Allgemein

Not just “flicker-free”: Healthy lighting must enter the time domain

April 30, 2026

From Percent Flicker and Flicker Index to SVM, PstLM, and PAVM — rethinking the dynamic relationship between light and humans

For more than a decade, the LED lighting industry has commonly used the statement: “Our lights are flicker-free.”

But the real question is: What does “flicker-free” actually mean?

Is it because camera sensors cannot capture visible banding?
Is it because the human eye cannot perceive flicker?
Is it because Percent Flicker is low?
Is it because SVM is within limits?
Or because PstLM passes compliance thresholds?

From the perspective of healthy lighting, human-centric lighting, and age-inclusive environments, the issue is far more complex than this.

Flicker should more precisely be understood within the framework of Temporal Light Modulation (TLM). It is not a single phenomenon, nor can it be defined by a single metric.

More importantly, flicker is not only a property of the light source itself. It is: the dynamic light exposure experienced by a person in a specific space, at a specific time, performing a specific activity, under a specific physiological and psychological state.

This is why next-generation healthy lighting cannot be limited to discussions of spectrum, illuminance, correlated color temperature, or color rendering index, nor even melanopic EDI / DER alone.

We must also incorporate the temporal quality of light into the system-level understanding of lighting.

1| Flicker is not a single phenomenon, but a set of phenomena

In the lighting industry, all temporal variations in light output are often broadly referred to as “flicker.” However, strictly speaking, variations in light output over time can lead to different types of human perceptual and physiological responses.

At minimum, these can be categorized into three major visual phenomena:

1. Direct Flicker: visible flicker

This is the most intuitive form of flicker. It refers to situations where the human eye directly perceives light as flashing, pulsing, or fluctuating.

This typically occurs under conditions such as:

low frequency operation
high modulation depth
poor driver quality
unstable dimming behavior

It most commonly leads to:

visual discomfort
eye strain
attention disruption
headaches
adverse reactions in sensitive individuals

This layer is typically described using metrics such as Percent Flicker, Flicker Index, and PstLM / Mp.

2. Stroboscopic Effect: motion discontinuity perception

This is not perceived as flicker in the light itself, but rather as temporal distortion of moving objects. For example:

fan blades appearing to stop or reverse
hand movements appearing discontinuous
rotating machinery appearing to change speed incorrectly
motion trajectories appearing segmented

This is not only a comfort issue. In environments such as industrial facilities, healthcare, sports, kitchens, and laboratories, it can become a safety risk.

This layer is primarily described using SVM (Stroboskopische Sichtbarkeitsmessung).

3. Phantom Array Effect: spatial-temporal image splitting

This is a historically underestimated phenomenon. It typically occurs during rapid eye movements (saccades). Instead of fixating steadily on a light source, humans constantly move, scan, shift gaze, and change focus.

In such conditions, certain lighting systems—especially:

high-intensity point sources
linear LED luminaires
vehicle headlights
stage lighting
retail display lighting

may produce the perception of multiple separated light images or streaks. This is known as the Phantom Array Effect.

It is particularly common in:

high-brightness point sources
exposed LED systems
linear luminaires
automotive lighting
entertainment lighting
retail environments
outdoor nighttime lighting
high-contrast visual scenes

In the latest TLM framework, this effect is increasingly quantified using PAVM (Phantom Array Visibility Measure).

This leads to a critical realization: Humans are not static lux meters. Humans move, scan, turn, fatigue, and respond to light in individualized ways.

2 | What do common flicker metrics actually represent?

Let us break down the main indicators.

1. Percent Flicker (modulation depth)

Percent Flicker is the most intuitive metric. It describes the relative difference between the maximum and minimum light output within a cycle.

In simple terms: the deeper the fluctuation, the higher the Percent Flicker.

If light output drops close to zero during a cycle, Percent Flicker becomes high. If light output remains nearly constant, Percent Flicker is low.

Its advantages:

simple
intuitive
easy to interpret

However, its limitations are significant. It does not tell us:

frequency of modulation
waveform shape
whether it is sine wave or PWM
duty cycle characteristics
perceptual visibility
stroboscopic risk
phantom array potential
fatigue or headache risk

Therefore, Percent Flicker only answers: How deep is the fluctuation?

It does NOT answer: Is this fluctuation harmful or perceptible to humans?

