The Optical Window: Which Wavelengths Reach Your Cells?
Share
The Optical Window: Which Wavelengths Reach Your Cells?
Someone messaged me last week. Been using their panel for three months, consistent, doing everything right on paper. Knee still the same. Could I take a look at what they'd bought?
I could. Within about thirty seconds I knew the problem. The device was a 630nm and 660nm only panel. Good red wavelengths, well-documented for skin. For a knee joint sitting under centimetres of tissue and muscle? Those photons were doing a solid job on the skin above the knee and essentially nothing else.
Three months. Wrong tool. Not because red light therapy doesn't work - because they didn't have the right wavelengths for the target.
Red light therapy (fotobiomodulasjon) works because specific wavelengths within the 600-1100nm range - the optical window - are absorbed by cytochrome c oxidase in the inner mitochondrial membrane. That absorption triggers ATP production, nitric oxide release, cellular repair. But wavelength determines whether the photon completes the journey. Light absorbed in your epidermis doesn't do anything deeper down. It just becomes warmth.
Why the Body Has an Optical Window
Your body is not transparent. Skin, blood and tissue are full of molecules that absorb light - melanin, haemoglobin, water, cytochromes. Below roughly 600nm, melanin and haemoglobin absorb aggressively. A blue or green photon doesn't get far. Above about 1100nm, water absorption takes over and light converts to heat before it penetrates.
Between 600nm and 1100nm these absorbers have relatively low absorption coefficients. Light can travel. How far depends on exactly where in that window you are.
A 630nm red device and an 850nm near-infrared device are both "red light therapy." They are not treating the same tissue. The industry mostly avoids saying this directly. I won't.
The Three Zones - What Each One Reaches
400nm to 545nm - above the window, epidermis only. UV through green. Absorbed in the epidermis. Cannot reach a joint, a tendon under tissue, or any meaningful muscle belly. Physics, not opinion.
600nm to 700nm - red light, dermal reach. 630nm and 650nm are among the most studied wavelengths in the photobiomodulation literature - solid evidence for skin rejuvenation, collagen, hair follicle activation, surface wound healing. 660nm and 670nm push slightly deeper. These wavelengths reach the dermis and the fibroblasts there.
What they don't reach is anything more than a centimetre or two below the skin surface. Deep muscle. A hip joint. A lumbar disc. If that's your target - and for most people who contact me it is - visible red alone won't do it.
800nm to 1100nm - near-infrared, deep tissue. NIR is absorbed far less by melanin and haemoglobin, which is what lets it travel. 810nm and 850nm are the workhorses - deep-tissue effects, joint pathology, muscle recovery, increasingly the brain. 1064nm is useful for very deep targets: water absorption is relatively low at that wavelength, surface heat is minimal, and it outperforms shorter NIR wavelengths for depth.
A 2017 review by Dr. Michael Hamblin confirmed the clinically relevant penetration difference between visible red and NIR - you can read it here: PMID: 29809255. If your target is deep, NIR is not optional.
Does that land differently now than when you started reading? It should. If you've been running 660nm sessions on a joint that hasn't changed, you've been treating the skin above that joint.
Condition by Condition
Skin rejuvenation. Evidence clusters around 630nm, 650nm, 850nm and 880nm. Visible red handles the dermal layer - collagen, elastin, fibroblast activation. NIR around 850-880nm reaches deeper vascular and connective structures supporting skin from below. Red-only for a skin protocol works but leaves something on the table. More detail at lighttherapy.no/blogs/english/red-light-therapy-for-skin-health-benefits-science-norways-need.
Hair growth. 630nm and 650nm are the core research wavelengths. 830nm shows up in several clinical trials. Most laser cap studies cluster in this range. If someone is selling you a high-power 850nm panel specifically for hair growth, ask for the studies.
Muscle performance and recovery. 630nm through 670nm in visible red, plus 810nm and 850nm in NIR. Ferraresi et al. documented significant exercise performance improvements with pre-exercise red and NIR exposure - PMID: 22540422. Red for superficial muscle, NIR to reach larger muscle groups. Full post at lighttherapy.no/blogs/english/red-light-therapy-for-post-workout-muscle-soreness.
Cognitive performance. 650nm, 810nm and 1064nm. The brain is about 75% water, which is part of why 1064nm is interesting transcranially - water absorption is low at that depth. Worth watching. Most studies are still small. I'd rather be honest about that than oversell it.
Deep joint and arthritis. This is where I spend the most time with people, for the obvious personal reason. 630nm and 650nm for superficial tissue, then 765nm, 810nm, 850nm and 910nm for the deep structures. My arthritis lives in my ankle (old rugby injury, bone started dying - avascular necrosis, if you want the term) and in my hands. January mornings my hands have their own agenda and it doesn't include coffee for a while. Years of using this on myself and talking to hundreds of customers: the NIR end is what reaches the joint. The red is doing something in the layers above. But the photons that matter for deep joint tissue are near-infrared. More at lighttherapy.no/blogs/english/red-light-therapy-joint-pain-arthritis-norway.
