Pulse Frequency, Duty Cycle and Dose in Red Light Therapy: Why Hz Alone Cannot Tell You What Light Will Do

When a precise number still does not mean precise biology

Red light therapy can look impressively precise. A display shows 10 Hz, 40 Hz or 100 Hz. Next to it, the user sees a duty cycle of 50%, 70% or 75%. Intensity is adjustable, treatment time is defined, and a preset mode may combine all of these parameters automatically.

Numbers create the impression of control.

The problem is that biological precision does not arise simply because a device can display a frequency down to the nearest hertz.

Photobiomodulation is the result of several variables acting together: wavelength, irradiance, exposure time, spatial light distribution, treatment distance, pulse structure, duty cycle, treatment repetition and the biological state of the tissue being exposed. One of the persistent problems in PBM research has been incomplete or inconsistent reporting of these parameters, incorrect dose calculations, and reliance on manufacturer specifications without independent verification of the actual optical output.

Jenkins and Carroll therefore proposed a reporting framework for photomedicine parameters as early as 2011. Later, Hadis and colleagues showed that a substantial proportion of published PBM research still failed to adequately describe or directly measure key radiometric parameters.

This is where the distinction begins between a panel that merely offers many adjustable settings and a system that can explain what those settings physically mean and what science actually allows us to infer from them biologically.

The most useful question is therefore not:

Which frequency is best?

A better question is:

What light stimulus is the tissue actually receiving, and do we have good reason to believe that this particular stimulus is appropriate for the biological objective we are pursuing?

Photobiomodulation is not a single variable

Red light therapy is often simplified into one of two numbers: device output or joules per square centimetre. Neither one, by itself, adequately describes the biological stimulus.

Light must first interact with molecules and structures capable of absorbing photons or responding to the consequences of photon absorption. Among the mechanisms most commonly discussed in PBM are mitochondrial and redox signalling, cytochrome c oxidase, nitric oxide, reactive oxygen species, changes in membrane potential, calcium signalling and light- or heat-sensitive ion channels.

The contemporary mechanistic picture therefore cannot be reduced to the statement that “red light produces ATP”.

A more accurate interpretation is that PBM acts as a regulatory stimulus capable of entering cellular signalling networks, with the resulting response depending on both the parameters of exposure and the biological state of the tissue.

For practical dosimetry, this means that we must distinguish at least:

wavelength, irradiance, exposure time, radiant exposure, illuminated area, beam profile, treatment distance, pulse structure, number of sessions and the intervals between them.

That is why rigorous PBM reporting frameworks do not define treatment with a single number called “dose”. They require a set of parameters, because without them the intervention cannot be reliably reproduced.

Irradiance and radiant exposure: two quantities that must not be confused

Irradiance describes optical power delivered per unit area. It is commonly expressed in mW/cm².

Radiant exposure, often called fluence or energy density in the PBM literature, describes energy delivered per unit area and is expressed in J/cm².

For continuous light, the simplified relationship is:

radiant exposure = irradiance × time

If the skin receives 50 mW/cm² for 200 seconds, the mathematical radiant exposure is 10 J/cm².

That does not mean, however, that every exposure of 10 J/cm² produces the same biological effect.

Ten joules per square centimetre can be delivered through a relatively short exposure at high irradiance or through a longer exposure at lower irradiance. The total energy may be identical while the biological response is not.

Experimental work and dose-response reviews have repeatedly shown that PBM depends not only on total energy but also on how that energy is delivered in time.

This is why J/cm² alone should never be treated as a complete description of dose.

There is another layer of complexity. The energy measured at the skin surface is not automatically the energy received by deeper target tissue. A proportion of incoming photons is reflected, a proportion is absorbed superficially and a proportion is scattered while travelling through tissue.

Wavelength, pigmentation, tissue thickness, tissue composition, optical geometry and target depth all influence how much light ultimately reaches the biological structure of interest.

For that reason, it is often more accurate to speak about surface radiant exposure rather than pretending that a surface measurement represents the true dose delivered to a muscle, joint or deeper anatomical target.

