Blog 

What J/cm² measures—and what it leaves out

For light arriving at the skin, the precise term is incident radiant exposure: joules of optical energy per square centimetre, accumulated over the exposure. Photobiomodulation (PBM) papers often call it “fluence,” “energy density” or “dose.” Here, J/cm² refers to exposure at a specified surface, not energy absorbed by a target beneath it. [1]

With constant irradiance—the optical power arriving per unit area—the calculation is:

Radiant exposure (J/cm²) = irradiance (mW/cm²) × exposure time (seconds) ÷ 1,000

If output varies, exposure is the irradiance accumulated over time. A time-averaged irradiance can be used when it accurately represents that interval. This is a physical calculation, not a prediction of a biological response. [1]

The products are equal. The exposures have different intensity and duration, and the calculation contains no wavelength information. It also says nothing about whether that value represents the centre of a beam, an area average or every point across the illuminated region.

There are three different comparison questions: Are the calculated numbers equal? Are the light exposures physically comparable? Have comparable clinical outcomes and safety been demonstrated? Answering the first does not answer the other two. Even a close physical match would still need evidence relevant to the intended outcome and population.

Why intensity and time must remain separate

The idea that increasing irradiance and reducing duration proportionally will preserve the biological effect is called exposure reciprocity. The arithmetic identity does not establish that relationship in living tissue. A radiometry review explains why irradiance and duration should be reported individually rather than reduced to one exposure number. [1]

One original LED study examined this question in a mouse pressure-ulcer model. Its abstract reports that varying irradiance and exposure time while targeting a specified energy density changed phototherapy outcomes. That finding challenges an unrestricted assumption of reciprocity. It does not identify a dose for human facial skin or prove that every pair of equal-dose LED protocols produces different results. The full methods were not verified for this article. [3]

The appropriate conclusion is therefore narrow: equivalence must be tested in the relevant setting, not inferred from multiplication. Keep wavelength, irradiance and duration in separate fields when reading a trial. A missing field is a limitation in the comparison, not an invitation to fill it with a convenient value.

Pulses and spatial distribution change the comparison

Peak irradiance is not average irradiance

A continuous source emits throughout the exposure. A pulsed source changes output over time. Pulse frequency, pulse duration, shape and duty cycle describe aspects of that timing; duty cycle is the fraction of each cycle during which the source is on. Frequency alone does not describe the complete waveform. FDA’s January 2023 PBM draft guidance lists multiple pulse descriptors and average power separately. It remains non-binding draft guidance for specified US medical-device submissions, not a consumer treatment standard. [2]

For an ideal rectangular pulse train with zero output between pulses, average irradiance equals on-pulse irradiance multiplied by duty cycle. For example, 20 mW/cm² during each pulse at a 50% duty cycle averages 10 mW/cm². Over 600 seconds, that is 6 J/cm². These are hypothetical values; they demonstrate a calculation, not a safe or effective setting. [1]

If a specification already gives the time-averaged irradiance, multiplying it by duty cycle again would undercount exposure. If it gives peak irradiance, using that figure as though it were continuous would overcount it under the stated ideal assumptions. Ask which value is reported and how the waveform was measured. Matching the correctly calculated exposure still does not demonstrate that pulsed and continuous protocols have equal outcomes.

An area average can hide different local exposures

Consider two hypothetical exposure maps with the same area average of 6 J/cm². One is uniform at 6 J/cm². The other delivers 3 J/cm² over one half and 9 J/cm² over the other half. Their averages match, but the local exposures do not. This is a mathematical illustration; no benefit or harm is assigned to these numbers.

Beam profile, illuminated area and measurement geometry matter when describing delivery. A central measurement cannot establish uniform coverage across the entire treatment area. Measurements at an emitter aperture also cannot simply stand in for measurements at the skin in a different setup. The radiometry review discusses target distance, geometry and spatial measurement as separate requirements. [1]

For a curved surface such as the face, ask whether the reported exposure describes the actual intended placement and area. Do not assume that a flat measurement plane represents every facial contour, or alter distance to recreate a study number without the device’s supporting instructions.

