630nm, 660nm and 850nm: Reading a Red Light Spec Sheet

Every red light spec sheet is studded with numbers followed by "nm": 630nm, 660nm, 850nm, sometimes 605nm and 460nm as well. Those numbers are nanometers, and a nanometer figure is a position on the light spectrum. Nothing more and nothing less. It is not a strength, not a dose, and not a grade.

That single fact clears up most of the confusion in this category, because a lot of listings present a wavelength as though a bigger number were a better product. This guide covers what the common figures actually describe, why one array can look bright and another almost switched off, what the plus-or-minus tolerance means, and what to do when a spec sheet gives you no wavelength at all - including 2 devices in our own catalog.

Two families, and one of them is invisible

Almost everything in this category sits in 1 of 2 bands.

Visible red runs roughly from 620nm to 700nm. This is the glow you buy the device for in a visual sense: 630nm and 660nm are the 2 figures you will see most often, and both of them are plainly, obviously red to the eye.

Near-infrared starts past the edge of human vision, around 700nm, and 850nm is the figure that dominates spec sheets. Your eye cannot register it. An array running near-infrared emitters on their own looks nearly off - a faint, dull red haze at most, coming from the small amount of visible light the emitter leaks.

That last point is worth its own paragraph, because it is the single most common source of "my device is broken" messages in this whole product category.

If your device has a near-infrared-only mode and it looks dim, that is the mode doing what it says. Trust the indicator light on the controller rather than the array. Our mask and panel comparison goes further into how the different formats present their modes.

The plus-or-minus you should look for

An LED does not emit a single mathematical wavelength; it emits a narrow band around a center figure. Careful manufacturers print the tolerance. Ours, on the flexible pad, is the clearest example in our range.

A tolerance is a small, useful signal about the maker. It says somebody measured, and it tells you that "630nm" on that sheet means a band from 625nm to 635nm rather than a marketing round number. A sheet with no tolerance is not automatically worse - most sheets have none - but a sheet that gives you one has told you something about how it was written.

What the Rubalume catalog publishes

Those last 2 rows are the honest part of this guide. We could have written a number in them - the internet is full of confident nanometer figures for devices whose makers never published one - and we are not going to. If a wavelength is not on the manufacturer's sheet, then as far as anyone outside the factory is concerned, it is unknown.

630nm or 660nm? Both are simply red

These 2 figures sit 30nm apart, well inside the visible red band, and you would struggle to tell them apart side by side. Why does a catalog like ours carry both? Because LED supply does. Emitters are manufactured and sorted into bins by their center wavelength, and a maker designs around the bins that are available, affordable, and consistent at the volume they build.

We are not going to tell you that one of these numbers beats the other, because nothing on a spec sheet supports that claim. What a spec sheet can support is the more useful comparison: how many emitters there are, how they are laid out, what the timer does, and how the device is powered. Those are the figures our buying guide is built around.

How one bead can carry more than one wavelength

A modern LED bead is a package that can hold several separate chips, and this is where the eye-catching totals on listings come from. Our belt is the cleanest illustration: 105 beads, each holding 3 chips, in a fixed arrangement of 1 red emitter and 2 near-infrared emitters. That is 105 red and 210 near-infrared, from 105 physical dots.

So when you compare 2 listings, check whether the headline number counts beads or chips. "720" and "180" describe the same pad. Neither figure is dishonest; they are answers to different questions, and only one of them tells you how many dots are physically in front of you.

The amber and the blue on a 4-mode pad

Our flexible pad is the only device in the range with more than 2 wavelengths: alongside 630nm red and 850nm near-infrared it carries 605nm amber and 460nm blue, and you move between the 4 modes by short-pressing the mode key on the wired controller. Both extra figures sit in the visible part of the spectrum, so unlike the near-infrared mode you can see exactly which one is running.

Frequently asked questions

Does a higher nanometer figure mean a stronger device?

No. A wavelength is a position, not a quantity. The specs that describe how much light reaches you are output power, the number of emitters, the area they cover, and how far away you sit. An 850nm array can be modest and a 660nm array substantial, or the other way round.

Why do so many devices pair red with 850nm?

Because the 2 bands are produced by different emitters, and putting both in one package is how a maker offers 2 modes without building 2 products. When a sheet lists "660nm and 850nm" it is usually describing a bead or diode that contains both, which is why the mode switch on the controller is often the only visible difference.

Can I check a wavelength myself?

Not reliably by eye, and not with a phone camera. Phone sensors do register some near-infrared, which is why a camera can show a faint glow where you see none - a neat trick for confirming an array is emitting something, and no help at all in telling 630nm from 660nm. Measuring a center wavelength needs a spectrometer.

What if a device lists a range, like 630 to 660nm?

Read it as a hedge unless a tolerance is attached. A stated band 30nm wide usually means the maker is describing a mix of emitters, or is leaving room for whichever bin the factory has. It is not a red flag on its own; it is a reason to look harder at the specs that are stated precisely.

Do the numbers change how long a session should be?

The timer is set by the device, not by the wavelength. The mask ends its session at 10 minutes on its own; the panel and belt offer steps from 5 to 30 minutes. Our session guide works through what each device's own manual asks for.

Keep reading

Rubalume devices are wellness devices, not medical devices. Our guides describe what the devices are and how they are used. For the formal details, see our Wellness Disclaimer.