Why “more” does not automatically mean “better” in photobiomodulation

One of the foundational concepts in PBM is the biphasic dose response.

If the stimulus is too weak, it may fail to produce a measurable biological response. Within an appropriate range, increasing the stimulus may strengthen the response. Beyond that range, however, increasing irradiance, exposure time or total energy does not necessarily produce greater benefit and may reduce or even reverse the desired response in some experimental models.

This phenomenon has been repeatedly described in cellular and animal research.

Classic experiments with cortical neurons, for example, demonstrated that the same wavelength delivered at different radiant exposures could produce different mitochondrial responses. A moderate range could be stimulatory while much higher exposure could become less effective or inhibitory.

WALT publications similarly emphasize tissue- and indication-specific dose titration and caution that energy density alone does not reproduce a successful PBM protocol unless power density, time and other treatment parameters are also known.

For the user, the principle is simple but fundamental:

A more powerful panel is not automatically a better panel, and a longer session is not automatically a better session.

What pulse frequency in hertz actually means

In pulsed operation, light is not continuously emitted. It is repeatedly switched on and off.

At 10 Hz, the system produces ten pulse cycles per second. At 40 Hz, forty cycles. At 100 Hz, one hundred.

Frequency alone does not tell us how long the light remains on within each cycle.

For that, we need another parameter:

duty cycle.

Frequency, pulse width and duty cycle together describe the temporal structure of pulsed light.

At 10 Hz and a 50% duty cycle, a simple square-wave system could remain ON for approximately 50 milliseconds and OFF for approximately 50 milliseconds during each cycle.

At 100 Hz with the same 50% duty cycle, each cycle is ten times shorter: approximately 5 milliseconds ON and 5 milliseconds OFF.

Average delivered energy may be similar.

The temporal structure is completely different.

That is why frequency is not a synonym for dose.

It is better understood as the temporal organisation of dose.

Duty cycle: 50% does not mean 50% of the biological effect

Duty cycle describes the proportion of time during which the light source remains active during pulsed operation.

At 50% duty, the light is ON for half of the total pulse interval.

At 75% duty, it is ON for three quarters of the interval.

For a simple square-wave pulse, with stable peak irradiance and all other parameters unchanged, average irradiance can be approximated as:

average irradiance ≈ peak irradiance × duty cycle

If peak irradiance is 100 mW/cm² and duty cycle is 50%, average optical output over time will be approximately 50 mW/cm².

That does not mean the biological response will be 50%.

Duty cycle is a physical descriptor of temporal output. Biological response depends on the light actually absorbed and on how the biological system responds to its temporal pattern.

The simplified calculation is also only valid if the real waveform and true peak irradiance are known.

Commercial devices can regulate “intensity” in different ways. Some reduce current. Some use PWM. Some combine current control with pulse-width modulation. If an intensity-control PWM system is then combined with a separate therapeutic pulse mode, the value displayed on the screen may no longer adequately describe the real optical output.

Therefore, 50%, 75% or 100% on a display must never automatically be assumed to mean 50%, 75% or 100% of real radiant output.

It must be measured.

This is particularly important because reliance on unverified manufacturer-stated values has been identified as a major methodological weakness in the PBM literature. In a systematic analysis of radiometry reporting, most of the reviewed studies did not describe how optical measurements had been made, and many failed to report essential parameters such as irradiance, pulse frequency, beam area or radiant energy.

Continuous wave or pulsed light: is pulsing biologically superior?

The current evidence does not support a universal yes or no.

Hashmi and colleagues reviewed the literature comparing continuous-wave and pulsed PBM and found evidence that pulsed light can produce a different or sometimes more favourable response in certain experimental contexts.

At the same time, the authors concluded that available data were insufficient to establish universal rules linking a particular pulse frequency with a particular disease, tissue state or biological objective.

Later reviews of PBM parameters reached a similar conclusion.

Some comparative studies favoured pulsed illumination.

Others found no meaningful difference.

In some situations, continuous wave performed better.

Most importantly, the literature does not support a reliable universal equation such as:

specific frequency = specific biological state

This matters enormously.

Not because frequency is biologically meaningless.

But because its meaning is context-dependent.