These are fundamentally different questions.

2. Flicker Index

Flicker Index extends beyond Percent Flicker by considering the area under the waveform over time, relative to the average level.

This makes it more sensitive to:

waveform shape
duty cycle behavior
PWM characteristics
temporal distribution of light output

For example, two systems with identical Percent Flicker (e.g., 50%) may have very different perceptual impacts depending on whether the waveform is smooth (sine-like) or sharp (square-like).

Flicker Index can better capture these differences. However, it still does not represent full human response.

It does not incorporate:

frequency-dependent visual sensitivity
task-dependent perception
motion-based stroboscopic effects
eye-movement-related phantom array effects
individual sensitivity differences
long-term fatigue or neurological response

Thus, Flicker Index helps describe waveform quality, but cannot be directly equated with health risk.

3. Frequency

Frequency is a fundamental parameter in all TLM analysis.

It answers: How many times per second does the light fluctuate?

Measured in Hz. However, frequency alone is insufficient for risk assessment.

Because at the same frequency (e.g., 1,000 Hz):

low modulation depth may be harmless
high PWM modulation may still cause issues
stationary viewing may reduce perception
rapid scanning of bright sources may induce phantom array effects
sensitive individuals may react differently

Therefore, frequency must always be interpreted together with:

modulation depth
waveform type
duty cycle
dimming level
viewing condition
task type
population sensitivity

In other words: Frequency defines the time scale, but not the full human risk profile.

4. PstLM / Mp: direct flicker perception metrics

PstLM is used to describe short-term visible flicker perception (direct flicker).

It helps answer: Can a typical observer perceive flicker under given conditions?

In newer frameworks such as TM-39, Mp is also used for similar evaluation of direct flicker. Its key value is that it shifts analysis from purely physical waveform characteristics to human perception-based assessment.

However, it has limitations. It does not fully capture:

motion discontinuity perception
phantom array effects during eye movement
migraine triggering potential
long-duration fatigue effects
cognitive load impact
neurological responses (EEG/fMRI-level effects)

Thus, PstLM / Mp is essential for direct flicker assessment, but not sufficient as a complete human discomfort model.

5. SVM: stroboscopic visibility measure

SVM is designed to describe the visibility of stroboscopic effects.

It answers: Will moving objects appear discontinuous under this lighting?

It is critical in:

manufacturing
machining
rotating machinery environments
laboratories
kitchens
sports facilities
medical procedure areas
logistics and warehousing
fast hand-motion tasks

Its importance is not whether light flickers, but whether motion perception is distorted. In environments with rotating machinery, stroboscopic effects can lead to misinterpretation of equipment motion, creating safety hazards.

However, SVM is not a universal indicator. It does not fully predict:

phantom array effects
headaches or migraine triggers
long-term visual fatigue
neurological sensitivity responses
non-visual physiological effects

Therefore, SVM is a key metric for dynamic visual safety, but not a comprehensive “overall comfort score.”

6. PAVM: Phantom Array Visibility Measure

PAVM is a relatively recent and important development.

It corresponds to the Phantom Array Effect, which refers to the perception of multiple separated light images or streaks when the human eye moves—such as during saccades, scanning, or head rotation.

This type of effect has historically been underrecognized, largely because conventional testing assumes a stationary observer.

However, in real-world conditions, humans are never static. We constantly:

walk
turn our heads
scan environments
shift attention between near and far objects
move between screens and lighting environments
visually browse objects in retail spaces
observe exhibits while in motion
scan vehicle headlights at night

Therefore, PAVM addresses a critical gap in traditional flicker evaluation.

It is particularly relevant for assessing:

retail environments
exhibition and museum spaces
stage and entertainment lighting
high-brightness point sources
automotive lighting
linear luminaires
outdoor nighttime lighting
educational and medical environments

The emergence of PAVM highlights a fundamental shift: Flicker evaluation can no longer assume static viewing conditions. It must account for human motion, eye movement, and real behavioral patterns.

