Light Is Good for You. Don't Overthink It.

Everyone wants a protocol to follow. If a modality exists, we assume someone must have found the best way to use it. And to be fair, this is true for many therapies and practices.

 

Many people who sell red light therapy devices like to tell you exactly how much you need. Six minutes for skin. Twelve for joints. Twenty minutes at twelve inches. Hit your target of 30 J/cm². Don't go over. Don't go under. Another line of reasoning is "match sunlight, don't go too much over as there may be negative consequences from over-saturating the body."

 

Having this kind of specific advice to follow feels good… but thousands of published studies all together tell us almost none of these recommendations map to reality.

 

It’s the same story with what specific frequencies ‘work the best’ and what "should be avoided."

 

In nature, almost everything tends to work in bell curves. This is exactly the case with light therapy. There are likely no magic frequencies, and there is no specific level of exposure that’s best for any given outcome.

 

Anyone telling you otherwise is selling confidence, not describing reality.

 

What the research on light therapy shows in aggregate…

Vladimir Heiskanen is a Finnish researcher who spent years (before AI) building by hand the most complete database of photobiomodulation research in the world, more than nine thousand studies covering every major outcome, every studied wavelength, every dose, and session time that's been tested on living tissue. It’s a great resource, you can find it here.

 

We pulled out every study in the database with clear power and time parameters available.

 

These span nearly 5 orders of magnitude over both time and irradiance!  That’s a one hundred thousand times (100,000x) difference between the lowest exposure and the highest studied. All with positive results.

 

Irradiance (how intense the light is) ranged from 0.04 to over 6,000 mW/cm² while session time ranged from 1 second to multiple hours. Total exposure (fluence, in J/cm²) ranged 1,000,000x (yes, one MILLION X!) from 0.03 to well over 10,000.

 

This isn't just noise across studies of different areas of the body. This is within and across different categories all together.

 

Muscle recovery studies with positive outcomes span doses over a 10,000x range.

 

Wound healing studies span 20,000x

 

Skin rejuvenation, over 170,000x

 

Even the tightest category we grouped “joint comfort" spans 300x

 

If there were a single "correct" dose for any of these outcomes, we'd surely have found it by now. Instead, the literature shows an extraordinarily wide effective range…a cloud where positive results show up almost everywhere on the grid.

(Click for interactive chart!)

Marker shapes: ● circles = LED panels and large-spot (≥1 cm² beam) studies, comparable to consumer red-light devices · ▲ triangles = small-spot laser studies (<1 cm² beam or "laser" in title), typically dental/acupuncture applications delivering high intensity to tiny tissue areas.
Source: Vladimir Heiskanen's PBM research database · 1,999 positive-outcome studies (colored) and 46 negative/null-outcome studies (black ×) · Irradiance: computed from power ÷ beam area where available, else reported value · Time: computed from fluence ÷ irradiance where available, else reported value · Rows flagged uncertain or where reported vs computed time disagree by >100× have been dropped · Values above 10⁵ mW/cm² omitted.
*Extended exposure with Ironforge and Ironforge Mini requires allowing tissue to cool between passes. If the device feels uncomfortably warm, move on and return after cooling.

Why is the range of effective exposure so wide?

Our cells have been absorbing light to help them work better since before our simple ancestors had skin. These cells have been doing the same job for about a billion years.

 

When a photon in the right range hits, nitric oxide releases, the electron transport chain picks up speed, and your cell makes more energy. One photon, one event.

 

This process doesn't seem to care whether the photons arrived over two seconds, twenty, or over several minutes. Same total, same effect. This principle of equivalent result from equivalent total photon count regardless of how fast you delivered them, is called reciprocity, and it's been tested in PBM research enough times to take it seriously.

 

To put numbers to it... 850 mW/cm² for 20 seconds in one area is more or less equivalent to 85 mW/cm² for 3 min 20 seconds. Both deliver the same total exposure to your tissue. The photons don't come with timestamps.