Energy at the surface is not energy absorbed by a target

The skin surface is a measurement boundary. Light reaching that boundary subsequently interacts with tissue; scattering changes its path, and absorption transfers energy to tissue components. The amount arriving at a particular target is therefore a different question from the incident J/cm². Tissue scattering and measurement geometry are discussed in the radiometry review. [1]

Wavelength matters because tissue chromophores—components that absorb light—interact with particular parts of the spectrum. Hernández-Bule and colleagues discuss chromophores and proposed PBM mechanisms in their dermatology narrative review. That mechanistic discussion does not establish a clinical benefit for an unspecified LED device. [4]

A combined red and near-infrared exposure also needs a spectral breakdown. Hypothetically, “6 J/cm² total” could mean 6 J/cm² from one band, or 3 J/cm² from each of two bands. Those descriptions have equal total incident energy but different spectral composition. An LED count or a list of nominal wavelengths cannot tell you the energy contribution of each band.

There is no absorption percentage supplied here that converts a surface reading into energy absorbed by a facial target. Such a conversion requires an appropriate tissue model or measurement with its assumptions stated. Neither a calculator nor a proposed cellular mechanism closes that evidence gap.

Audit the protocol before copying a study’s dose

Use the following checklist to assess comparability, not to design a treatment. Jenkins and Carroll’s reporting paper identifies wavelength, power, duration, beam area at the target, pulse parameters, anatomical location and treatment scheduling as necessary descriptors. Its accessible abstract supports that reporting framework; its full text was not verified here. [5]

The checklist separates exposure description from outcome evidence. An abstract may establish what was studied without supplying enough methods to reproduce it. Likewise, matching technical fields does not turn a study of one outcome into evidence for another. A direct equivalence claim needs an appropriate comparison and analysis; “no statistically significant difference” alone is not proof of equivalence.

Common shortcuts that do not establish equivalence

• “Double the intensity and halve the time.” This preserves the calculated exposure under constant-output assumptions. It does not validate the biological or safety consequences.
• “The same red wavelength means the same protocol.” Duration, irradiance, spectral width, pulse structure and delivery geometry still need comparison.
• “Matching J/cm² transfers a laser study to an LED mask.” A shared unit does not remove differences in source and delivery, or establish population and outcome relevance.
• “Add up the whole course and compare the total.” A cumulative number discards the timing of individual exposures and their intervals. It cannot by itself establish an equivalent regimen. [1] [5]

FAQ

Can two devices have the same J/cm² but different total joules?

Yes. For a uniform exposure, total incident energy equals radiant exposure multiplied by illuminated area. Hypothetically, 6 J/cm² across 10 cm² equals 60 J; across 100 cm² it equals 600 J. These are arithmetic examples, not recommended exposures. For a non-uniform field, total energy must account for the exposure across the whole area. [1]

Does an electrical wattage rating let me calculate skin exposure?

No. Electrical input is not optical irradiance at the skin. A calculation of incident exposure needs relevant optical measurements and duration. The distinction between light-source output and irradiance at a target is central to radiometric reporting. [1]

Does the same unit make LED exposure comparable to IPL or UV?

Only as a unit of energy per area. It does not make the technologies interchangeable. Spectrum, temporal delivery and intended interaction with tissue differ. FDA’s PBM draft explicitly distinguishes its scope from ablative or coagulating light devices, including high-power lasers and intense pulsed light (IPL). Do not use an LED dose number as a protocol for those procedures, UV phototherapy or photodynamic therapy. [2]

Compare the exposure description, then the evidence

J/cm² is useful when its measurement conditions are clear. Preserve the separate parameters, identify what has not been measured, and distinguish a plausible comparison from demonstrated equivalence. If the full protocol or corresponding instructions are missing, the comparison remains incomplete.