3 | Core differences between key metrics

We can summarize the relationships between the main indicators as follows:

Metrics	Main Description	Layer Classification	Main Purpose	Maximum Limitations
Flimmeranteil	Depth of light wave fluctuations	Description of physical stimuli	Fast judgment of modulation degree	Does not consider frequency and human body perception
Flimmerindex	Waveform area and duty cycle	Description of physical stimuli	Determine waveform quality	Not a human response model
Frequency	Speed of light fluctuations	Time parameters	Determine time scale	Cannot judge risk individually
PstLM / Mp	Directly visible flicker	Visual response model	Direct Flicker	Does not cover SE / PAE / physiological discomfort
SVM	Jumping sensation of moving objects	Visual response model	Stroboscopic Effect	Not suitable for representing overall health risk
PAVM	Images separated by eye movement	Visual response model	Phantom Array Effect	Still mainly a visual model

Therefore, we can draw a key conclusion:

Percent Flicker, Flicker Index, and Frequency describe the physical stimulus itself;
PstLM / Mp, SVM, and PAVM describe different layers of visual response;
but none of them alone can fully define human discomfort or neurological response.

4 | Why we cannot simply say “flicker-free”

The term “flicker-free” is fundamentally too vague.

When a product claims to be “flicker-free,” it should at minimum clarify:

Is Percent Flicker very low?
Is Flicker Index very low?
Does it comply with PstLM requirements?
Does it comply with SVM requirements?
Has PAVM been measured?
Was it tested at 100% output, or also at 10%, 20%, and 50% dimming levels?
Is this based on single-luminaire testing or full spatial system testing?
What dimming method is used—sine-wave, DC, PWM, or quasi-square waveform?
Were sensitive populations considered, such as children, elderly users, or individuals prone to migraines or neurological sensitivity?

If these questions are not addressed, then “flicker-free” is merely a marketing phrase.

A genuinely professional statement should instead specify: Under defined output levels, defined dimming conditions, and defined control methods, the luminaire exhibits the following values: Percent Flicker, Flicker Index, Frequency, PstLM / Mp, SVM, and PAVM.

Only then can performance be verified, compared, and meaningfully applied in healthy lighting design.

5 | Flicker must be understood within the framework of “human × space × time × activity”

One of the most common mistakes in healthy lighting is reducing complex human responses to a single metric.

In the past, lighting was simplified into lux. Later, it became CCT. Then CRI.
Today, many people reduce healthy lighting to melanopic EDI.

But real-world conditions are not that simple.

Human-centric lighting must return to four dimensions: human × space × time × activity

Flicker should be understood in the same way.

1. Human: different people have different sensitivity to flicker

Not everyone responds to light in the same way. Under identical lighting conditions, some people may experience no issues, while others may report glare, irritation, headaches, or fatigue.

Groups that require special attention include:

Children and adolescents

They spend long hours in classrooms, tutoring centers, study desks, and screen-based environments.

Flicker may not always be explicitly reported, but it can manifest as:

reduced attention
reading fatigue
visual discomfort
irritability
decreased learning efficiency

Classroom lighting should not be evaluated solely by illuminance levels. It must also consider whether light remains stable, especially under dimming conditions, and whether low TLM performance is maintained in real operation.

Office workers and sub-healthy populations

Many office workers already experience:

dry eyes
headaches
sleep disruption
neck and shoulder tension
reduced attention
visual fatigue

High temporal light modulation (TLM) in office environments may not be the sole cause, but it can act as an aggravating factor. This is particularly relevant in:

open-plan offices
smart lighting systems with adaptive dimming
sensor-based control systems
environments mixing multiple lighting brands and drivers

In such cases, system-level TLM management becomes essential.

Migraine-sensitive and neurologically sensitive individuals

These individuals may not only be sensitive to brightness, but also to:

flicker
high contrast
bright point sources
dynamic lighting changes
phantom array effects
visual noise

For them, lighting discomfort is often more pronounced and complex. Therefore, healthy lighting should not only be designed for the “average user,” but also for sensitive populations.

Elderly users

The issue for elderly populations is not simply “more light.” They generally require:

stable illumination
soft visual environments
low glare
low flicker
safe dynamic perception
non-disruptive nighttime lighting

In corridors, kitchens, staircases, bathrooms, and long-term care environments, both stroboscopic effects and phantom array effects can significantly impact safety and spatial perception.

Medical and long-term care populations

In healthcare environments, users are often in a vulnerable physiological and psychological state. They may experience:

poor sleep
anxiety
fatigue
pain
increased sensitivity to environmental stimuli

In such contexts, light is not just illumination—it is part of the recovery environment.