The effect of pulse frequency depends on wavelength, irradiance, pulse width, duty cycle, target tissue, penetration depth, biological status, exposure time and probably additional variables that were not consistently controlled in much of the historical literature.

Why it is not scientifically justified to say “3 Hz = sleep”

Commercial frequency maps are increasingly common.

They may assign:

3 Hz to sleep,

7 Hz to relaxation,

10 or 15 Hz to regeneration,

30 Hz to recovery,

40 Hz to the brain.

These maps look extremely precise.

Their scientific foundation is often far less precise than their visual presentation.

PBM studies have used pulse frequencies ranging from single-digit hertz to tens, hundreds and thousands of hertz, with positive outcomes reported across very different parameter combinations.

That heterogeneity makes it impossible to conclude that one frequency carries a universal biological meaning independent of wavelength, irradiance, duty cycle, treatment site and tissue state.

A precise-looking number can therefore create pseudo-precision.

If a program is called Recovery and uses 30 Hz, that does not prove that 30 Hz is an optimal “recovery frequency”.

If a Sleep program uses 3 Hz, that does not automatically make 3 Hz a “sleep frequency”.

A scientifically responsible protocol must distinguish between:

a frequency that happened to be used in a particular experiment

and

a frequency that has been demonstrated to be optimal for a particular biological objective.

Those are fundamentally different claims.

The 40 Hz example: a highly interesting signal, but not a universal answer

Forty hertz is a useful example of how PBM research should be interpreted.

In a 2023 randomized sham-controlled study, 56 healthy young adults received transcranial photobiomodulation using 660 or 850 nm light.

The study compared sham stimulation, continuous wave, 40 Hz pulsed stimulation and 100 Hz pulsed stimulation over eight minutes at an average power density of 250 mW/cm².

The pulsed groups showed improvements in some measures of vigilance, attention and memory, together with changes in EEG activity.

An earlier randomized pilot study using pulsed 810 nm transcranial PBM at 40 Hz also demonstrated modulation of neural oscillations measured by EEG.

These findings are scientifically interesting.

They do not mean that a 40 Hz whole-body panel automatically improves Focus.

These were transcranial PBM interventions.

The optical geometry was different.

The irradiance was different.

The anatomical application site was different.

The target organ was different.

Taking 40 Hz from a transcranial study, placing it into a whole-body preset and claiming an equivalent effect would be exactly the kind of extrapolation TANVEA wants to avoid.

Forty hertz can therefore be viewed as a meaningful neurobiological research signal.

It should not be treated as a universal biological constant.

Another important distinction: pulse frequency is not treatment frequency

The word “frequency” is used in more than one way in PBM research.

Recent randomized studies in facial rejuvenation, for example, have examined frequency of application — how many treatment sessions are performed per week.

In a 2025 clinical trial, researchers used a 660 nm LED mask at approximately 6.4 mW/cm², around 8 J/cm² radiant exposure and a treatment duration of 21 minutes.

They compared two versus three applications per week.

That is treatment frequency.

It is not pulse frequency in hertz.

These concepts must not be confused.

10 Hz means ten pulse cycles every second.

Three sessions per week describes how often the entire treatment exposure is repeated.

Biologically, these operate on completely different timescales.

The same panel, a different distance, a different dose

Parameters from large LED panels cannot be interpreted independently of treatment distance.

Moving closer can substantially increase local irradiance and may reduce field homogeneity.

Moving farther away generally lowers local irradiance while allowing the light cones from neighbouring LEDs to overlap more extensively, potentially improving spatial uniformity.

That is why a specification such as:

“150 mW/cm²”

is insufficient.

We need to know:

where the measurement was taken,

at what distance,

at which location in the light field,

with what detector,

and whether the value represents peak, local or spatially averaged irradiance.

Methodological PBM publications repeatedly point out that beam area at the target surface may differ from the size of the emitter aperture and that optical power should be experimentally measured rather than simply copied from a manufacturer specification.

For large LED matrices, it is also inappropriate to mechanically apply the elementary inverse-square law as though the panel were a single point source.

