21 CFR 1040.10 takes that into consideration - I believe I correctly teranslated this into 63 joules per square centimeter of an "ideal" "lambertian" light source surface for exposure period of 10,000 seconds as an eye exposure limit to neither exceed nor match. And if your eye exposure is less than 1 millijoule in any 2500 second period, you are OK. And if you violate one but not both of these, you are OK.
And if the light source is not a laser, then 21 CFR 1040.10 is not a law that you are legally required to follow, although I suspect your lawyer might advise it to be a guideline for CYA-ing with. (And 21 CFR 1040.10 is a USA thing - I do not guarantee any advice of mine good for so much as what you paid me to read it anywhere else!)
Up to a certain area that is small enough to be a point...
21 CFR 1040.10 makes me think that 21 CFR 1040.10 has a point source no worse than a source with surface area up to about 1/32 millisteradian, or roughly .31 square centimeter (or roughly .63 mm dianmeter circle) at 1 meter distance. This is for exposure time 1 to 2500 seconds - the equivalent safe source area varies proportionately with exposure time (within a 10,000 second periond) from 2500 to 10,000 seconds and then things level off with maximum safe exposure having point and area source exposure being equal at about 1/8 millisteradian of viewed surface.If I screwed up here, someone please say how!
- Don Klipstein ( snipped-for-privacy@misty.com)