Therefore, in addition to illuminance, glare, and spectrum, TLM must also be considered as part of lighting design in hospitals, wards, nursing stations, and rehabilitation spaces.

2. Space: single-luminaire compliance does not guarantee system compliance

Many flicker evaluations remain at the single-luminaire level. However, in real environments, people are not exposed to one lamp—they experience:

multiple luminaires
multiple angles
multiple reflections
multiple control circuits
multiple drivers
multiple dimming states
multiple visual tasks

A single compliant luminaire does not guarantee a compliant space. For example:

phase differences between luminaires
beat frequency between different drivers
waveform inconsistencies across dimming groups
reflections amplifying high-brightness artifacts
differences between eye-level exposure and desk-level measurements
perceptual effects during walking or scanning (including PAE-related phenomena)

Therefore, future lighting acceptance testing should not rely solely on single-luminaire reports.

Instead, it should evaluate real spatial exposure, including:

user eye-level conditions
desktop plane conditions
wall and surface reflections
circulation paths
lighting scenes and modes
nighttime operation modes
multi-luminaire simultaneous operation
scene transitions

This marks the shift from “luminaire flicker” to “spatial TLM exposure.”

3. Time: flicker risk varies across operating conditions

Flicker should not be evaluated only at 100% output.

In practice, many LED systems exhibit their most problematic behavior under dimmed conditions, especially:

10% dimming
20% dimming
low-level nighttime modes
sensor-driven dimming systems
scene switching events
daylight harvesting adjustments
standby modes
emergency operation modes
low-duty-cycle PWM control

This is why healthy lighting products must provide TLM performance data across multiple dimming levels—not just full output conditions.

Time also includes human biological timing. Morning, daytime, evening, nighttime, and pre-sleep conditions all require different lighting strategies. Nighttime healthy lighting should not be defined only by “low blue light.”

It should also ensure:

low illuminance
low glare
low melanopic stimulation
low TLM
low visual stress

A truly sleep-friendly lighting environment is not just warm in color temperature—it is stable, low-stimulus, low-fluctuation, and low-disruption as a system.

4. Activity: different tasks require different flicker management

The same luminaire can present very different risk profiles depending on the activity. For example:

Reading

Key concerns:

direct flicker
eye fatigue
attention interference
long-term visual stability

Office work

Key concerns:

interaction between display and lighting TLM
long-duration visual load
stability under dimming conditions
risk of eye strain and headaches

Industrial environments

Key concerns:

rotating machinery
moving objects
stroboscopic effects
safety misperception risks

Healthcare environments

Key concerns:

precision tasks
patient sensitivity
long working hours and fatigue
nighttime lighting in care stations

Retail and exhibition spaces

Key concerns:

high-brightness point sources
rapid visual scanning
phantom array effects
balance between visual comfort and product presentation

Residential and hospitality environments

Key concerns:

emotional comfort
relaxation
sleep preparation
nighttime safety
low visual stimulation

Therefore, healthy lighting should not be defined as “one scenario = one metric set,” but rather as a system where different activities require different lighting recipes.

6 | The real meaning for all-age and sub-healthy populations

All-age healthy lighting is not about making light brighter or warmer.

Its real purpose is to address: the integrated lighting needs of different ages, physiological states, spaces, times, and activities.

Within this framework, flicker management represents the temporal quality of light. If we define lighting functions as:

illuminance → visibility (“can I see clearly?”)
CCT → visual atmosphere (“what does the light feel like?”)
CRI → color accuracy (“are colors rendered correctly?”)
EDI / DER → spectral biological stimulus (“how does the spectrum interact with physiology?”)

Then TLM management addresses something different: whether light is stable, non-disruptive, and temporally compatible with human perception and physiology.

This is especially important for sub-healthy populations.

Because sub-health is rarely a single condition—it is a cluster of chronic stress factors, wie zum Beispiel:

poor sleep
eye fatigue
mental tension
headaches
anxiety
reduced attention
heightened environmental sensitivity

In such states, every environmental stimulus can become a load.

Flicker, glare, high-intensity point sources, incorrect color temperature, and excessive nighttime stimulation can all accumulate into physiological stress.

Therefore, future healthy lighting should not only aim to be “bright” or “visually appealing,” but should instead focus on: a measurable, manageable, and verifiable low-burden lighting environment.

