Which wavelengths, irradiance levels and total doses actually matter in a device?

Strip away the diode counts and the marketing language and you're left with three numbers that decide whether a light device can do anything at all for your hair. Everything else on a spec sheet is either calculated from those three or it's decoration. If a product won't publish all three, you can't work out what your follicles are actually receiving.

Which wavelengths have the strongest clinical evidence behind them for scalp hair growth?

The evidence base is a lot narrower than the shelf full of devices suggests. Nearly all the controlled scalp research sits in one tight red band, and everything outside it is either a plausible companion or a claim with nothing behind it. Knowing where the line falls saves you from paying extra for emitters that have never been shown to grow a hair.

Why does the red band around 630 to 680 nanometres dominate while near infrared sits in a supporting role?

It comes down to a lucky overlap between what your skin lets through and where the target sits. Red light in that band hits a strong absorption peak while still getting past melanin and blood, which is what produces a chemical signal instead of just warmth. Near infrared goes deeper, but depth stops being worth anything once you're already past the follicle.

• Optical window: Skin passes roughly 600 to 1100 nanometres; blood absorbs below it, water above it.
• Target peaks: Cytochrome c oxidase absorbs near 620, 660, 760 and 825 nanometres.
• Depth needed: The bulge and dermal papilla sit within about 4 millimetres of the surface.
• Skin tone: More melanin cuts red harder than near infrared, the one solid case for adding it.

What irradiance at the scalp surface is actually enough, and what happens outside that window?

Irradiance is power per square centimetre where your scalp actually sits, and it's the figure most often left off the devices that need it most. Total output in milliwatts tells you nothing until you divide it by the area it's spread across. A cap with 300 diodes at 5 milliwatts each sounds huge until you spread 1500 milliwatts over 250 square centimetres and land at about 6 milliwatts per square centimetre before any losses.

How is total dose calculated, and what joules per square centimetre show up in published protocols?

Fluence is the one figure that ties wavelength and power into something you can compare across devices, and the arithmetic fits on a phone screen. Run it in reverse and you can tell in seconds whether a manufacturer's recommended session is even capable of delivering a dose anyone has studied. When a sheet leaves fluence out entirely, the fair reading isn't that it's proprietary, it's that the irradiance is too low to produce a flattering number.

1. Start with measured irradiance: Milliwatts per square centimetre at the scalp contact plane, not at the diode face.
2. Multiply by exposure time in seconds: A single session, not a week's worth.
3. Divide by 1000: That gives joules per square centimetre. So 10 milliwatts per square centimetre for 600 seconds is 6 joules per square centimetre.
4. Check it against the protocols: Low level light devices generally deliver about 1 to 10 joules per square centimetre, while scalp trials have reported per session surface fluences of 47.9 and 67.3.
5. Remember it's a surface number: What reaches a papilla at 4 millimetres is a fraction of the meter reading, often put at an order of magnitude lower.

What is the biphasic dose response and how does too much light stall results?

Here's where people who are trying hardest get punished for it. More light isn't better, and doubling your session or going from three days a week to daily 40 minute runs won't speed anything up. You can slide down the far side of the curve while feeling diligent, and you won't feel a thing while it happens.

Do laser diodes and LEDs change which numbers matter?

Source type changes the shape of the delivery problem far more than it changes the biology. Coherence, the property laser marketing leans on hardest, is the weakest part of the pitch, since light loses coherence in the first layers of tissue. What survives as a real difference is how narrow the output is and how it spreads.

How do scalp coverage and beam distribution change the dose a follicle actually receives?

A dose figure quoted for a device is an average, and averages hide the thing that decides whether your follicle responds. In a real treated area some follicles sit at a studied dose, some sit below threshold, and the number on the box describes neither of them. What's between the emitters matters as much as what's under them.

• Emitter spacing: Scalp under a diode can receive several times the point midway between diodes 30 millimetres apart.
• Standoff: A rigid helmet sits flush at the crown and millimetres off at the temples, where irradiance drops sharply.
• Hair as a filter: Dense, long or dark hair absorbs much of the incident red before it reaches skin.
• Zone to treat: Vertex, mid-scalp and frontal region, not only the visibly thin patches.

Why do the power figures printed on a product page rarely tell you the dose at the scalp?

Almost every number printed large on a product page was chosen because it's large, not because you can turn it into a dose. Each one drops a piece of the arithmetic, and without that piece there's no path from the claim to joules per square centimetre at your scalp. You can sanity check a claim in about thirty seconds by dividing claimed total output by a realistic treated area.

How do session length and weekly frequency turn a device specification into a real treatment protocol?

Specs turn into a protocol at the point where irradiance and your target dose decide how long the thing sits on your head and how often. The schedule in the trials isn't arbitrary, because a follicle cycles over months and nothing visible can outrun a growth cycle no matter what dose you throw at it. The device you'll actually use beats the one with the better number on paper.

1. Set session length from irradiance: Trials ran roughly 5 to 25 minutes, so 20 milliwatts per square centimetre hits 5 joules per square centimetre in about four minutes while 5 milliwatts per square centimetre needs about seventeen.
2. Run it three times a week or every other day: Spacing is deliberate, since the signalling cascade takes time to play out and stacking sessions adds nothing.
3. Give it 16 to 26 weeks before judging: That's when the trials made their primary assessment.
4. Track it with standardised photographs: Fixed intervals and fixed lighting, because two weeks of change is smaller than daily variation in styling.
5. Plan for maintenance: The published trials carried no long-term follow-up after treatment stopped, so treat it as an ongoing routine rather than a course with an end date.

Which specifications should someone ask for before buying, and which claimed numbers are meaningless?

Reduce the purchase to five questions and most of the market answers itself, because a seller who can answer all five has almost certainly measured their product. Regulatory clearance is worth having but it's frequently misread: it says a device met a safety and equivalence standard, not that this unit delivers the dose used in the supporting studies. The most useful check of all is whether the trials a product cites were run on that product or on a different device at a different wavelength and geometry.