Red Light Therapy Manufacturer Testing: How a Professional Factory Tests and Reports Joule Dose

Update date: 2026.5.25 | Reading time: 15 minutes

This article is for buyers, private-label brands, distributors, and clinics evaluating a red light therapy supplier. It describes the manufacturer's testing-and-reporting system — the SOP, instruments, records, and deliverables — not treatment protocols and not a general dose calculation tutorial.

Three sentences will tell you whether you are talking to a manufacturer or a reseller.

"Can you send us your irradiance test report at four distances?"
"Can you provide a one-page joule dose report we can drop into our spec sheet and influencer kit?"
"Which spectroradiometer model did you use, and when was it last calibrated?"

A serious factory has those answers in a folder before you ask. A reseller — even one with a nice catalogue, a real warehouse, and confident marketing copy — improvises, deflects, or quietly offers to "check with the engineer." That conversational moment is more diagnostic than any product photo, any LED-count number, any "150 mW/cm²" headline on a product page.

This article is not a tutorial on how to calculate a joule dose. The arithmetic — irradiance × time ÷ 1000 — is trivial. This article is about the verification system a manufacturer must run before any joule number that comes out of that arithmetic can be defended: the standard operating procedure that controls how the irradiance reading is produced, the instruments and calibration regime that earn the word credible, the locked records architecture that lets every J/cm² claim be traced back to a specific sensor reading on a specific date by a specific tester, and the one-page customer-facing dose report that becomes the deliverable a private-label brand, clinic, or reviewer can actually use.

Red light therapy panel test

If you are buying for a clinic, building a private-label brand, or just trying to separate a credible manufacturer from one that has memorised the right specs, this is the system you are looking for.

Why the testing system matters more than any single number

A red light therapy panel's joule dose is calculated from a stunningly simple equation: irradiance in mW/cm², multiplied by time in seconds, divided by 1000, equals J/cm² at the skin. A school-aged child can do the arithmetic.

What is not simple is the irradiance number that goes into the equation. That number is the output of a testing process, and the credibility of the final J/cm² figure is exactly equal to the credibility of that process. Garbage in, garbage out — except in this case the "garbage" looks like a confident product-page number, which is much harder for a buyer to identify than obvious garbage would be.

A manufacturer's testing system is what converts a raw sensor reading into a defensible specification. It controls who measures, with what instrument, at what distance, in what mode, after what preheat, on what grid, with what archiving, signed off by whom, and traceable to what calibration certificate. A panel without that system behind it can still produce a number — but the number is a marketing artefact, not a measurement.

The closed loop: SOP → Raw data → Review → Customer report

A credible joule claim does not come out of a single measurement. It comes out of a four-stage closed loop in which each stage produces an input to the next, and each stage is auditable from any other stage. Break the loop at any point and the number on the spec page becomes unverifiable.

The point of describing it as a loop, not a sequence, is that each stage references the others. The customer report cites the raw data reference. The raw data reference points back into the SOP under which it was produced. The reviewer signing off in stage 3 is confirming that stage 1 was followed. If a buyer six months after delivery asks "where did this 57 J/cm² come from?", every stage of the loop has to answer the same way — and a manufacturer who has built the system can produce that answer in minutes.

The rest of this article walks through each stage in detail.

The five-stage SOP every serious manufacturer runs

A professional red light therapy lab does not improvise irradiance tests. It follows a controlled sequence, in the same order, every time. The sequence has five stages, and skipping any one of them invalidates the others.

Stage 1 — Preparation

Before the panel is even powered on, three things must be true.

The instrument is in calibration. A spectroradiometer is not a forever-accurate device. It drifts. It needs a current calibration certificate from a recognised lab, dated within the manufacturer's calibration interval (typically annual). A test report that does not state the calibration date is reporting a number with an unknown error band.

The environment is recorded. Ambient temperature and humidity are written into the test record. Stray light is excluded — the test is run in a dark or low-light room, with no reflective surfaces or sunlight angles that would inflate the reading. A test conducted next to a window at midday is not the same test as one conducted in a controlled enclosure, and the difference is not small.

