Independent measurements of consumer LED devices keep turning up the same finding: what’s advertised and what’s delivered don’t always match. One recent independent study of four commercially available LED face masks, as reported by a light therapy review publication, found that measured wavelengths, irradiance and dose didn’t consistently match manufacturer claims — including light sold as one colour measuring closer to another, and hotspots leaving parts of the treatment area under-covered. The practical response for a brand is to stop relying solely on a supplier’s own numbers. This guide covers how to commission testing yourself.
RedVance manufactures red light therapy panels. This guide describes how independent testing works and how to commission it; it is not a statement of independent test results for any RedVance product.
Why a supplier’s report isn’t enough on its own
A supplier’s test report may be entirely accurate. The problem is structural rather than personal: the party producing the number has a commercial interest in what it says, and controls the sample, the conditions and the choice of what to report. That doesn’t make supplier data worthless — it makes it one input rather than independent verification.
How to read a supplier’s report critically is covered in our irradiance report guide and spectrum report guide. Independent testing is the step beyond reading: generating your own evidence.
What to measure
| Measurement | What it tells you |
|---|---|
| Irradiance, multi-point | Delivered output at stated distances across the surface — not one centre reading |
| Spectral output per channel | Actual peak wavelength and bandwidth versus the nominal figure |
| Total optical output | How much light the device emits overall, independent of distance |
| Output over a session | How irradiance changes as the device warms up |
| Uniformity | Hotspots and low zones across the treatment area |
| Flicker (optional) | Whether output modulates rapidly under the driver |
The multi-point and uniformity rows matter more than most brands expect. A single centre reading can look excellent on a device with significant low zones between LEDs — the problem described in our coverage guide. Thermal behaviour is covered in our thermal drift guide.
Where the sample comes from
The independence of the result depends on the independence of the sample:
| Sample source | Strength |
|---|---|
| Bought anonymously on the open market | Strongest — reflects what customers actually receive |
| Pulled at random from a production batch | Strong — especially if you or your inspector pull it |
| Golden sample approved before production | Useful as a reference, but may not represent production |
| Selected and shipped by the supplier | Weakest — the supplier chose it |
For a product you’re about to manufacture, the practical route is to have your inspector pull units at random during pre-shipment inspection — covered in our pre-shipment inspection guide — and compare them against the approved sample from our sample evaluation guide. Testing more than one unit also shows unit-to-unit variation, which a single test can’t.
Choosing a laboratory
- Appropriate calibrated equipment. Typically a spectroradiometer for spectral and irradiance measurement, and an integrating sphere for total optical output. Ask when each instrument was last calibrated.
- Relevant accreditation. ISO/IEC 17025 is the general standard for the competence of testing laboratories. Check that the accreditation scope actually covers the test methods you need, not just the lab in general.
- No commercial relationship with the supplier. A lab recommended by the supplier isn’t automatically compromised, but ask the question.
- Method disclosure. The report should state instrument, distance, measurement points, warm-up time and conditions. A lab that won’t state its method produces a number, not evidence.
Agree the protocol before testing
Results depend entirely on conditions, so define them before anyone sees data:
- Distances — the real operating distance of the format, plus at least one further point to show fall-off.
- Measurement grid — number and position of points across the treatment surface.
- Warm-up — time powered on before measurement, matching a realistic session.
- Channels — all channels on together, as the user runs it, plus individual channels if needed.
- Ambient temperature and power supply — recorded, and matched to the target market’s voltage.
- Units tested — how many, and from where.
Reading the results against the claims
- Compare like with like. Measured irradiance at 15 cm against a claimed figure at 15 cm — not against a figure measured at the surface.
- Check each channel. Multi-wavelength devices can meet a combined figure while one channel falls short — see our multi-wavelength guide.
- Look at the spread, not just the average. A good average with large low zones is a uniformity problem.
- Check LED counting. If output seems low for the stated LED count, the counting convention may be the explanation — see our LED count guide.
- Note unit variation if more than one unit was tested.
A shortfall isn’t automatically a scandal — it’s information. It either corrects your spec sheet before a customer does, or opens a conversation with your supplier before the next batch.
Publishing results
Independent data is increasingly published openly by brands as a trust signal. If you publish:
- State the method, lab, date and model or batch tested alongside the numbers.
- Publish the unflattering figures too. Selective publication of independent data undoes the point of commissioning it.
- Keep it to physical performance. Test data supports statements about irradiance, spectrum and uniformity. It doesn’t support outcome claims — the boundary in our guide to reading research without overclaiming.
- Retest on change. A new component, driver or production batch can change the result.
Test data is one part of the full evidence documentation package, and it feeds directly into a spec sheet that survives verification.
The independent testing checklist
- Measurements defined: multi-point irradiance, spectrum per channel, total output, session behaviour, uniformity.
- Sample sourced as independently as practical — market purchase or random production pull.
- More than one unit tested where budget allows.
- Lab checked for calibrated equipment, relevant ISO/IEC 17025 scope and independence.
- Protocol agreed in writing before testing.
- Results compared like-for-like against claims, per channel.
- Published results include method, lab, date and batch, without outcome claims.
- Retesting scheduled for component or batch changes.
This is one of four guides in our product definition series. Terms are defined in our glossary.
What a manufacturer can usefully contribute: production units for random selection, full access for a buyer’s inspector to pull samples, the technical detail a lab needs to set up a realistic protocol, and no objection to results being published. A supplier who resists independent testing of its own product is telling a buyer something worth knowing. Our panel range and OEM and ODM pages describe what we build.
Frequently asked questions
Why commission independent testing of a red light device?
Because advertised specifications and delivered performance do not always match. Independent measurements of consumer LED devices have reported differences between claimed and measured wavelength, irradiance and dose, along with uneven coverage. Testing by a party with no commercial stake in the result gives a brand documentation that does not depend on the supplier’s own word.
What should an independent red light test measure?
At minimum: irradiance at stated distances across multiple points on the treatment surface, spectral output including peak wavelengths and bandwidth for each channel, total optical output, and how output changes over a full session as the device warms. Electrical safety and electromagnetic compatibility are separate certifications, not part of performance testing.
Where should the test sample come from?
The more independent the sample, the stronger the result. A unit purchased anonymously on the open market reflects what customers actually receive. A unit pulled at random from a production batch is the next best option. A sample selected and shipped by the supplier is the weakest, because it may not represent production.
How do I choose a laboratory for red light device testing?
Look for appropriate calibrated equipment — typically a spectroradiometer, and an integrating sphere for total output — accreditation to ISO/IEC 17025 covering the relevant test methods, no commercial relationship with the supplier, and willingness to disclose the measurement method in the report. A lab that will not state its method produces a number, not evidence.
Why does the test protocol need to be agreed before testing?
Because the result depends on the conditions: distance, warm-up time, number and position of measurement points, ambient temperature, which channels are active and the power supply used. Defining these before testing prevents choosing favourable conditions after seeing the data, and makes results comparable over time and across products.
Can independent test results be used in marketing?
They can support statements about what the device physically delivers — irradiance at a distance, spectral output, uniformity — when published with the method, date and model or batch tested. They do not support outcome or health claims, which require different evidence and are governed by advertising and device regulation in each market.
Want to test our panels independently?
We’ll supply production units for random selection, give your inspector access to pull samples, and provide the technical detail your lab needs to set up the protocol. Publish whatever the results say.
Arrange independent testing →
EN