The Mythical “Flicker-Free” Luminaire
Promoting Proper Lighting Through Metrics and Understanding
by Daniel Murby, Optical Engineer
What is Flicker?
Everyone has experienced light flicker. Whether noticed or not, light flicker refers to the change of visible light intensity with respect to time. However, all flicker is not created equal and different observers respond to flicker in different ways. The lighting industry uses a broad term, Temporal Light Artifacts (TLA), to describe this light behavior. Within TLA there are two subcategories: 1) Flicker and 2) Stroboscopic Effects. Flicker categorizes lower frequency, higher-impact light modulations that are readily perceptible by most occupants. Flicker leads to negative health consequences including headaches and dizziness. In applications that require high levels of visual acuity, such as manufacturing, flicker can be dangerous. Stroboscopic Effects, on the other hand, are more nuanced and exist at higher frequencies. These effects tend to manifest when objects or light sources are moving at high speeds - for example in sports or while operating a motor vehicle. Although there are legitimate applications where flicker is preferred or advantageous, such as emergency vehicle lighting or entertainment strobes, the vast majority of illumination applications aim to minimize flicker and its effects. In order to achieve this goal, it is necessary to understand the origins of flicker, how to measure it, and how to implement mitigation solutions
Where Does Flicker Come From?
A light source, whether LED or not, does not inherently flicker on its own. Instead, flicker is introduced by the power delivery, conversion, and control system to which the LED is coupled. Although flicker manifests itself visually in light output, flicker actually starts as a function of voltage. Typical powerlines provide alternating current (AC) voltage (a sinusoidal waveform with a typical frequency between 50 and 60 cycles per second, or hertz (Hz)). LEDs cannot be powered by AC voltage on their own. Instead, the power input must be altered to direct current (DC) voltage. This conversion process is handled by the power supply. In most cases, the power supply will return a flicker frequency at twice the line frequency due to how the sinusoidal power is corrected, or rectified. For example, it is very common to see flicker frequencies of 100Hz in European markets and 120Hz in North American markets, since those markets operate on 50Hz and 60Hz mains frequencies respectively.
Beyond power conversion, the other key source of voltage-induced flicker occurs when an LED lighting product is dimmed. Depending on the dimming system and its compatibility with the luminaire, additional flicker characteristics may appear - especially when operating at dimmed light levels. One dimming method common in many digitally-controlled lighting solutions is the use of pulse-width modulation (PWM). In a PWM system, to reduce perceived brightness, the same LED is rapidly switched on and off. The perceived light level corresponds to the percentage of time spent in the "on" state. Since the low ("off") and high ("on") states are already defined, the expected flicker mitigation technique is to increase the switching frequency; however, this approach often requires more expensive electrical hardware, increased heat generation (or phrased alternatively: lower electrical efficiency), and in some cases, the risk of audible noise generation.
How is Flicker Measured?
There are several different flicker measurement techniques used in the industry. These techniques have various positive and negative attributes. Fundamentally, a successful flicker measurement needs to account for three critical characteristics: the magnitude, frequency, and waveform of the light modulation.
Likely the easiest to calculate, flicker percentage is used to measure the magnitude of the perceived flicker. Flicker percentage is calculated by taking the difference between the peak light output value and the trough light ouput value and dividing it by the sum of those two values. This difference is then converted to a percentage. While straightforward to calculate and representative of magnitude, flicker percentage does not take frequency nor waveform into consideration and cannot be used to characterize temporal light artifacts alone. At best, a flicker percentage may provide a rough estimate for how flicker-prone a luminaire may be - at worst, it can disqualify suitable fixtures, such as those using PWM-dimming systems.
Flicker index approaches measurement by first calculating the average light output for a given flicker cycle. Flicker index then calculates the total area under the waveform curve above, and the total area below that average light output. Finally, the total area above the average light output is divided by the sum of both areas. The advantage to this method is that it relies on relatively simple calculations but considers the shape of the waveform rather than solely relying on amplitude values like flicker percentage. Flicker index is typically represented on a 0-1 scale, with a lower score representing a lower flicker perceptibility.