The panel is preheated for 10 to 15 minutes. LED output drifts as the panel warms — typically downward for cheaper drivers without thermal management, and toward a stable plateau for well-engineered panels. A reading taken thirty seconds after power-on is not the panel's steady-state output. It is the panel's transient cold-start output, which can sit 5–15 percent above the value the panel actually delivers across a real user session. Preheat is the boring step everyone wants to skip; it is also the step that separates a steady-state irradiance number from a flattering snapshot.

Stage 2 — Setup

With the panel preheated and the instrument verified, the geometry is locked.

Distance is set with a fixed jig, not held by hand. The standard tier covers 15 cm, 30 cm, 45 cm, and 60 cm — the range real users actually treat at, from close-range high-density use to whole-body coverage. A tape-measured distance read by eye is not repeatable; a machined jig with a positive stop is.

Sensor angle is perpendicular to the panel centre. A tilted sensor returns a depressed reading because of its cosine response. A sensor tilted to catch more light returns an inflated one. The mounting must be deliberate, and the test report must say so.

Mode is declared and locked. Red only, NIR only, or combined mode. Each is measured separately, because the three are not equivalent and do not simply add — combined mode often shows slightly less than the arithmetic sum of the individual bands due to shared driver current and higher thermal load. A panel that publishes only a combined number is hiding the band ratio the buyer needs to evaluate it.

Pulse mode is disabled during baseline irradiance characterisation. Pulsed output is a separate measurement requiring separate documentation, not a way to inflate the steady-state reading.

Stage 3 — Sampling (this is where raw data is born)

Now the panel is read.

The sensor records the centre point, then a grid pattern — a 3×3 grid at minimum for any panel, a 5×5 grid for clinical-grade and premium private-label products. Each grid point is recorded three times and averaged, so a momentary fluctuation does not become the published number.

At each point, the reading is broken down by band: red mW/cm² (typically 620–680 nm), NIR-1 mW/cm² (typically 800–900 nm), and any third band if the panel includes one (940 nm or 1060 nm). Peak wavelength, full-width-at-half-max (FWHM), Effective Energy (EE) coefficient, and flicker percentage are recorded alongside the irradiance values. A spectroradiometer such as the OHSP-350-IRF series exports per-nanometre spectral irradiance to Excel, which is what makes per-band calculation possible from a single measurement.

Irradiance grid mapping

This is the stage where the most fundamental honest-versus-dishonest decision happens. A factory recording only the brightest centre point at the closest distance is selecting its highest number and discarding the rest. A factory recording the full grid at multiple distances is collecting the data needed to publish a real specification — even when most of that data will look less impressive than the cherry-picked peak. The raw export from this stage is the asset that feeds the next two stages of the loop; if it does not exist as a saved file with a reference number, nothing downstream is auditable.

Stage 4 — Calculation

With the raw data collected, the calculation step is mechanical but error-prone.

Units are converted to a single standard, typically mW/cm². If the spectroradiometer exports W/m², the conversion is mW/cm² = W/m² × 0.1. The conversion factor appears explicitly in the test record so a reviewer can audit it.

Joule dose is computed at the panel-average level, not the centre-peak level:

D_avg = t ÷ 1000 × Σ(Eι × Aι) ÷ ΣAι

Where Eι is the irradiance at grid point i, Aι is the area that grid point represents, and t is the session length in seconds. The output is the area-weighted average J/cm² across the panel face — the dose figure that maps to what a real body surface receives, rather than the inflated number a one-cm² centre patch would receive.

Band-specific doses are computed separately:

Dose_band = E_band × t ÷ 1000

So a panel with average red irradiance of 40 mW/cm² and average NIR irradiance of 55 mW/cm² over a 10-minute session delivers approximately 24 J/cm² of red and 33 J/cm² of NIR at the skin, with a combined surface dose of 57 J/cm².

Total incident energy is calculated for the treatment area:

Total J = J/cm² × Area(cm²)

This is the figure useful for system-level comparisons and large-coverage panels, but it should never replace the per-cm² figure on a customer-facing spec sheet. A 100 kJ "total energy" claim sounds impressive, but it does not tell the user what dose any given patch of their skin actually receives.

Stage 5 — Verification, sign-off, and report generation

The final stage is what turns a measurement into a record — and a record into a deliverable.