A large panel is an extended source composed of many overlapping emitters.

For practical dosimetry, the more meaningful approach is to construct a real spatial irradiance map at the intended treatment distance.

Light-field homogeneity is part of dose

Imagine a panel for which a meter placed at the centre measures 100 mW/cm².

If the left side of the field delivers 55 mW/cm² and the right side delivers 140 mW/cm², the statement “this panel delivers 100 mW/cm²” does not describe whole-body exposure very well.

The same problem applies to spectral homogeneity.

A panel may have the correct overall RED-to-NIR ratio while concentrating 630 nm in one region and 850 nm in another.

Different parts of the body would then receive different spectral stimuli.

This is why LED count alone is not enough.

The light field itself must be engineered.

This becomes increasingly important in multispectral systems.

The more wavelengths a system contains, the less informative the total LED count becomes and the more important it is to understand how the wavelengths are spatially distributed, independently controlled and radiometrically calibrated.

A multispectral program is not a single frequency

When a device operates with multiple wavelength groups, a preset cannot be adequately described by one number in hertz.

A program is actually a simultaneous state of multiple variables.

It may use:

a higher output of primary NIR wavelengths,

a lower contribution from supporting RED wavelengths,

a defined exposure time,

continuous or pulsed operation,

and a specified treatment distance.

The biological character of the protocol arises from the combination, not from one isolated parameter.

This is why building the marketing identity of a preset around its pulse frequency can be misleading.

“30 Hz Recovery” sounds highly technological.

But unless we can show why 30 Hz specifically improves the value of that spectral and dosimetric architecture, continuous wave may be scientifically more defensible than introducing precision the evidence does not yet support.

A preset should be physically calibrated before it is biologically named

This is one of the most important principles behind the TANVEA approach to light-system development.

If we want to create a mode called Recovery, Skin, Rejuvenation or Focus, selecting percentages on a screen is not enough.

First, we need to know what those percentages mean physically.

If the interface displays:

L3 = 70%

we need to know the actual irradiance generated by L3 at 70% output at the recommended working distance.

If the program uses pulsing, we need to know:

waveform,

peak irradiance,

frequency,

pulse width,

duty cycle.

If it uses multiple wavelengths, we need to know the real radiant contribution of each spectral group.

If the device is a large panel, we need to know whether the field remains homogeneous from centre to edge.

If the program lasts ten minutes, we need to know the radiant exposure produced during those ten minutes.

And if LED output changes as the device heats up during the session, we need to know that too.

Jenkins and Carroll specifically recommend measuring optical output not only initially but also during experimental operation because light-source performance can change with temperature and device age.

Only when these layers are known does a preset stop being merely a software button.

It begins to become a calibrated protocol.

Why continuous wave may be scientifically more precise than a “sophisticated” pulsed preset

This may sound counterintuitive.

Continuous wave is technologically simpler.

Pulsing looks more advanced.

But scientific quality is not determined by the number of available features.

If, for a particular use, we have credible evidence for wavelength, irradiance and radiant exposure under CW conditions, but no convincing evidence anchor for an exact setting such as 18 Hz with a 65% duty cycle, then CW may actually be the more scientifically precise choice.

This does not mean pulsing is biologically irrelevant.

It means TANVEA distinguishes between:

a technical capability of the device

and

a biological claim.

A device may technically generate hundreds or thousands of pulse frequencies.

TANVEA does not need to label all of them as biologically optimised protocols.

What we currently know about whole-body photobiomodulation — and what we do not

This caution is particularly important for whole-body PBM.

Local photobiomodulation has been investigated in a broad range of clinical and experimental contexts including muscle function, pain, skin, wound healing and exercise performance.

A 2024 meta-analysis, for example, found that pre-exercise local PBM can improve some measures of muscular endurance and recovery, although outcomes depend on population and protocol.

Whole-body PBM currently has a much smaller evidence base.

A 2025 systematic review identified only five relevant studies involving a total of 105 physically active participants.

Two studies found signals suggesting improvements in sleep quality.

None of the five demonstrated convincing benefits for fatigue biomarkers or exercise performance/recovery.