 

This is why the "correct dose" question is a category error. You're not aiming at a target. You're filling a bucket, and the bucket is very generous.

We get a TON of light from a day outside in the shade…

For context on these numbers, a walk in the sun and shade delivers 8 to 15 mW/cm² of red and near-infrared light to your skin. Spend six hours outside in the sun and shade and you've picked up 170 to 320 J/cm² … this is much more than many red light protocols suggest as an entire session.

 

Nobody has ever been "overdosed" with red light by spending a day outdoors. (I'm not promoting burning, seek shade as needed and don't burn!) The structures that absorb red and NIR wavelengths grew up under exactly this kind of long daily exposure. The body has had roughly a billion years to learn how to handle red and NIR photons.

 

The real concern is thermal. If an exposure feels uncomfortably warm, you're likely past useful. Move on, or if you're outside, seek shade. Let your tissue cool. That's the practical safety rule. Check out this article we wrote here on the safety of high irradiance light therapy.

Why 850nm + 810nm NIR and 630 + 660nm Red dominate the literature

If you look at which wavelengths appear most in the research, three stand out as really well studied: 630 + 660nm red, and 810nm + 850nm NIR. It’s easy to draw the conclusion that this must be because these are the most effective ranges for our biology, but that’s not the case.

Research can only be done on devices that exist. Those wavelengths became dominant because they were the cheapest and most available LEDs when most of this research began. 660nm and 630nm because it's the standard ‘red’ in consumer and industrial LED production, 810nm and 850nm because they are the standard wavelengths for nighttime security cameras. Researchers use the hardware they can afford. When a 630nm LED costs a fraction of a 720nm LED, you get a disproportionate number of 660nm studies.

Positive results in any range also lead to more studies of that range, reinforcing the trend that only appeared because of the economics of LED production in the first place!
(Click For Interactive Chart!)

Each dot (●) is one study with a positive outcome. Each ✕ is a negative or null outcome. Dot colour approximates the actual wavelength. X-axis is wavelength (nm). Every research area shows positive results spread across hundreds of nanometers. Negative results are similarly scattered. There is no magic wavelength.

This chart shows what happens when you look at the full distribution of light therapy studies. Positive outcomes extend from below 400nm to beyond 1000nm across every category that's been studied. The 82 studies flagged as negative or null are scattered just as broadly… no wavelength consistently fails, any more than a specific wavelength consistently succeeds.

The effective window for red light therapy isn't three wavelengths wide. It's most all of the red-to-NIR spectrum. And remember, just because a specific frequency range isn't studied, doesn't mean it is not bioactive!

Further… a 660nm LED does not only emit 660nm, there’s meaningful output from 645-675 and beyond. This holds true for nearly every single LED spectra. Again, we’re often dealing with bell curves here, not single frequency lasers (lasers are closer to true single frequency sources).

Consistency matters for results…

Many randomized trials that showed meaningful results have a common thread: participants used the protocol consistently, over weeks or months.

 

Consistency is the key. Not a specific exposure level.  Our bodies expect to be outside for most of the day, over most of the year. We live really far away from that norm in the modern world, and as a consequence, we’re all starved for light.  Light therapy isn’t like a massage, where you’ll always feel the impact immediately, it’s reintroducing the light your body has been missing over most of our adult lives indoors.

 

In practical terms: use a device powerful enough to deliver a meaningful exposure in a few minutes. Use it regularly. Listen to your body. If it feels overly warm, allow the area to cool before more exposure. Don't stress over the stopwatch.

 

More red light isn't always better. But the range of what works is wide enough that you don't need to hit a specific target. Start small, work up slowly over time until you are feeling an impact, and don't exceed that by much. Biology varies person to person and you may need significantly more or less exposure than somebody else for a similar effect.