The raw Excel export is archived under a unique reference number. The completed test sheet is signed by both the tester and an independent reviewer. The instrument serial number, calibration date, and environmental log are attached. The entire package becomes part of the panel's Design History File (DHF) for engineering validation work, or its Device Master Record (DMR) for production-released specifications.

From that record — and only from that record — the customer-facing one-page joule dose report is generated. The report is not written from memory, not summarised from a marketing brief, and not lifted from a competitor's spec sheet. It is a direct extract of the verified measurement, with the conditions and instrument explicitly stated. This is the deliverable that completes the closed loop introduced earlier: SOP → raw data → review → customer report.

Without this stage, a J/cm² number cannot be defended. With it, the number can be traced back to a specific sensor reading on a specific date by a specific tester using a specific calibrated instrument — which is exactly what a clinical client, a private-label brand's compliance officer, or a regulatory reviewer in a destination market will eventually ask for.

The equipment tier: what belongs in a professional test bench

A buyer can tell a great deal about a manufacturer from how they answer a single question: "What did you test it with?"

The professional solution is a spectroradiometer. Instruments such as the OHSP-350S (spectrometer) or SM-206 (solar power meter) — or equivalent devices — feature a wavelength range of 380–1100 nm. This means they are capable of reading not only the standard 660/850 nm bands but also the emerging 1060/1070 nm bands utilised by some next-generation panels. The package includes a calibration certificate as well as an annual calibration plan. These instruments carry a price tag running into the thousands of dollars and require trained personnel to operate. These are the very instruments cited in authoritative test reports.

Irradiance instrument tests the phototherapy panel

Supplementary equipment rounds out the bench. A standard tape measure or, better, a positioning jig for repeatable distance. A tripod or test stand that holds the sensor perpendicular and steady. A thermometer and hygrometer to log environment. A thermal imaging camera to verify thermal stability during long-duration tests. A locked Excel template that takes the spectroradiometer's export and applies the standard formulas without manual override. None of these alone make a credible report, but their absence makes a report incomplete.

What should never stand alone as the basis for a published spec:

A solar meter sold for photovoltaic panel testing. These instruments are calibrated for the broadband solar spectrum, not the narrow LED emission lines of red light therapy panels. As reviewers including Alex Fergus at Light Therapy Insiders have repeatedly documented, solar meters can read 2 to 3 times higher than calibrated spectroradiometers on the same LED panel. A "150 mW/cm²" number from a solar meter is not a measurement error — it is a category error.

Total electrical input power divided by panel area. This calculation is sometimes presented as "the panel draws 900 watts, so it delivers about 50 mW/cm² across its face." It is wrong on multiple levels: it ignores LED efficiency (typically 30–45 percent of electrical input becomes radiant output), it ignores driver losses, it ignores optical losses through diffusers and lenses, and it ignores beam pattern. A panel's input wattage tells you what it costs to run, not what dose it delivers to skin.

A centre-only reading turned into a panel-wide claim. A 95 mW/cm² centre reading on a panel that averages 74 mW/cm² across its face overstates the average user dose by more than 20 percent.

An "X joules in Y minutes" number with no stated distance, no mode, no grid description. This is a slogan, not a spec.

A manufacturer whose entire dose claim rests on any of these is not running a professional testing system. They are running a marketing one.

The data record template — making every joule traceable

A measured joule value is only as defensible as the record it sits in. A professional test bench uses a fixed template with locked formulas and required fields, so a reviewer six months later can reconstruct exactly what happened. The minimum field set:

Two details matter here.

The template should be a locked spreadsheet with formulas pre-built and cells protected. Manual overrides of calculated results — for any reason — invalidate the record. A spec figure that originated as "the engineer typed it into the cell" is not a measurement.

The completed record should feed directly into the Design History File (DHF) for engineering validation and into the Device Master Record (DMR) for production specifications. This is the difference between a test report that supports a marketing page and one that supports an ISO 13485 / MDSAP design-control audit. Buyers operating in regulated channels — medical aesthetic, veterinary, clinical — should specifically ask whether the irradiance test record is part of the DHF/DMR. The answer is binary, and it is informative.