The authors therefore explicitly cautioned against treating whole-body PBM as simply a larger version of local PBM.

This is an important observation for the entire large-panel market.

Illuminating more skin does not automatically produce a larger biological effect.

Dose should follow the target tissue

Although there is no single universal J/cm² value appropriate for all PBM applications, current literature supports the principle that different target tissues call for different wavelength and dosimetry strategies.

Recent reviews in skeletal-muscle regeneration, for example, summarize that superficial tissues are frequently targeted with red wavelengths around 630–660 nm, whereas deeper musculoskeletal targets more commonly involve NIR wavelengths in the approximate 810–980 nm range.

The same reviews also emphasize that pulsed illumination remains scientifically interesting but insufficiently standardized.

This does not mean those wavelength ranges constitute a universal prescription for every device.

The broader principle is more important:

Spectral and dosimetric architecture should follow the biological target, not the marketing need to activate every available feature.

Why TANVEA does not want “magic hertz”

TANVEA Light Systems™ does not treat pulse frequency as a sales gimmick.

If we use pulsing, there should be a reason.

If that reason is not sufficiently established, CW is not a technologically inferior option.

If we use a research anchor such as transcranial 40 Hz PBM, we must also state clearly where that evidence came from and where it cannot automatically be extrapolated.

If we use a defined duty cycle, we need to understand what it does to average optical output.

If a preset has a specific biological name, its physical architecture needs to be reproducible.

And if a parameter cannot be scientifically defended, we would rather not present it as biologically precise.

That may be less dramatic than a screen filled with highly specific frequencies.

But it is much closer to science.

From a technical setting to a real dose

A truly sophisticated PBM system should be capable of answering, for every protocol:

Which wavelengths are being emitted?

What is the measured optical output of each spectral group?

What irradiance reaches the skin at the recommended working distance?

How uniform is the field?

How linear is the dimmer response?

What waveform does the pulse generator produce?

What are the peak and average irradiances?

What is the duty cycle?

How does output change as the panel warms up?

What radiant exposure does the skin receive during the complete session?

And what scientific evidence supports the biological objective used to name that program?

Only when these questions can be answered does a preset become more than a software setting.

It becomes a calibrated light protocol.

In summary

Photobiomodulation is not a technology in which biological effect can be inferred from a single number.

Hertz do not tell us what dose the tissue received.

Duty cycle does not tell us what percentage of the biological effect will occur.

J/cm² without irradiance and time does not describe how energy was delivered.

An intensity percentage on a display does not necessarily represent the same percentage of optical output unless this relationship has been measured.

And a mode called Recovery or Sleep does not become scientifically optimised merely because it includes a specific pulse frequency.

PBM is a multidimensional light stimulus.

Its character is shaped by spectrum, irradiance, time, spatial field, pulse structure, duty cycle, distance, repetition and the biological context of the organism.

That is why TANVEA does not define a high-quality Light System as a panel with the greatest possible number of adjustable parameters.

A high-quality system is one that can measure, explain and reproduce its parameters — and use them only where their physical and biological meaning makes sense.

Because the highest form of technological precision is not the number of values shown on a display.

It is knowing which numbers actually matter — and which merely look precise.

---

FAQ

Is pulsed light better than continuous light?

Not universally. Some experimental and clinical studies have favoured pulsed illumination, others have found no meaningful difference, and in some circumstances continuous wave has performed better. Current evidence does not support a general rule that pulsed PBM is always more effective.

Is there an optimal pulse frequency for regeneration?

No single universal frequency has been established as optimal for regeneration independently of wavelength, irradiance, duty cycle, target tissue and other parameters. Reviews of pulsed PBM specifically caution that a reliable universal map linking individual frequencies to individual biological states cannot currently be created.

Does a 50% duty cycle mean half the dose?

For a simple square-wave pulse with unchanged peak irradiance, it approximately halves average optical output over time. It does not automatically mean half the biological effect. Real output also needs to be measured because electronic control may not be perfectly linear.

Is J/cm² the most important parameter in red light therapy?

