View Full Version : Graininess of LED light--interference?

2007-Oct-01, 03:37 AM
Ok, for astronomical observing purposes, I have this Ray-o-vac LED Headlamp with a red light mode so I can see in the dark while still maintaining the ability to see in the dark! But, while photographing the moon last night, I stepped back inside for a "break" and looked in the mirror with my headlamp still shining red light.

The reflection of the LED light was a diffuse pool of red, as expected, but it was "pixelated"! I was standing, say, 2 feet to 3 feet from the mirror and the pool of light seemed to have a grain to it--it looked like pixels maybe 5mm in diameter. I moved my head and the pixels moved with the pool of light, proving it wasn't (just) something in the glass.

There are, in fact, two LEDs on the headlamp. And there is the fact that both the reflective backing of the mirror and, to a lesser extent, the surface of the mirror reflect, so there are at least two ways light waves could interfere with each other. However, I am not sure that a cheap LED would emit anything close to being coherent, especially with the warping of the light caused by the red casing it is in, but it is probably close to being monochromatic.

So, does anybody happen to know about this--was it interference, or did I see something completely different?


2007-Oct-01, 03:53 AM
Actually, Todd, a monochromatic source can still produce destructive interference if the light passes through 2 or more slits or pinholes....without having to be originally emitted from a coherent source.


Likewise rear surface / front surface interference doesn't require coherent light .... That is how thin film coatings stop reflective glare. It is usually based, however, upon a front to rear surface distance that corresponds to the wavelength .


BTW, it is possible 2 or more small ridges in the plastic casing can act like slits.....As long as it DIFFRACTS the wavefront in two different positions it can create a interference pattern from monochromatic light.

2007-Oct-01, 01:51 PM
LED's are not monochromatic, but I imagine a red casing(filter) can help.

2007-Oct-01, 04:47 PM
ok, thanks--

I tried again last night--it only works if I am exactly the right distance (about 3 feet) from the glass. Also, covering one LED or the other changes nothing (but the brightness).

So, given that red is about .7 microns or so, the 5mm or so thickness of the mirror isn't the "culprit". It must be flaws in the red casing acting as a diffraction mask.

Incidentally, the 5mm period of the "bumps" or pixels made me think of "beats" in music, though I don't see how two (as opposed to lots if not exactly monochromatic) frequencies would arise.


Robert Carnegie
2007-Oct-03, 12:39 AM
I wondered if an LED capsule comes with a lensed cover, to scatter the light. But according to a description here http://www.fiber-optics.info/articles/LEDs.htm a surface-emitting LED emits light spread around, anyway. But what I don't know is how small the actual light-emitting surface is. After all, an incandescent lamp room-lighter bulb isn't a point-source of light, but you usually use it with a cover or a frosted glass shell, so it doesn't dazzle when you look straight into it. Come to think, that is behaviour that you're warned against with some LED lighting - whoops!

I also found http://www.patentstorm.us/patents/6649939.html which appears to refer to a patent applied for in 2002 by Osram, granted in 2003, for a light-emitting diode inside a scattering lens (actually pyramid) structure that I had in mind for your situation.

...you don't have one of those experimental silicon chip retinas in your eye, that could be pixelated, possibly (I don't know if they are)? ;-)

Does this effect interfere with your observations, besides keeping you studying your mirror instead of the telescope??

2007-Oct-03, 02:25 AM
Not really, since I have to be 3 feet from the mirror. I tend to look at notes, etc., close-up.

Ok--on the other hand, its brightness can affect observations, even though red is not supposed to damage night vision--it merely damages it *less* than white light. SO, if I were to look right into it, I might have to wait before I can see faint objects.

2007-Oct-03, 03:57 AM
Can you get a picture of the effect? It'd be interesting to see.