Five categories of error a real lab actively controls

A credible joule number is not produced by being lucky. It is produced by understanding what can go wrong and engineering each source of error out of the result. A professional lab tracks five categories.

Instrument error. Calibration drift, detector cosine response, limited wavelength range. Controlled by annual calibration, choosing a spectroradiometer with a flat cosine response and full 380–1100 nm coverage, and cross-checking against a reference instrument when possible.

Test error. Distance offset, angle offset, ambient light interference, insufficient preheat. Controlled by positioning jigs, dark-room enclosures, and a written setup checklist that is signed off before sampling begins.

Sample error. LED batch variation, driver current drift, lens or diffuser transmittance differences, thermal output decay. Controlled by recording the BOM revision and LED batch on every test, by testing both pre-production and production samples, and by running a thermal stability check that confirms the panel's output is steady at the moment of measurement.

Calculation error. Unit conversion mistakes (mixing W/m² with mW/cm²), reliance on centre-point readings, inaccurate treatment-area estimates, area-weighted-average formulas applied incorrectly. Controlled by the locked-formula template described above and by independent reviewer sign-off.

Use error. User sits at a different distance than tested, skin reflectance differs, session length differs from spec. Controlled at the documentation level — the user manual must state the distance, time, and mode that the published dose corresponds to, so the user can replicate the conditions or knowingly deviate from them.

Every credible test report names the error sources in its uncertainty disclosure: instrument accuracy class, calibration date, number of repeat readings, standard deviation across grid points, and stated distance tolerance. A report without any of this is not stating that the panel has zero uncertainty. It is stating that the uncertainty has not been characterised.

Where joule data sits in a real design-control chain

For manufacturers serving private-label brands or regulated channels, the irradiance and joule data are not just marketing inputs. They are required records in a design-control flow that mirrors ISO 13485 and MDSAP expectations.

A simplified view of the chain:

1. Design Input. Target dose range in J/cm² for a stated distance and session length, wavelength configuration, treatment area, preheat tolerance, uniformity floor. Written into the Product Requirements Document (PRD).
2. Design Output. LED Bill of Materials, optical/structural drawings, driver current settings, control firmware modes, labelling specifications. The "we will build this" deliverable.
3. Design Verification. Spectral measurements, irradiance grids, joule calculations, temperature-rise checks, flicker measurements, stability testing across a defined burn-in period. The "did we build what we said" deliverable.
4. Design Validation. User-scenario testing, customer experience evaluation, user-manual readability, competitor benchmarking. The "does it work in the world we built it for" deliverable.
5. Production Release. Incoming materials inspection on critical components (especially LEDs), in-process QC, finished-product sampling, label-claim cross-check against the test record. The "every unit shipped matches the spec" deliverable.

A buyer asking a manufacturer "where in this flow does your joule number live?" is asking a question that filters strongly. A factory that can point to the design verification record and connect it to a specific PRD target sits at one end of the spectrum. A factory whose joule number lives only on the website sits at the other.

Bad spec versus good report: a side-by-side

The single most useful filter a buyer can apply, before any sample is shipped or any audit is booked, is to put a supplier's published spec language next to what a good test record would actually say. The gap between the two is the gap between marketing and measurement.

The right-hand column does not look as marketable as the left. It looks credible. Those are different goals, and a buyer has to decide which one they are paying for.

What the product specification should actually say

Most red light therapy product pages publish a number and a vibe. "150 mW/cm² irradiance. Up to 300 J/cm² per session. Industry-leading power." These statements look authoritative. They are also unfalsifiable, because they do not state the conditions under which they are true. The same panel measured at a different distance, mode, or grid point produces a different number — and the spec page is silent on which one was chosen.

A specification written from a real test record reads differently. Three example sentences a serious supplier should be able to write:

"At 15 cm distance, combined mode, after 15-minute preheat, the panel's average grid irradiance is X mW/cm². A 10-minute session delivers an area-weighted surface dose of Y J/cm² (red contribution Z₁ J/cm², NIR contribution Z₂ J/cm²). Test instrument: spectroradiometer model OHSP-350S, calibrated [date]. Grid pattern: 9-point. Uniformity: U%."

