Red light for recovery vs red light for sleep: what's the difference?

Type "red light" into a search bar and two very different shops appear side by side, apparently selling the same thing at wildly different prices.

One is offering a floor-standing panel the size of a door, photographed next to a very fit person in shorts, for anywhere from a few hundred to a couple of thousand pounds. The other is offering a small bedside lamp for roughly the price of a takeaway.

If you are the parent who just wants to see what you are doing at 3am without waking all the way up, this is genuinely confusing. Is the cheap lamp a knock-off of the expensive panel? Do you have to spend £400 to get anything useful at night? No, and no. And the reason is more interesting than "one is dear and one is cheap".

These are two entirely different jobs that happen to share a colour. They differ in mechanism, dose, intensity, duration, timing and, most importantly, in the direction of the effect you want. One is about giving your body something. The other is about giving it as close to nothing as you can manage while still being able to see.

Job one: a dose you deliberately sit in front of

The expensive panels belong to a field usually called photobiomodulation, or, in the older literature, low-level laser therapy. The proposed mechanism is specific and physical: photons of red and near-infrared light are absorbed by cytochrome c oxidase, an enzyme complex in the mitochondrial respiratory chain, which is thought to alter cellular energy production and downstream signalling.^[1] Two things follow from that, and both matter here.

First, it is a dose. Photons have to arrive at the tissue in sufficient number to do anything at all. Reviews of the field describe device powers spanning roughly 1 to 1,000 mW and an effective "optical window" of about 600 to 1,070 nm, with shorter red wavelengths used for superficial tissue and near-infrared for deeper structures.^[1] A meta-analysis of phototherapy and exercise found the most consistent results with power outputs between 50 and 200 mW and doses of 5 to 6 joules per point.^[2] Joules per point. That is the language of a treatment, not of a lamp.

Second, and almost nobody mentions this in the marketing, more is not automatically better even here. Photobiomodulation is characterised by what researchers call a biphasic dose response, sometimes described using the old Arndt-Schulz curve: too little light does nothing, an intermediate dose produces the effect, and too much can suppress it again.^[1][3] One laboratory study of human stem cells in a dish found the optimal dose for 810 nm light was roughly 10 to 100 times higher than for 980 nm light, and that overshooting either lost the effect.^[4] That is in-vitro work on cells, not people, so treat it as mechanism rather than proof. But it tells you something: in this field, dose is a target you can miss in both directions.

Job two: an absence you happen to be standing in

The night job runs on an entirely different system. Light reaching the eye at night does not just let you see. It also reaches a small population of retinal cells containing the photopigment melanopsin, which feed directly into the circadian clock and the pathway controlling melatonin. The classic action spectrum work in humans put peak sensitivity of that system at roughly 446 to 477 nm, squarely in the blue and clearly distinct from the pigments we use for ordinary vision.^[5]

The system is also alarmingly sensitive. In a controlled study of 23 healthy volunteers, roughly half the maximum melatonin suppression and half the maximum phase shift produced by very bright evening light (around 9,000 lux) could be achieved with just over 1% of that light, about 100 lux, which is ordinary dim room lighting.^[6] You do not need a floodlight to send the signal. You need a landing light.

So the target for night-time lighting is not a dose. It is the smallest possible signal. An international expert consensus published in 2022 put numbers on it, recommending a melanopic equivalent daylight illuminance below 1 lux in the sleep environment and below 10 lux in the three hours before bed, while pushing daytime exposure above 250 lux.^[7]

This is where long wavelengths earn their place. Because the melanopsin system is so much less sensitive at the red end, a light can be visible to you while contributing very little to that melanopic figure. In one study comparing wavelengths at matched irradiance (65 µW/cm² per diode), 470, 497 and 525 nm light advanced the melatonin rhythm by roughly 40 to 65 minutes, while 595 nm amber and 660 nm red produced no significant shift at all.^[8] Separately, researchers have shown you can hold illuminance and colour temperature constant and still roughly halve or eliminate melatonin suppression purely by changing which wavelengths are in the mix.^[9] Spectrum is a lever independent of brightness.

The two jobs side by side

The counterintuitive bit: for sleep, a bigger panel is worse

A powerful red therapy panel is not a better version of a small dim red lamp. For the night-time job it is arguably the wrong tool, because everything that makes it good at job one makes it bad at job two.