The latest flicker measurement techniques, including stroboscopic effect visibility measure (SVM) and flicker perception, have a critical difference from Flicker Index and Flicker Percentage. Specifically, these "perception metrics" rely on a Fourier transform of the light output to convert it from time domain to frequency domain. This conversion successfully accounts for both frequency and amplitude, creating a more comprehensive representation of the perception of light flicker. A Fourier transform will dissect complicated light output signals into the constant frequency "building blocks", allowing a light source to be evaluated for presence of flicker at frequencies known to cause issues to occupants. To create a meaningful score, flicker perceptibility metrics compare the fundamental frequencies with empirical data surrounding human impact. These frequencies can be weighted against empirical human perception test results to determine the risk of negative impact on human occupants. While empirical tests can introduce subjectivity, it allows a direct comparison of light sources against each other and often makes selecting light sources easier. Some of these metrics, such as Flicker Perception, limit the frequency range to what would classify as flicker, typically <120Hz. Other metrics, such as IEEE's PAR1789-2015 standard, expand the frequency range to 1000Hz to account for stroboscopic effects as well. Frequency-domain flicker analysis involves more testing and data analysis than earlier flicker scores such as Flicker Index and Flicker Percentage, but by combining change amplitude and change frequency, these metrics provide the greatest amount of information on luminaire performance.
The "Flicker-Free" Luminaire
It is not uncommon to see luminaires marketed as "flicker-free", especially if they are LED. While this labeling might ease some anxiety about LED technology adoption, these statements can be misleading. Almost all light sources, whether a candle, an incandescent bulb, or an LED luminaire, exhibit changes in light output with respect to time and therefore - flicker. This is especially true of any luminaire that is powered over AC mains - the filament will produce a measurable flicker as it reacts to alternating current supplied to it. The language used should be changed and "flicker-free" statements, whether from luminaire manufacturers or power supply and driver manufacturers, should be challenged. Instead, the goal should be to eliminate the harmful impacts of flicker and stroboscopic effects in the environment in which the luminaires are installed. The two critical paths for eliminating the impact of flicker, rather than intrinsically flicker itself, are lowering the modulation percentage and increasing the frequency. Different applications and project budgets will demand different levels of performance, but any regularly occupied space should be actively designed to eliminate the risks of exposure to flickering artificial light. In all cases, flicker treatment involves design tradeoffs like cost and fixture size - generally greater flicker reduction requires more electrical components and higher fixture costs. When evaluating the merit of a luminaire's flicker claims, look for both a modulation percentage and a frequency, as those two pieces together will provide the clearest indication of flicker performance.
Sometimes considering human physiological impact is not enough and lighting design requires additional flicker investigations and precautions. One critical application area is television and video production. Flicker is important to monitor in these situations because the recording equipment may "see" flicker that the eye does not naturally detect based on the frame rate of the recording. For example, slow motion replays frequently used in sports broadcasting require higher frame rates that can exacerbate flickering field illumination. For this reason, professional sports arenas have strict guidelines on what light sources can be used during broadcasted events. Depending on the light source used, cameras may need to be ‘synced' to the light source to ensure consistent light levels frame-to-frame. Lumenpulse designs many of its fixtures to utilize high enough frequencies to maintain compatibility with video production.
Lumenpulse relies on flicker studies to understand the characteristics of the light output and to ensure the highest quality light delivery in all applications. Light flicker, whether visually detectable or not, can create problems for occupied spaces. By understanding the nature of light flicker, how to measure it, and marrying the lighting design to the desired application, proper illumination can be realized. Chasing the "flicker-free" luminaire is not a worthwhile endeavor and could end up costing more than the lighting application requires. Instead, Lumenpulse recommends understanding your application and the requisite levels of flicker to minimize any negative impact. This includes fully understanding any compliance requirements, for the space - WELL Standard, for example, has flicker requirements listed. Flicker will continue to grow in importance as more human impacts of artificial light are studied and specified - investing in a quality flicker meter that can output the right flicker measurements and using the meter as a tool to make design decisions, mockup evaluations, and product qualifications is a key step for any specifier.