Each sentence ties the claim to a condition that can be checked. The buyer can replicate the test if they want to. An influencer can verify it without ambiguity. A regulatory reviewer can match the marketing copy to the underlying record. This is what "verifiable spec" actually means in practice.

A manufacturer that allows the kind of language in the left-hand column of the comparison table above onto a spec page is showing the buyer how they will market the panel after the buyer ships it.

The one-page customer joule dose report

The deliverable a serious private-label brand, distributor, or clinic should ask for is a single page that summarises the panel's tested joule dose under specified conditions, with the source data archived behind it. The format matters because the page will be used in three different places — the technical spec section of a customer-facing product page, the technical-question response a clinic sales rep sends to a clinical buyer, and the press kit that ships with a sample panel sent to a reviewer or influencer.

For example:

Product Model: RD 1500 Panel
Test conditions: 15 cm distance | Combined mode | 15 min preheat | 9-point grid average | 10 min session

Average total irradiance	Red dose	NIR dose	Total surface dose
95 mW/cm²	24 J/cm²	33 J/cm²	57 J/cm²

Test instrument: Spectroradiometer OHSP-350S (calibrated 2026-02-14)
Raw data archive reference: RD-LT-2026-001
Test notes: Surface incident dose, calculated from measured irradiance × session time. Different distances, angles, or session lengths produce different doses.

That is the entire useful body of the page. Logo and contact details flank it. A QR code linking to the underlying test record completes it. The page is repeatable across every SKU in a product line, so a buyer comparing four panels from the same supplier reads four versions of the same one-page format — not four marketing pages that omit different variables.

Joule dose report template

This is the deliverable that converts an interested buyer into a confident one. It is also the deliverable that signals to the buyer that the supplier can produce it — which is itself a filter.

Four questions that identify a credible supplier in under five minutes

A buyer evaluating a red light therapy supplier — for an OEM order, a private-label launch, a clinical distribution deal, or a serious bulk personal purchase — can shortcut almost the entire evaluation by asking four questions in sequence.

1. "Can you send me your full irradiance test report at 15 cm, 30 cm, 45 cm, and 60 cm, in red mode, NIR mode, and combined mode, with a 9-point or 25-point grid at each distance?" The credible answer is yes, here it is, attached. The incredible answer is some version of "we have data, let me check what we can share."

2. "What instrument and calibration date were used? Can you attach the calibration certificate?" A spectroradiometer model number, a calibration date inside the last twelve months, and a willingness to share the certificate are the credible answer. A vague reference to "professional testing equipment" is not.

3. "Can you provide a one-page joule dose report we can use in our product page and reviewer kit?" A supplier who has built the system above produces this page from an internal template in under a day. A supplier who has not will improvise or quietly take a week.

4. "How does this test record feed into your design-control file and your production QC?" This question filters strongly between OEMs running a real quality system and traders sourcing finished panels from a third party. A real OEM can describe the DHF/DMR linkage in concrete terms. A trader cannot.

The supplier shortlist filters itself within those four exchanges. There is no need for site visits, audits, or proxy indicators — the supplier's ability to produce the deliverables described above is the indicator.

From "LED count competition" to "verifiable dose competition"

The red light therapy industry spent its first commercial wave competing on the wrong axis. Spec sheets stacked LED count, electrical wattage, and centre-point irradiance peaks. Buyers compared on those numbers, because those were the numbers manufacturers published. Whichever panel had the highest headline figure won the click.

That model has aged badly. Clinicians, sophisticated end consumers, and the better independent reviewers have moved on. They now expect — and increasingly demand — the second tier of competition: verifiable, condition-bounded, audit-grade joule dose. Per-band, per-distance, per-mode, with the test record attached.

A manufacturer competing in the verifiable-dose tier publishes a different kind of product page. It states test conditions. It separates red and NIR. It provides a distance tier. It cites a real spectroradiometer. It offers a one-page customer report on request. It links the customer-facing number to an internal record that survives an audit. It treats the joule figure as an engineering output of a controlled process, not a marketing input.

That competition rewards real engineering. It punishes the panel that looked great in product photos but never had a defensible measurement behind it.