Power Density - Two Things That Pull Against Each Other
First: deep tissue needs adequate power density. Low irradiance at your treatment distance means fewer photons start the journey and fewer survive it. The inverse square law applies whether the marketing copy mentions it or not.
Second - and this is the one almost nobody talks about openly - the biphasic dose response. Arndt-Schulz law, well-documented by Hamblin et al. - PMID: 20011653. Low doses stimulate. Moderate doses stimulate well. Excessive doses - generally above 50-100 J/cm² depending on wavelength and tissue - start inhibiting the same processes you're trying to activate. The curve goes back down.
Last autumn I had a conversation with a supplier about a panel that was measuring significantly above its claimed output at close range. Not a fun call. Their customers were potentially overdosing without knowing it - sitting too close for too long, thinking more is better, wondering why they'd stopped improving.
The generic advice you see everywhere - "5 to 20 minutes, three to seven times a week" - tells you almost nothing useful without knowing your device's actual measured irradiance at your treatment distance. Your dose is power density multiplied by time. Change either and you change the dose.
I test every device with a spectrometer before I stock it. I've sent panels back when the measured output didn't match the spec sheet. Full post on what this looks like in practice at lighttherapy.no/blogs/english/is-your-red-light-panel-actually-working-the-irradiance-question.
Start From the Window, Work Backwards
Skin, hair, surface wound healing - visible red 630-670nm. Deep muscle, joints, tendons under tissue - NIR from 810nm upward, with enough power density to actually reach the target. Most conditions benefit from both, which is why the better panels run a dual-wavelength spectrum.
If you're treating joint or deep muscle issues and getting nothing: check the wavelengths first. Red-only device? That's probably the problem. Not your consistency. Not your timing. The tool.
Message me if you want to work through what makes sense for your situation. Panel range with irradiance data at lighttherapy.no/collections/red-light-panels. Portable devices for targeted work at lighttherapy.no/collections/portable-and-specialist.
This post is educational and not medical advice. Red light therapy is not a substitute for professional medical assessment. If you are managing a diagnosed condition, consult your healthcare provider before starting a photobiomodulation protocol.
Frequently Asked Questions
What is the optical window in red light therapy and why does it matter when choosing a device?
The optical window is the wavelength range - roughly 600nm to 1100nm - where the body's main light-absorbing molecules (melanin, haemoglobin, water) have low enough absorption to allow light to travel through tissue rather than converting to heat at the surface. Below this range light stops in the epidermis. Above it water absorption takes over. Within it, red and near-infrared can reach the dermis, muscles, joints, and brain depending on wavelength and power density. A device emitting outside this range won't produce deep-tissue effects regardless of how long you use it.
Does red light therapy actually reach deep joints like knees, hips, or the spine?
Visible red (630-700nm) reaches the dermis but won't reliably penetrate to a joint under significant muscle and connective tissue. For deep joints - knees, hips, shoulders, lumbar spine - you need near-infrared in the 810-910nm range, at adequate power density. NIR is absorbed far less by melanin and haemoglobin than visible red, which is what lets it travel further. Research on arthritis and deep joint pathology consistently uses NIR wavelengths. A red-only device is the wrong tool for deep joint work regardless of session length.
Can you overdose red light therapy?
Yes. The biphasic dose response means excessive doses can inhibit the same cellular processes that moderate doses stimulate. Possible signs: feeling temporarily worse after sessions, plateau or regression after early progress, increased localised inflammation. Research puts the inhibitory threshold generally above 50-100 J/cm², varying by wavelength and tissue. Your actual dose is power density at your treatment distance multiplied by session time - not the diode rating on the box.
Kan jeg bruke rødt lys mot leddsmerter hjemme i Norge?
Ja - og bølgelengdevalget er avgjørende. Synlig rødt lys (630-670nm) er dokumentert for hud og overfladiske strukturer, men rekker ikke ned til leddbrusk og senefester under vev av betydning. For dype ledd trenger du nær-infrarødt i området 810-910nm med tilstrekkelig effekttetthet. Konsistens og riktig dosering er like viktige som bølgelengdevalg. Ta kontakt hvis du vil ha hjelp til en protokoll for din situasjon.
What wavelength is best for cognitive or transcranial use?
810nm has the broadest research support for transcranial applications. 1064nm is increasingly studied because its water absorption coefficient is relatively low at depth - penetration to deeper brain tissue looks better than shorter NIR wavelengths. 650nm appears in some cognitive studies, possibly via retinal photoreceptors. Most of this research is still small and preliminary. Promising, not settled.
References
- Hamblin MR. Mechanisms and applications of the anti-inflammatory effects of photobiomodulation. AIMS Biophysics. 2017. PMID: 29809255
- Ferraresi C et al. Effects of low-level laser therapy on muscle performance and exercise recovery. Photomedicine and Laser Surgery. 2012. PMID: 22540422
- Hamblin MR et al. Arndt-Schulz biphasic dose response in photobiomodulation. PMID: 20011653