630nm, 660nm and 850nm: How to Actually Read a Wavelength Spec Sheet
The Number on the Listing Is Not the Whole Story
Scroll through any red light device listing and you will see the same three numbers: 630nm, 660nm, 850nm. They appear in titles, bullet points, and comparison charts — and for many buyers, the decision ends there. But two devices can both say "660nm" and deliver noticeably different light, because a wavelength figure on a spec sheet carries more information than the number itself. This article walks through how to read that documentation properly: what a nominal peak wavelength means, what tolerance and bandwidth add to the picture, how dual-chip LEDs complicate the count, and why every irradiance figure needs a distance attached to it before it means anything.
This is a spec-sheet literacy guide, not a general buying checklist — the broader selection factors are covered in our device choosing guide. Here we stay on the numbers themselves.
Nominal Peak, Tolerance, and Bandwidth
What "660nm" actually states
When a spec sheet says 660nm, it is stating a nominal peak: the wavelength at which the LED emits its maximum intensity. It is not a guarantee that every photon comes out at exactly 660nm. Real LEDs emit across a spread of wavelengths centered on that peak, and manufacturing variation shifts the center slightly from unit to unit and even from LED to LED within one array.
Tolerance: the ± that separates suppliers
Well-documented devices state a tolerance, typically written as 660nm ±5nm or ±10nm. A tight tolerance means the supplier has binned their LEDs — sorting production output so each unit's channels sit close to the target. A supplier who cannot state a tolerance, or who cannot explain where the number comes from, is usually quoting the LED datasheet rather than measuring the finished device. This is one of the quiet differentiators in sourcing: the difference between a marketing number and a tested number.
Bandwidth and why it matters less than you think
LEDs have a relatively wide emission band compared with lasers — a 660nm LED may emit meaningful intensity from roughly 620nm to 700nm. This is normal and not a defect. What matters is that the stated peak and tolerance reflect the assembled device, which is why per-unit spectrometer testing on the production line is worth asking about. Our article on how red light devices are tested before shipping covers what that process looks like in practice.
The Three Common Channels, Structurally Speaking
Without making any claims about outcomes, it is useful to know why these particular numbers dominate the category:
- 630nm — red light at the shorter end of the visible red band. Commonly included in facial devices where visible red output is part of the product experience.
- 660nm — deep red, the most widely used single figure in the category. It sits at the center of the visible red range most device designs target, and it is the number customers will search for.
- 850nm — near-infrared. It is invisible to the eye, which surprises end customers the first time they power a device on. Near-infrared channels are commonly paired with red channels so a device offers both a visible and a non-visible mode.
Some designs add amber (around 590nm–605nm) or blue channels. The sourcing question is the same for every channel: is the peak documented per channel, with a tolerance, and tied to measured output?
Dual-Chip LEDs: Read the Count Carefully
Many modern devices use dual-chip LEDs — one package containing both a red die and a near-infrared die. This changes how you read the numbers. A panel listing "300 LEDs" with dual chips is not 300 red LEDs and 300 infrared LEDs; it may be 300 packages delivering two channels each, or a mixed ratio such as red-heavy or IR-heavy depending on the design intent. A clear spec sheet states the chip configuration per package and the channel ratio. If a listing leads with an LED count and does not clarify the configuration, that is the first question to send the supplier.
Irradiance: Every Number Needs a Distance
mW/cm² means nothing without "at X cm"
Irradiance — the power delivered per unit area, quoted in mW/cm² — is the other half of the spec sheet, and it is the number most often quoted misleadingly. Light intensity falls off with distance from the source, so a panel measured at 3cm will show several times the figure of the same panel measured at 30cm. Listings that state "120 mW/cm²" without a distance are technically quoting something, but the figure is not comparable with anything else. A trustworthy spec sheet states irradiance at a defined measurement distance, and ideally provides a curve at several distances rather than a single point.
Device type changes the working distance
This is also why irradiance figures cannot be compared across device categories. A face mask sits close to the skin and works at a short distance; a panel is used further away; a wearable wrap sits against the body. Each has its own realistic working distance, so each needs its own measurement conditions. Our red light waist wrap, for example, is documented for contact use, and comparing its contact-distance figure against a panel's 30cm figure would be meaningless.
Dose is the product of two numbers
Spec sheets may also quote radiant exposure in J/cm². That figure is not a property of the device alone — it is irradiance multiplied by session time. Any device with an intensity setting or wavelength modes delivers a different dose per mode, which leads to the last section.
Wavelength Modes: Each Mode Needs Its Own Numbers
Most multi-wavelength devices let the user switch modes — red only, near-infrared only, or both. Mode switching changes which LEDs are active, which changes both the spectral output and the irradiance delivered. A complete spec sheet documents irradiance per mode, not just for the "all on" setting. When evaluating a device, ask for the per-mode table; if the supplier only has one irradiance number for the whole product, the documentation is incomplete.
Five Questions That Test a Spec Sheet
- Is the peak wavelength stated per channel, with a tolerance?
- Is the tolerance based on binned LEDs and measured on finished units, or copied from a component datasheet?
- For dual-chip configurations, what is the package structure and the channel ratio?
- At what distance is irradiance measured, and is there a per-mode breakdown?
- Can the supplier produce per-unit optical test records, not just a batch summary?
Seeing It on a Finished Device
To make this concrete, look at how a finished product documents its configuration. The RXpre521 silicone LED facial care mask states its wavelength channels with mode-based operation, and the full documentation set — spec sheet, LED layout, and per-channel figures — is available on request. Comparing that layout against the checklist above is a good exercise before you evaluate any supplier's documentation.
Want the complete spec sheet and per-mode irradiance table for a model you are evaluating? Message us with DM SPEC and the model name, and we will send the full documentation package.