Intensity is the whole point of a therapy panel. Intensity is the enemy of the night job. Melanopsin is far less sensitive to red than to blue, but "less sensitive" is not "blind", and the melanopic figure those consensus recommendations are built around scales with how much light actually reaches the eye.^[7] Put a very bright red source in front of your face at 3am and you have taken a spectrum designed to be quiet and turned the volume up on it.

Which answers the other common question: could using a recovery panel in the evening interfere with sleep? It depends on intensity, on how close to bedtime you use it, and on how much of the output your eyes receive rather than your quadriceps. The literature gives no clean yes or no, but it does give a principle: the risk scales with the light your eye gets and how close it is to your bedtime.^[6][7] Using a panel earlier in the day, and not staring into it, are sensible low-cost precautions.

How good is the recovery evidence, really?

If you are deciding where to put a few hundred pounds you deserve a straight answer rather than a brochure. The short version: there is a real signal, and the literature is genuinely mixed in quality.

• A meta-analysis of 24 randomised and crossover studies in athletes concluded the evidence supports improved muscular performance, while explicitly stating that a definite therapeutic effect is yet to be established and calling for larger, more robust trials.^[10]
• A systematic review of 16 randomised trials in healthy people, totalling just 340 participants, found improvement in fatigue recovery but no improvement in strength and none in functional capacity. Device parameters varied enormously, spanning 655 to 905 nm.^[11]
• A review comparing photobiomodulation with cryotherapy found in favour of light, but rested on four studies and 66 participants in total, with certainty of evidence rated moderate to very low.^[12]
• In dermatology, reviews describe red LED light in the 630 to 760 nm range being used for skin rejuvenation, scar healing and hair growth, with mechanisms attributed to mitochondrial and immune modulation.^[13]

Small samples, heterogeneous protocols, plausible mechanism, unsettled results. Not nothing, but not settled either. If a panel appeals to you for muscle soreness or skin, that is a defensible purchase made with open eyes. It is simply not a purchase about your night feeds.

One small study muddies the neat divide, and it would be dishonest to skip it. Ten of twenty elite basketball players received 30 minutes of whole-body red light nightly for 14 days, and that treated group showed better sleep questionnaire scores and higher serum melatonin than a group given no light at all.^[14] Twenty participants, no sham light for the controls, and a design that cannot separate the light from the ritual of lying still every evening. It is a hypothesis, not a finding, and not strong enough to carry the claim that therapy panels improve sleep.

So which one do you actually need?

A therapy panel, if...

• You train hard and want to try something for muscle soreness or perceived recovery, understanding the evidence is preliminary and mixed.^[10][11][12]
• You are interested in the skin and hair applications described in the dermatology literature.^[13]
• You are happy to sit in front of a device for several minutes as a deliberate daytime routine.

A red night light, if...

• You are up at 2am with a baby and the overhead landing light currently costs you an hour of lying awake afterwards.
• You want to find the loo, the stairs or a muslin without switching on anything that broadcasts to your body clock.
• You are a shift worker or a light-sensitive sleeper protecting the hours you do get. Worth knowing the shift-work lighting literature is itself unresolved: a meta-analysis of bright light in nurses found benefits under a fixed-effects model that disappeared under random effects, and the authors concluded the controversy persists.^[15]

If you want both, buy both, for different reasons. Just do not buy the panel expecting it to solve the 3am problem, and do not skip the lamp because you assume it is the budget option. On the night job it is not the budget option. It is the correct one.

What this means for what we make

Driftback does not make therapy panels, and we will not pretend the two categories compete. We make lighting for the second job, a much narrower brief: be visible enough to function, be as quiet as possible to the melanopsin system, and cost what a lamp should cost.

That is why our bedside lamp uses a true 670nm red and a warm 1700K amber rather than a coloured cover over a white LED (a tinted shade still leaks the short wavelengths you are trying to avoid), why the brightness goes genuinely low rather than merely lower, and why the plug-in hallway light switches itself off after 60 seconds instead of glowing all night. Not because dimmer is a cheaper way to build a light. Because for this job, dimmer is the specification.

You are not trying to treat anything at 3am. You are trying to see the baby and then get back to sleep. Those are the only two requirements, and they do not cost £400.