How REDDOT LED supports private-label brands with this system

Everything in this article is what REDDOT LED's internal testing system is designed to produce. For private-label clients, that system translates into specific deliverables:

• A distance-tier irradiance test report for every panel SKU, covering 15, 30, 45, and 60 cm in red, NIR, and combined modes, with 9-point or 25-point grids and band-split data.
• A one-page customer joule dose report, branded to the client where required, that can be used directly on product pages and in influencer/reviewer kits.
• Underlying raw data archive references that the client's compliance or regulatory team can request when needed for destination-market filings.
• OEM/ODM design support that writes target J/cm² ranges into the Product Requirements Document upfront, so the panel is engineered to a verifiable dose target rather than measured after the fact to whatever number it happens to produce.
• Influencer and reviewer sample programs where REDDOT panels can be sent for independent verification, knowing the published spec will match what the reviewer measures — because the same testing standards were applied at the factory.

The product-page parameter section, the customer-support technical reply, the influencer review kit, and the regulatory tech file all draw from the same source data. That is the goal of the system described in this article: a single test record that supports every external use of the panel's joule claim.

Bottom line

A red light therapy manufacturer's testing-and-reporting system is not a backstage detail. It is the actual product behind the product. The panel is a piece of hardware; the panel's specification is a piece of evidence. Without the SOP, the calibrated instrument, the locked record template, the design-control linkage, and the customer-facing one-page report, a J/cm² number on a product page is a guess presented as a fact.

For buyers, the practical implication is straightforward. Ask for the report. Ask for the calibration certificate. Ask for the distance tier and the band split. Ask where the test record lives in the supplier's quality system. The suppliers who can answer all four are running a manufacturing operation. The ones who cannot are running a sales channel. Both ship product. Only one of them ships product whose dose claims you can defend to your own customers.

Request a Light Dose Report

Companion guides

You may also find these helpful:

• Red Light vs Near-Infrared Dose: Why Multi-Wavelength Panels Need Band-Split Joules
• Why do testing distance and uniformity alter the joule dosage of red light therapy devices?
• Why Red Light Therapy Panels Should Not Be Judged by Wattage and LED Count
• How to Calculate Red Light Therapy Dose: A Practical Guide from mW/cm² to J/cm²
• Joules vs Irradiance: Why mW/cm² Is Just the Starting Point, Not Your Final Red Light Therapy Dose
• What Are Joules in Red Light Therapy? Why Dose Matters More Than Watts and LED Count
• Is More Red Light Therapy Better? Understanding the PBM Dose Window

References

• Light Therapy Insiders / Alex Fergus. The $2,500 Gadget I Use to Test Red Light Therapy Devices. 2025.
  https://www.lighttherapyinsiders.com/
• Hopoocolor. OHSP-350-IR / OHSP-350-IRS / OHSP-350-IRF spectroradiometer product pages and specifications.
  https://www.hopoocolor.com/product/detail/OHSP350IR.html
• Chung H, Dai T, Sharma SK, Huang YY, Carroll JD, Hamblin MR. The Nuts and Bolts of Low-level Laser (Light) Therapy. Annals of Biomedical Engineering, 2012. https://pmc.ncbi.nlm.nih.gov/articles/PMC3288797/
• Huang YY, Chen ACH, Carroll JD, Hamblin MR. Biphasic Dose Response in Low Level Light Therapy. Dose-Response, 2009/2011. https://pubmed.ncbi.nlm.nih.gov/20011653/
• THOR Photomedicine. Calculating PBM Therapy Dosage.
  https://www.thorlaser.com/PBM/calculating-PBM-dosage.php
• ISO 13485:2016. Medical devices — Quality management systems — Requirements for regulatory purposes.
  https://www.bonnier.net.cn/download/d_20170812100731.pdf
• Medical Device Single Audit Program (MDSAP). Companion document and audit model — design and development records expectations.
  https://www.fda.gov/media/166672/download

This article is for educational and engineering reference only and does not constitute medical advice. For specific therapeutic applications, consult published clinical literature and a qualified healthcare professional. Any claims related to therapeutic use, indications, or efficacy should be evaluated by the regulatory authority of the destination market and supported by registration documentation appropriate to that market.

Reposting requires indicating the source.