It is important, but insufficient on its own. The same radiant exposure can be delivered through very different combinations of irradiance and time, and the biological response may not be identical. Rigorous PBM reporting therefore includes wavelength, optical power, irradiance, beam area, pulse parameters, exposure time, treatment site and session intervals.

Why does treatment distance matter so much?

Because it changes both irradiance and field homogeneity. Two ten-minute sessions performed at different distances may therefore represent substantially different exposures.

Is 40 Hz scientifically supported?

Forty hertz has interesting experimental evidence, particularly in transcranial PBM and neuromodulation. Randomized studies have reported EEG changes and improvements in certain cognitive measures. These findings should not, however, be automatically extrapolated to whole-body PBM.

Does a more powerful panel mean a shorter and better treatment?

Not automatically. Higher irradiance can deliver the same radiant exposure in less time, but PBM follows a biphasic dose-response pattern. Excessively high irradiance or total exposure may not produce a better biological outcome.

---

Scientific references

1. Jenkins PA, Carroll JD. How to report low-level laser therapy (LLLT)/photomedicine dose and beam parameters in clinical and laboratory studies. Photomed Laser Surg. 2011;29(12):785–787. PMID 22107486.
2. Hadis MA, et al. The dark art of light measurement: accurate radiometry for low-level light therapy. Lasers Med Sci. 2016;31:789–809. PMID 26964800.
3. Huang YY, Sharma SK, Carroll J, Hamblin MR. Biphasic Dose Response in Low Level Light Therapy – An Update. Dose Response. 2011.
4. Hashmi JT, et al. Effect of Pulsing in Low-Level Light Therapy. Lasers Surg Med. 2010;42(6):450–466. PMID 20662021.
5. Zein R, Selting W, Hamblin MR. Review of light parameters and photobiomodulation efficacy: dive into complexity. J Biomed Opt. 2018;23(12). PMID 30550048.
6. Freitas LF, Hamblin MR. Proposed Mechanisms of Photobiomodulation or Low-Level Light Therapy. IEEE J Sel Top Quantum Electron. 2016.
7. WALT. Photobiomodulation therapy in management of cancer therapy-induced side effects: position paper 2022.
8. Zomorrodi R, et al. Pulsed Near Infrared Transcranial and Intranasal Photobiomodulation Significantly Modulates Neural Oscillations. 2019. PMID 31004126.
9. Pulsed transcranial photobiomodulation generates distinct beneficial neurocognitive effects compared with continuous wave transcranial light. Lasers Med Sci. 2023. PMID 37668791.
10. Li BM, et al. Can pre-exercise photobiomodulation improve muscle endurance and promote recovery? A meta-analysis of randomized controlled trials. Lasers Med Sci. 2024. PMID 38758297.
11. Álvarez-Martínez M, Borden G. A systematic review on whole-body photobiomodulation for exercise performance and recovery. Lasers Med Sci. 2025;40:55. PMID 39883205.
12. Bragato EF, et al. Role of photobiomodulation application frequency in facial rejuvenation: randomized, sham-controlled, double-blind clinical trial. Lasers Med Sci. 2025. PMID 40167796.
13. Lee SY, et al. A prospective, randomized, placebo-controlled, double-blinded, split-face clinical study on LED phototherapy for skin rejuvenation. J Photochem Photobiol B. 2007.
14. Barolet D, et al. Importance of pulsing illumination parameters in low-level-light therapy. J Biomed Opt. 2010. PMID 20799848.

---

Safety note

This article is intended for educational purposes and does not replace individual medical evaluation, diagnosis or treatment. Photobiomodulation should be used according to the technical parameters of the specific device and an appropriate usage protocol. People with photosensitivity, those taking photosensitising medication, individuals with significant eye disease, active medical conditions, pregnancy, or uncertainty about whether PBM is appropriate for them should discuss its use with a qualified healthcare professional.

Do not stare directly into high-intensity LED sources. Follow the manufacturer’s recommendations concerning eye protection, treatment distance and exposure time.

If you have questions about photobiomodulation, dosimetry or the correct use of TANVEA Light Systems™, contact us. Our goal is not simply to deliver light, but to help people understand what kind of light stimulus the body is actually receiving.