the high intensity portions of the beam, but in regions of low intensity, the spectroradiometer may again suffer with insurmountable signal to noise errors. Third, we know that the rules for far-field photometry dictate a certain minimum separation between light source and receiver. For very narrow angle beams, the extended measurement distance may cause the light level reaching the spectroradiometer to fall below the minimum detectable signal level, especially if the lamp is low power.

The reality is that filtered photometers or colorimeters are relied upon as the primary sensor in nearly all professional, accredited lighting test laboratories and in National Measurement Institutes (NMIs). Spectroradiometers should not be used as the main optical detector for goniometric measurements, however their use as secondary detectors for spectral computations (e.g. colour rendering) is recommended.