7 | What product development should look like

Future healthy lighting products should not only specify:

power
luminous flux
color temperature
CRI
beam angle
UGR

They should also clearly provide:

Flimmeranteil
Flimmerindex
Frequency
PstLM / Mp
SVM
PAVM
waveform characteristics
duty cycle
dimming curve behavior
driver ripple characteristics
TLM performance at 100%, 50%, 20%, and 10% output levels
performance under different control protocols

Especially for products claiming:

gesundes Licht
eye protection lighting
learning environments
sleep-friendly lighting

Full transparency of temporal behavior is essential. Because “flicker-free” is not a scientific endpoint.

A truly professional product should instead state: Under defined output conditions, control strategies, and application scenarios, the TLM-related risks are controlled within specified ranges.

8 | What lighting design practice should evolve into

Lighting design should not rely solely on illuminance calculations or photometric files.

A more complete data structure is required: Photometry + Spectrum + Alpha-opic + TLM Profile

A healthy lighting design dataset should include:

photometric distribution data
spectral power distribution (SPD)
EDI / DER metrics
CCT / Duv
CRI / TM-30
Flimmeranteil
Flimmerindex
Frequency
PstLM / Mp
SVM
PAVM
dimming states
control profiles
scene tags

Only with this level of integration can we begin to build a true digital twin of healthy lighting.

Not just simulating how bright a space is, but simulating: how a human actually experiences dynamic light exposure over time, across activities and spatial contexts.

9. How site acceptance testing should evolve

Commissioning should no longer be limited to single luminaires or desktop illuminance measurements.

Especially in schools, offices, healthcare facilities, long-term care environments, hotels, high-end residential projects, and industrial spaces, evaluation should progressively include:

single-luminaire TLM
multi-luminaire combined TLM
eye-level user exposure TLM
task plane TLM
surface reflection TLM
dimming-state TLM
nighttime-mode TLM
scene-transition TLM
PAE risk during movement and scanning
SVM risk under dynamic mechanical operation

This marks the transition from: “product compliance” → “spatial compliance”

It is a critical step in moving healthy lighting from concept to engineering verification.

10 | Next-generation healthy lighting: beyond spectrum, beyond flicker

The lighting industry has gone through several stages of evolution:

from “having light” → “enough brightness”
from “brightness” → “energy efficiency”
from “efficiency” → “light quality”
from “light quality” → “healthy lighting”

But today, healthy lighting cannot be reduced to isolated indicators. We cannot only talk about:

high CRI
low blue light
no flicker
high EDI
full spectrum
eye protection
sleep support

These are all meaningful, but incomplete. The next step is to return lighting to real human life:

What kind of person?
In what kind of space?
At what time?
Performing what activity?
In what physiological and psychological state?
Receiving what spectral, spatial, temporal, and dynamic light exposure?

This is the true foundation of human-centric lighting.

Conclusion: stop asking only “is there flicker?”

We should stop asking: Does this light have flicker?

And instead begin asking: Under what frequency, waveform, dimming condition, spatial configuration, task context, and population sensitivity does this system produce direct flicker, stroboscopic effects, phantom array effects, and potential visual or physiological discomfort?

This is not overcomplicating the problem. It is clarifying it. Because humans are not static instruments. Spaces are not single-luminaire test setups. Time is not an average value. And activities are not abstract scenarios.

True healthy lighting must shift from static parameters to dynamic exposure:

from spectrum → time
from single luminaire → spatial system
from average user → real human diversity
from product metrics → lived experience

This is the direction flicker research must evolve toward.

And it is the challenge that next-generation healthy lighting, all-age lighting environments, and sub-healthy-friendly spaces must address.

Lawrence Industry Observation

I believe the healthy lighting industry is approaching a major inflection point.

One group of companies will continue to rely on marketing language, wie zum Beispiel:

flackerfrei
eye-friendly
full-spectrum
low blue light

Another group will move into engineering language, including:

SPD
EDI
DER
PstLM
SVM
PAVM
dimming profiles
spatial exposure models

And one step further, the real leaders will operate in a human-centric language framework: human × space × time × activity × individual sensitivity

This is the direction I have consistently emphasized. Healthy lighting is not about optimizing a single metric in isolation.

It is about building a lighting environment system that is:

measurable
verifiable
manageable
continuously optimizable

Within this system, flicker management is not optional. It is a fundamental component.