How red light therapy actually works.
No jargon, no hype. Just the mechanism, the wavelengths, and the research — explained like you're talking to a friend who happens to have the studies open.
Light in. Energy out.
Your skin cells contain mitochondria — the structures that produce energy (ATP) for everything your body does. Certain wavelengths of light are absorbed by a protein inside those mitochondria called cytochrome c oxidase.
When that happens, the cell produces more energy. More energy means more of the things skin needs to renew itself: collagen synthesis, cell turnover, repair.
This process is called photobiomodulation. It's not heat. It's not UV. It's light at specific wavelengths triggering a cellular response — and it's been studied for over 40 years.1,2
The right wavelength, at the right strength, for long enough — that's the whole science.The three variables that determine whether a session does anything: wavelength (which cells absorb the light), irradiance (how much light reaches the skin), and time (how long it's applied). Get all three right, and photobiomodulation works. Miss one, and nothing happens — no matter how good the other two are.
What each wavelength does — and where the evidence stands.
Not all wavelengths are equal. Here's what the research says about each one, honestly — including where the evidence is strong and where it's thinner.
Red · 630nm
Absorbed in the upper dermis where fibroblasts live. Research links 630nm red light to increased collagen density, reduced fine lines, smoother texture, and improved complexion — consistently, across multiple controlled trials.1,2,3
Evidence: strong. Multiple RCTs with measurable outcomes.
Near-Infrared · 850nm
NIR passes through the epidermis and reaches deeper into the dermis. Research on 830–850nm shows increased collagen and elastic fiber production, wrinkle reduction, and improved skin elasticity when combined with red light.1,5
Evidence: solid. Best studied in combination with red, which is how Lumé delivers it.
Blue · 460nm
Blue light in the ~415–460nm range is absorbed by porphyrins produced by C. acnes bacteria, generating reactive oxygen species that destroy them. A landmark RCT of 107 patients found blue light reduced inflammatory acne lesions by 63%.4
Evidence: solid for acne. Well-established antibacterial mechanism.
Yellow · 590nm
Yellow/amber light at ~590nm is associated with calming surface redness and supporting a more even tone. Clinical experience reports (including a 900-patient open-label series) suggest benefits for photoaged skin and post-treatment redness.
Evidence: emerging. Mostly open-label and self-reported data. No large RCTs yet — we're honest about that.
We include yellow modes but feature them less prominently than red and blue, reflecting the current evidence.
Why irradiance decides everything.
Wavelength tells you what kind of light. Irradiance tells you how much of it reaches your skin — measured in mW/cm².
This is the number that determines whether a session delivers enough energy to trigger photobiomodulation. A mask with the perfect wavelength but weak irradiance will do nothing — you can't fix it by wearing it longer. Below a certain threshold, the light is simply too faint to produce a cellular response.1
Dose = irradiance × time. Most published studies use a dose of roughly 15–30 J/cm², delivered over 10–20 minutes. That requires an irradiance in the range of roughly 25–50 mW/cm² at the skin's surface.
You can't fix a weak mask by wearing it longer.One more thing: the measuring distance matters. Light fades fast with distance, so the same mask can read strong at the LED surface and far weaker at the distance you actually wear it. A published irradiance without a stated distance isn't telling you what your skin actually receives.
Every mode, measured on a spectroradiometer.
We tested the Lumé mask on a HAAS-3000 spectroradiometer — the same instrument used in optical testing labs. Every mode, measured at 2.5cm (the distance the mask sits from your face when worn). Here are all six readings.
Tap to enlarge
Red · 630nm · 37 mW/cm²
Tap to enlarge
Red + NIR · 630+850nm · 41 mW/cm²
Tap to enlarge
Blue · 460nm · 32 mW/cm²
Tap to enlarge
Blue + NIR · 460+850nm · 37 mW/cm²
Tap to enlarge
Yellow · 590nm · 18 mW/cm²
Tap to enlarge
Yellow + NIR · 590+850nm · 23 mW/cm²
All six pages of the spectrometer report, measured at 2.5cm, under controlled lab conditions (23°C, 65% humidity). The spectrograms show the peak wavelength, and the Ee (mW/cm²) reading is the irradiance — the number that matters.
We publish these because we think you should be able to see exactly what you're getting — not a marketing number, but the actual lab reading.
What the studies actually found.
Here are the key peer-reviewed studies behind the claims. We cite these because they're the evidence base for the wavelengths our mask uses — not because they were conducted on the Lumé device (they weren't). The research is about the light, not us.
Collagen and wrinkles (red / red+NIR): Wunsch & Matuschka (2014) enrolled 136 volunteers in a prospective, randomized, controlled trial. Subjects treated with red and near-infrared light showed statistically significant improvements in skin complexion, skin roughness, and collagen density vs. controls. Blinded photo evaluation confirmed the results.1
Collagen metabolism (red): Barolet et al. (2009) used pulsed 660nm LED light and found a 31% increase in type-1 procollagen and an 18% reduction in MMP-1 (the enzyme that breaks collagen down) in tissue-engineered skin. In the clinical arm, over 90% of subjects showed reduced wrinkle depth.2
LED mask for skin aging (red): Couturaud et al. (2023) tested a 630nm LED mask (12 min, twice weekly, 3 months) on 20 women. Results were progressive — improvements in wrinkle depth, firmness, elasticity, and complexion, with effects lasting up to one month after stopping use.3
Blue light for acne: Papageorgiou et al. (2000) randomized 107 patients across four groups (blue light, mixed blue+red, white light, benzoyl peroxide) for 12 weeks. Blue light alone reduced inflammatory acne by 63%. Combined blue+red achieved 76% improvement — outperforming benzoyl peroxide.4
NIR + Red for skin rejuvenation: Lee et al. (2007) conducted a prospective, placebo-controlled, double-blind, split-face trial. Patients receiving 830nm (NIR), 633nm (red), or both showed statistically significant wrinkle reduction (26%, 33%, and 36% respectively). Tissue analysis revealed increased collagen and elastic fibers.5
Yellow light (590nm): Weiss et al. reported clinical experience with 590nm yellow LED across 900 patients with photoaged skin, noting self-reported softening and reduction in fine lines. While promising, the evidence base for yellow remains mostly open-label and self-reported — no large RCTs exist yet.
See how Lumé delivers this.
630nm + 850nm. 41 mW/cm², measured at 2.5cm. Every mode published.
See the mask → — $239.00