Blog 

Start with a comparison of conditions

Irradiance describes optical power arriving per unit area at a surface. It is not electrical input power, LED count or total optical output. A value in milliwatts per square centimetre (mW/cm²) therefore needs a defined measurement surface and location. [1]

Before comparing numbers, place both reports in the same worksheet. For each field below, record the original entry, its report page and whether it matches. Leave a blank field marked unknown; do not replace it with a setting from a brochure.

This is a practical document-review checklist, not a universal test standard. The FDA’s January 2023 PBM draft guidance illustrates why wavelength, source-to-site distance, output mode, irradiance and pulse characteristics need separate description. It concerns specified US PBM medical devices and remains a nonbinding draft, not a salon operating protocol. [2]

What an apparently higher number can hide

Hypothetical example, not supplier data: Report A states 50 mW/cm² at the central point of a target plane. Report B states 400 W/m² as an area mean over a mapped field. Converting B gives 40 mW/cm², but that does not establish that A delivers more across the field. The spatial statistics differ, and distance, spectrum and operating mode still need matching.

Request a centre-point result and a field map for both devices under matched conditions. If their intended geometries differ, preserve that difference and compare each report with its own intended configuration as a separate question. Do not quietly label unlike setups equivalent.

Normalize units without changing the meaning

The following conversions are dimensional arithmetic, not treatment recommendations. They apply only when the reports describe the same quantity:

• 1 W/cm² = 1,000 mW/cm².
• 1 mW/cm² = 10 W/m²; divide W/m² by 10 to obtain mW/cm².
• A spectral value in W/m²/nm is per wavelength interval. It is not a band-integrated W/m² value; the spectrum must be integrated over stated wavelength limits before those quantities can be compared. [8]

CIE defines irradiance as a surface quantity. NIST distinguishes radiometric units from photometric units, which are weighted for human vision. Lux and lumens therefore cannot be converted into a general LED irradiance figure using a single spectrum-independent factor. Brightness photographs do not replace a radiometric report. [1] [6]

Separate spatial averaging from time averaging

“Average” needs two labels: average over which area and average over which time interval? An area mean can still describe output during the on-period of a pulse. A time average can still be a reading at only one point. Ask the laboratory to define both rather than relying on the word alone.

For an ideal rectangular pulse train with zero output between pulses, the time average equals the on-period irradiance multiplied by the duty-cycle fraction. That arithmetic cannot reconstruct an unspecified waveform. Request the actual pulse definition and the instrument’s method for measuring or deriving the average; frequency alone is insufficient.

Do not allocate a combined figure to every colour

If a report lists a red-plus-near-infrared total, ask for each band’s contribution while that combined mode is operating. Separate red-only and near-infrared-only tests do not establish the combined-mode result unless the report demonstrates that the same channel outputs are maintained.

Adding band values is defensible as arithmetic only when they cover non-overlapping spectral ranges and share the same geometry, spatial/time averaging and simultaneous operating state. Otherwise, retain the reported values separately and request a matched combined-mode measurement. These are comparison rules; they are not findings about any supplier.

Keep radiant exposure separate from irradiance

Radiant exposure is the time integral of incident irradiance. For constant irradiance, or a valid average over the same exposure interval, J/cm² = mW/cm² × seconds ÷ 1,000. [7] This describes incident exposure at the measurement surface. It does not establish absorbed tissue energy or clinical equivalence. Do not use a supplier comparison to invent session times, distances or frequencies.

Ask what supports the measurement

NIST’s spectral irradiance calibration reference describes measurement procedures, uncertainty and links to SI units. Its abstract does not validate a particular salon LED or a supplier’s test method. [3] For the report in front of you, request:

• Instrument identity: meter or spectroradiometer model, serial number and detector/accessory configuration.
• Calibration evidence: certificate identifier and date, covered quantities and wavelength range, and any corrections applied.
• Test details: laboratory, report identifier/revision, test date, setup diagram, stabilization conditions and repeated readings.
• Method suitability: how the detector’s spectral and angular response, measurement range and temporal response suit the tested mode.
• Result uncertainty: uncertainty for the reported measurement, its units or percentage basis, and the contributing factors considered.

“Traceable to NIST” needs supporting documentation. NIST explains traceability through a documented calibration chain with associated uncertainties; owning a calibrated instrument does not automatically make every subsequent measurement traceable. Traceability also does not guarantee that uncertainty is small enough for the intended comparison. [4]

Read a ± figure before treating a difference as meaningful

Ask whether a quoted ± value is repeatability, instrument accuracy, standard uncertainty or expanded uncertainty. These labels should not be substituted for one another. NIST expresses expanded uncertainty as U = k × u_c, where k is a coverage factor and u_c is combined standard uncertainty. Request the stated factor and coverage interpretation. [5]

Hypothetical uncertainty example: Matched reports show 40 ± 4 and 42 ± 4 mW/cm², each with the same stated expanded-uncertainty basis. The headline difference is 2 mW/cm²; the quoted intervals overlap. That alone neither proves equivalence nor settles which output is higher. Ask the laboratory to assess uncertainty in the difference, including any shared calibration effects, before making a superiority statement.

Keep measurement uncertainty separate from variation across units. A report on one identified sample does not tell you the spread across a production batch. Ask what sampling or quality-control evidence supports a claim about the units a salon will receive.

Resolve gaps before making a comparison

Use three document-review outcomes. These describe the evidence available, not the quality or honesty of the supplier:

1. Comparable for a stated optical question: the quantity, conditions and uncertainty basis align. State exactly which output comparison the data support.
2. Clarification required: a missing field might be resolved by the underlying report or laboratory explanation. Keep that field unknown until documented.
3. Not comparable from current documents: conditions differ or a material conflict remains. Request a matched test; do not force a ranking.

If a brochure and report disagree, record both values with their document dates and page numbers. Ask which report revision applies to the exact model and mode, and whether the difference reflects settings, units, a changed configuration or a correction. No response means the conflict remains unresolved; it is not evidence of fraud.

A concise clarification request

Please provide the optical report for the exact offered configuration, including its test date and revision. Confirm the distance reference, target-plane geometry, mapped area and averaging method; spectrum and integration limits; active channels and continuous/pulsed settings; instrument and calibration details; and measurement uncertainty. Please identify whether each figure is a point value, area mean, pulse peak or time average, and reconcile any conflicting figures by report page.

Preserve the reply beside the report and comparison worksheet. A documented explanation can resolve a label; an unexplained assertion should remain an open item.

Questions worth asking before accepting a report

Is an older test report automatically unusable?

No fixed age cutoff is established here. Ask whether the tested configuration is still current, whether the instrument’s calibration was applicable at the test date, and whether later changes affect the result. Age alone cannot answer those questions.

Can a laboratory logo establish independence?

A logo is not enough to identify who commissioned, performed or interpreted the test. Request the original report and verify its identifier with the issuing laboratory if authenticity is uncertain. Record the tested sample and scope; do not imply independence solely from presentation.

Can one flat-plane map describe exposure across a face?

Keep the map’s geometry explicit. It describes the sampled plane and setup, not a measured distribution over every facial contour. Ask for evidence suited to the intended device arrangement before treating the map as a face-wide exposure estimate.

Does comparable output make two salon services equivalent?

No. An optical match answers a measurement question. It does not supply missing evidence about the intended outcome, population or clinical protocol. Keep clinical claims and device suitability as separate reviews.

The useful result is a defensible comparison

A completed worksheet should show what matches, what differs and what remains unknown. Accept a narrow optical conclusion when the documents support it. Where they do not, the next step is clarification or matched testing—not a larger claim.