Where this leaves Chroma devices

The Chroma Ironforge delivers ~850mW/cm² at the faceplate across five wavelengths, 630, 670, 760, 810, and 850nm, with about 75% of output in the near-infrared range for a focus on deep tissue work, as longer wavelengths penetrate deeper into tissues on average. The Ironforge Mini delivers about 650mW/cm² at the faceplate, in a smaller, travel friendly form using the same five wavelengths. Our Chromatorch is battery powered at ~85mW/cm² for longer sessions when a wall outlet isn't handy. Each can be used at a distance to reduce the irradiance and increase the exposure area if desired.

 

Yet, even at the faceplate, all three sit comfortably inside the well-studied zone. They land where positive outcomes have been documented across essentially every major PBM research category from skin, muscle, joints, wound healing, brain, sleep, and hair.

 

What these devices share is enough power to matter in a few minutes. That's the feature that separates a device that becomes part of your routine from one that sits in a drawer.

 

Listen to your body, spend more time outside, and when you can’t get outside, supplement your light exposure.

Nobody really knows how photobiomodulation works (and that’s okay!)

If you’ve read anything about red light therapy, you’ve probably seen the explanation involving cytochrome c oxidase (CCO) in your mitochondria, which releases nitric oxide and speeds up ATP production. It’s a clean story. It shows up on every product page in the industry, including ours.

The nuance is, it might not be right. Or at least, it’s probably not the whole picture.

A 2020 review from the Medical College of Wisconsin put it bluntly: "No reliable demonstration of any PBM-related light-induced mechanistic effect on CCO has been reported." Their alternative: the real action might be light releasing nitric oxide from hemoglobin and myoglobin, not from CCO directly.

A 2019 study used cells genetically engineered to have no functional CCO at all. Red light at 660nm still increased proliferation and ATP. The enzyme that’s supposed to be doing all the work… wasn’t active. And it still worked.

Other researchers have proposed that NIR light is absorbed by water layers on mitochondrial membranes, changing their viscosity and letting the ATP synthase motor spin more freely. Another group showed that different bands of NIR light work through completely different mechanisms with fundamentally different biology.

There are more ideas: opsins in non-eye tissues, reactive oxygen species as signaling molecules, direct effects on cell membranes. The field is genuinely unsettled.

We’re not pointing this out to undermine the therapy. Nine thousand studies show us it works. What we’re saying is anyone who tells you they know exactly why it works, and that you need exactly some set of wavelengths at some dose they recommend to get the benefit, is selling certainty the science hasn’t demonstrated yet.

The honest engineering response to this uncertainty is to cover more ground, not less. The Ironforge uses five wavelengths across the red and NIR spectrum. If CCO is the main target, we’re covering its documented absorption peaks. If it’s NO release from blood proteins, we’re in the right range. If it’s water absorption or ion channels, broader coverage helps there too. We designed for the proposed biology and left room for the unknown.

At the end of the day, engineering is about what works. We have found, over nearly a dozen iterations of our devices, a combination of wavelengths and power that is consistently reported as being very helpful to our customers. Science may catch up eventually and tell us why.

Frequently Asked Questions

How long should I use red light therapy per session?

There is no single correct session length. Across more than nine thousand published photobiomodulation studies, positive outcomes have been documented with session times ranging from under one second to over an hour. What seems to matter is total energy delivered to the tissue, which is a function of both irradiance and time, this also has a very wide range of positive outcomes. A high-irradiance device like the Chroma Ironforge (850 mW/cm² at the faceplate) delivers a meaningful dose in 30seconds to three minutes per area. A lower-power device may need fifteen to forty minutes for a comparable dose. Start with shorter sessions, increase gradually, and stay consistent.

What is the best wavelength for red light therapy?

No single wavelength is "best." The research literature shows positive outcomes across the full range from visible red (~620nm) through deep near-infrared (~1100nm), across every category that has been studied. The two most commonly studied wavelengths, 660nm and 810nm, appear frequently not because they are biologically superior but because these were the cheapest and most available LEDs when most of this research was conducted. A device that covers multiple wavelengths across the red and NIR spectrum provides broader biological coverage than one that targets a single peak.

