IR LEDs typically report wavelength and radiant intensity (mW/sr). There are many warnings abound on the internet and from other local engineers to be careful of eye damage, but I find it difficult to pin down what threshold of safety exists for our eyes. In particular, I use 950nm IREDs for communication and am wondering about my safety.

According to one source I found (from Vishay), this working limit is 10 mW/cm2. This seems low compared to the output of off-the-shelf IR TX/RX pairs.

Obviously the IR intensity striking your eye varies based on distance from a diverging LED source, so how do safety limits work for existing IR emitting products (such as remote controls) and what should a reasonable safety limit on output optical power look like?

EDIT: I am interested in using the LED VSLY5940 in an open outdoor space with potentially other people working in the area and only intermittent use of the IRED. At peak driving, the datasheet states 5100 mW/sr. Should I be worried about potential eye damage?

  • \$\begingroup\$ The emitted power has low efficacy compared to power input applied to LED 10 mW/cm2 is not same as 10 mW/sr \$\endgroup\$ Commented Apr 5, 2018 at 22:53
  • \$\begingroup\$ Talk to the places that do testing such as Intertek Testing Services or similar. This area is pretty complex and depends on many things. As of around 10 years ago, France had a different interpretation than the rest of the EU so that the requirements for LED's without diffused lenses were very strict indeed. \$\endgroup\$
    – user57037
    Commented Apr 6, 2018 at 0:38
  • \$\begingroup\$ @TonyStewart I was using 10 mW/cm2 from the 100 W/m2 in the document. I did the conversion while typing and forgot the initial document units. \$\endgroup\$ Commented Apr 6, 2018 at 14:07
  • \$\begingroup\$ 10mW/cm^2 is low, but the eye can focus it to 100x that power density or more \$\endgroup\$ Commented Jun 4 at 12:12

2 Answers 2


There are no issues with NIR led power levels like those of remotes.

Laser eye safety standards cover this area. Lasers can have very high pulse power levels at modest average power, and a narrow beam can get the energy through the aperture of the pupil. A narrow source can be focussed onto a narrow spot on the retina, where the flux will be high.

So hazard requires 3 things:

  • A significant energy must fit through the pupil aperture (7mm) at viewing distances
  • The source area must be small enough that it can be focused by the eye to a concentrated spot on the retina. (this is what makes the sun dangerous)
  • the localised flux in the spot on the retina has to be high enough to cause damage, for NIR/vis that basically means heating

It the past LEDs were not considered able to put hazardous energy levels through the aperture of the pupil, and were not regulated

However there are now very high power leds. The eye protects itself from visible light by closing the iris down, but not from near-IR (or UV), which it can still focus onto the retina as the eye is transparent.

A modern 3W led die can emit 1W of IR from 1mm sq. That is a flux of 1MW/sqm (at the die surface) by comparison sunlight at earths surface is 1.3kW/sqm.

This is not quite as alarming as it might seem, solar flux is reached at 11mm from the die (180deg beam, ~40mm for 30deg beam), and at that distance you cannot focus to the tiny spot on the retina, that you can looking at the sun.

Nevertheless, led flux is now very high, and we gang up lots of leds into massive floodlight, spotlights etc.

At the other end of the spectrum - Blue, Near UV and UV there is probably more actual hazard. Blue light (and more so UV) is biologically damaging - it breaks down organic bonds. (This is probably why plant use red light, and reflect green light)

  • 1
    \$\begingroup\$ yes that's true but the Eye lens and back of the eye are sensitive to high flux of IR and not necessarily felt. Just as you should not stare at a 20,000 mcd 5mm LED up to your eyeball @ 20mA which is normally measured at 10x the distance and is still painful to watch (visible) the IR pulsed at 100mA may not cause pain. Path loss from 10mm to 1m is 40dB \$\endgroup\$ Commented Apr 6, 2018 at 14:25
  • \$\begingroup\$ Indeed you should not look closely at them. However once an NIR led flux is equal to that of the sun, the hazard is from focusing to a concentrated spot on the retina. I am talking about NIR 850nm, where the eye is still transparent. Your comment about FIR and cataracts was interesting. Older LEDs simply couldn't get there, but modern ones can at practical distances. i.e. you can get your eye within 30mm-11mm of an LED if you try. \$\endgroup\$
    – Henry Crun
    Commented Apr 6, 2018 at 21:22
  • \$\begingroup\$ I would go further to say that where they are used in a continuous setting (eg illumination, heating) we (engineers) should design the system so the people do not look straight at undiffused die. e.g. an NIR floodlight should have a front diffuser, so the field of die is not directly viewable. The same for blue, near-uv floodlights \$\endgroup\$
    – Henry Crun
    Commented Apr 6, 2018 at 21:25
  • 1
    \$\begingroup\$ I would be totally unsurprised to see issues with blue and eye damage cropping up in 10 years time, especially with so many chinese leds that seem to have phosphors that lose efficiency within 10's of hours operation, so the perpendicular emission seems to be 70% blue \$\endgroup\$
    – Henry Crun
    Commented Apr 6, 2018 at 22:11
  • 1
    \$\begingroup\$ wandering completely off topic, I replaced the kids (15yo) bedroom lightbulbs with 2200k orangey led ones, and lo, he goes to sleep now instead of staying awake for 2 hours. \$\endgroup\$
    – Henry Crun
    Commented Apr 6, 2018 at 22:54

Obviously don't stare up close against the eye ball but intensity like Friis Loss is inverse squared with distance. Since "remotes" are typ. 100mA pulsed the average intensity is much reduced.

  • Emitters are listed on the Vishay page you posted where the limit of IR interest is the cornea/lens risk in the wavelength range from 780 nm to 3000 nm.

  • Irradiance \$E_e = 100 W/m^2\$

  • Intensity \$ I_e = 4 W/sr \$ at standard 20 cm distance
  • \$I_e = E_e * r^2\$

in addition to this range , you ought to know that Far infrared or FIR can also cause cataracts (catarogenic) such as blast furnaces and bonfires when long term exposure is endured. Other than the eyes, FIR is very healthy and therapeutic and beneficial so it is used by all Dr.'s of Accupuncture.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.