How to Tell If Your Red Light Therapy Device Is Actually Flicker-Free (No Lab Equipment Needed)

I was recently filming inside an infrared sauna, and I pointed my camera at the red light strip mounted near the ceiling. On screen, it was pulsing - visible dark bands rolling through the frame. The far-infrared emitters around the walls were doing their job perfectly well. But the "red light therapy" being sold as part of that sauna was a small strip of LEDs of undeclared wavelength, at an undeclared irradiance, and it flickered.

Then I filmed a Hero960 next to it from the same distance. Near-infrared on: nothing. Red on: nothing. Just a flat wall of light - as it should be.

That's the whole test. It takes just ten seconds and your phone.

What "flicker" actually is

Mains electricity is alternating current. LEDs need direct current. Every LED device therefore has a driver that converts one to the other - and how well it does that job varies enormously.

A well-regulated driver smooths the output into something close to a flat line while a cheap one leaves ripple, so light output rises and falls in time with the mains, typically at 100 Hz in Australia, the UK and most of Europe, or 120 Hz in North America and Japan. The engineering term is temporal light modulation; the industry benchmark for it is IEEE Standard 1789-2015, which defines how to measure modulation depth (percent flicker) and sets recommended limits by frequency (IEEE, 2015).

To summarise, flicker occurs when an LED is unable to maintain a steady stream of light output, kind of like an uncontrolled pulsing.

Just because you can't consciously see it doesn't mean it isn't affecting you

Conscious flicker fusion (where we perceive it as a constant light source) sits somewhere around 60-80 Hz for most people, which is why 100 Hz ripple looks like steady light. The research on what happens above that threshold is more interesting than you may expect.

In a double-blind crossover study, Wilkins and colleagues (1989) tracked headaches and eyestrain among office workers under two lighting conditions: conventional fluorescent ballasts producing roughly 43-49% modulation at 100 Hz, versus high-frequency electronic ballasts running at about 32 kHz with under 7% residual 100 Hz modulation. The average weekly incidence of headaches and eyestrain was more than halved under the high-frequency lighting. Notably, none of the participants attributed their symptoms to the lighting.

Later work showed the visual system responds well above the classical fusion threshold. Roberts and Wilkins (2013) found that during rapid eye movements, flicker becomes discriminable as a "phantom array" at frequencies averaging almost 2 kHz across observers - and at 120 Hz, the effect appeared whenever modulation exceeded 10%. Davis and colleagues (2015) demonstrated perceptible artefacts at 500 Hz. IEEE 1789-2015 summarises the potential consequences of temporal light modulation as including headache, eyestrain, fatigue, blurred vision, migraine, reduced visual task performance and, at low frequencies, seizure risk in photosensitive individuals (IEEE, 2015; Wilkins, Veitch, & Lehman, 2010).

You are going to sit in front of a red light therapy panel for ten to twenty minutes at a time, several times a week, often with the light filling most of your visual field. That's the exposure profile this literature is concerned with.

Another problem: flicker makes dosing difficult

Photobiomodulation dose is energy density - irradiance multiplied by time, expressed in J/cm². If a driver is switching the LEDs fully off for part of every cycle, which is what often occurs when devices flicker, then the light is only actually on for a fraction of your session. For example, at a 50% duty cycle (duty cycle is the percentage of time a system is on for) you lose half the energy; your twenty-minute session becomes a ten-minute session.

That matters more than "half" might sound, because the dose-response relationship in photobiomodulation is not linear. It's biphasic: below the effective threshold you don't get a reduced effect, you get no effect, and past the optimum more light produces less benefit rather than more (Huang, Chen, Carroll, & Hamblin, 2009; Huang, Sharma, Carroll, & Hamblin, 2011). Underdosing doesn't just cost you proportionally, it can prevent benefits altogether.

One caveat: Not all flicker is switching. Some drivers produce ripple that oscillates around a mean without ever reaching zero, and in that case the average irradiance may be roughly preserved - the energy is delivered unevenly rather than lost. Still problematic, but not as bad. Your phone can't tell you which kind of flicker it is, making this test a red flag rather than a verdict. A panel that bands on camera is a panel whose driver isn't tightly regulated, whose modulation depth is undisclosed, and whose published irradiance figure you therefore can't convert into a reliable dose. It also raises a question nobody answers on a spec sheet: was that mW/cm² number a peak reading or a time-averaged one? On a heavily modulated panel those are very different numbers, and only one of them tells you what your skin actually receives. An article on this is coming soon.

The camera test, step by step

1. Dim the room. Turn the device on.
2. Open your phone's camera in video mode. If your phone has an anti-banding or "flicker reduction" setting, turn it off.
3. Stand about a metre back and film the panel for five to ten seconds, panning slowly across it.
4. Repeat in slow-motion mode if you have it - higher frame rates make banding more obvious.
5. Watch the playback. Rolling dark bands, a strobing pulse, or brightness that breathes in and out means the light is flickering.

Why this works: most phone cameras use a rolling shutter, reading the sensor line by line over roughly 15-35 milliseconds rather than capturing the whole frame at once. When the light source is modulating at 100 or 120 Hz, different sensor rows catch different points in the brightness cycle, and the mismatch shows up as banding - a well-documented artefact in imaging engineering (Deegan, 2018).

Two caveats. First, a phone is a screening tool, not a photometer; a clean result means low modulation, not necessarily zero, and it won't quantify modulation depth for you. Second, modern computational photography sometimes suppresses banding automatically, so a positive result (visible flicker) is far more informative than a negative one. If you see bands, you've learned something definitive. If you don't, you've learned that it's probably fine (but maybe your phone is hiding it!).

Back to the sauna

None of this is an argument against infrared saunas. Far-infrared emitters produce longer wavelengths which work through a fundamentally thermal mechanism, these have their own literature and its own benefits (Vatansever & Hamblin, 2012). That's a different modality with a different mechanism from photobiomodulation, and you can't compare the two (in fact many of our customers put HeroPanels Elite inside their infrared saunas as they can work well synergistically!).

The problem is what often gets bolted on: A short strip of red LEDs in the corner of a sauna cabin, with no published wavelength, no published irradiance at a stated distance, and visible flicker on camera, is not a red light therapy device. It's a red light. Enjoy the sauna for the sauna.

Five things worth checking before you buy anything

1. Does it flicker on camera? Ten seconds. Free.
2. Which exact wavelengths, in nanometres, and how many LEDs of each? "Red and infrared" is not a specification. See our Wavelength Research page for what the literature actually covers.
3. What is the irradiance, at what distance, and was it measured with a spectrometer? Numbers quoted at 0 cm are meaningless - nobody uses a panel with their skin touching it - and we'll be sharing an article soon on why solar meter measurements are worthless.
4. Can you actually calculate J/cm² from what's published? If not, you can't compare it to any study. We've created a Dosage Estimator to make this easy for HeroPanels and HeroPanels Elite owners.
5. What is the company's reputation like, do they have real reviews or does it seem like another internet scam? It's also worth checking their certifications and testing.

Have any questions? Feel free to reach out and we'll be happy to help.