Can you overdo red light therapy?

Photochemical overdose in the body from red and near-infrared light has not been demonstrated at irradiance levels used by consumer devices. A six-hour walk in the sun and shade delivers 170 to 320 J/cm² of red and NIR light to your skin, far exceeding typical device protocols, and nobody has ever been harmed by a day outdoors from these wavelengths. The practical limit is thermal: if the device feels uncomfortably warm against your skin, move it to a different area, or increase your distance. Allow tissue to cool before continuing. More exposure is not always better, but the effective range is wide enough that precise dosing is unnecessary.

Does red light therapy actually work?

More than nine thousand peer-reviewed studies have investigated photobiomodulation across conditions including muscle recovery, wound healing, skin rejuvenation, joint comfort, neurological function, and pain. The overwhelming majority report positive outcomes across a wide range of wavelengths, irradiances, and session times. The mechanism is well understood. Life has been responding to these wavelengths for roughly a billion years.

Is red light therapy a scam?

The science behind photobiomodulation is robust and well-documented across thousands of studies. What is sometimes misleading is how companies market their devices, particularly around precise dosing protocols and inflated specifications. Claims like "you must use exactly 30 J/cm² at exactly 660nm" are not supported by the breadth of the research, which shows positive results across enormous ranges of dose and wavelength. The therapy itself is legitimate. The marketing around specific protocols and magic numbers is where skepticism is warranted.

How often should I use red light therapy?

Consistency matters more than any single session parameter. Studies showing meaningful results typically involve regular use over weeks or months. Daily or near-daily sessions of a few minutes generally produce better outcomes than occasional long sessions. The goal is to reintroduce light exposure that modern indoor living has eliminated, not to achieve a precise one-time dose. Think of it as a daily habit like drinking water, rather than an occasional treatment like a massage.

*This article is for educational purposes and reflects current peer-reviewed research. It is not medical advice. Chroma devices are general wellness products designed to support your body's natural processes. If you have a medical condition, consult your healthcare provider.*

References

1. Heiskanen V. Photobiomodulation (PBM) research — a comprehensive database. Continuously updated. Over 9,000 entries covering PBM studies across all major research areas, wavelengths, and dosing parameters. Data source for both charts in this article. Available at: bit.ly/PBM-database
   
   

2. Karu TI. Action spectra: their importance for low level light therapy. Summary of action spectrum research measuring DNA synthesis rate and cell adhesion in HeLa cell monolayers using monochromator-generated narrow-band light (10-14nm FWHM). Identified four CCO absorption peaks at ~620nm, ~680nm, ~760nm, and ~830nm corresponding to different redox states of the enzyme's copper centers. Available at: photobiology.info/Karu.html
   
   

3. Wong-Riley MTT, Liang HL, Eells JT, et al. Photobiomodulation directly benefits primary neurons functionally inactivated by toxins: role of cytochrome c oxidase. J Biol Chem. 2005;280(6):4761-4771. Independent functional confirmation: 670nm and 830nm (near Karu's peaks) restored CCO activity in poisoned rat neurons; 728nm (between peaks) was least effective.
   
   

4. Mason MG, Nicholls P, Cooper CE. Re-evaluation of the near infrared spectra of mitochondrial cytochrome c oxidase. Biochim Biophys Acta. 2014;1837(11):1882-1891. Spectroscopic confirmation of CCO absorption bands using purified beef heart enzyme: CuA maximum at 835nm, broad 715-920nm band from perturbed binuclear center spanning the 760nm region.
   
   

5. Bunsen R, Roscoe H. Photochemical researches — Part V. On the measurement of the chemical action of direct and diffuse sunlight. Proc R Soc Lond. 1862;12:306-312. Original demonstration of the reciprocity principle (Bunsen-Roscoe law): biological effect depends on total dose (irradiance × time), not delivery rate alone.