The inconceivable nature of nature

February 21, 2009 11:41 AM

How much do criss-crossing electromagnetic waves affect one another?

Richard Feynman famously said this:

http://www.youtube.com/watch?v=AU8PId_6xec

The picture is that the electric field consists of a lot of jiggling, and that information is going every which way. In that case, when one stream of information is going one way, how much is it getting distorted by myriad streams that are passing right through it from other directions?

posted by Eiwalker to science & nature (8 answers total)

My understanding (from physics lessons half my life ago) was that the interference only has an effect specifically where the signals coincide; the interference does not disrupt the ongoing propagation of the stream.

For example, two signals (sine modulated for simplicity) are travelling in an X pattern; there may be local interference observed where the two waves cross - the signals will either increase the observed magnitude, or cancel each other out (or somewhere between these two extremes), but the signal measured after the cross over is identical to the signal prior to the cross over.

There's only really an issue if two signals are being transmitted in the same direction/polarity as they will travel together and it is impossible to determine the original signals from the observed superposition.

Or something.
posted by Chunder at 11:54 AM on February 21, 2009

That's kind of one of the questions in physics. The answer is... complicated.

But on the macro-level, the answer is "Hardly at all." True, certain experimental setups can detect photon interference, but if you shine two beams of light such that their paths cross, neither beam will be affected by the other. Same goes for invisible wavelengths too.

Think about it: your cell phone and radio all work perfectly fine, but they're constantly bombarded by broadcasts in every conceivable spectrum, from x-rays through radio and visible light and up to gamma rays with absolutely no interference. When we do talk about radio interference we don't mean that the signals are actually affecting each other, we mean that we can't listen to two broadcasts on the same wavelength at the same time. Both signals reach the receiver perfectly clearly, but there isn't any meaningful way of sorting them out, so the result is like listening to two people talk at once.

But the actual information being broadcast is completely unaffected by the presence of competing broadcasts, even those on the same wavelength.
posted by valkyryn at 11:54 AM on February 21, 2009

It's not.

Wrong. Google the term "photon-photon scattering". There is very much an amplitude for photon scattering in QED, the simplest contribution involving a photon -> electron, positron pair -> photon loop. In a sense, the vacuum has an index of refraction for sufficiently high energy photons (>10^20 GeV, if I recall. Yes, that's huge.). It is too small to be directly measured in the lab, however this effect (and many others) adds to the directly measured and confirmed magnetic moments of fundamental particles.
posted by fatllama at 12:09 PM on February 21, 2009

Well, at the macro level, not at all. His comment is basically noting how weird that is... it seems like it should.
posted by phrontist at 12:10 PM on February 21, 2009

Wrong

Photon-photon scattering might be the technically correct answer, but that's not the helpful answer. The scattering cross section is proportional to the square of the fine structure constant for photons of reasonable energies. You're right that the effect is there, but at that level it's totally undetectable. Pointing out it's useful for QED doesn't answer the question well.


On topic now: They don't. Think about waves in water. If I set up some waves such that they will cross each other at a right angle, what happens when they meet? They pass right through each other. At the intersection the waves are superimposed on each other, which is critical. The two waves aren't competing for which direction the water gets to be "shaken", instead they add together in the overlapping region, and come out as normal waves on the other side. It all comes from superposition: the electromagnetic field in a region where two waves meet is the same as the EM field from one wave plus the EM field from the other wave. Simple, but very useful.

Alternatively, point two flashlights such that the beams cross. The intensity of both beams stays the same (to within 0.001%), and you don't get any weird third beam from the "collision" of the other two. Visible light and radio waves are both just electromagnetic waves, so radio waves would have the same behavior.

(This is at least the case in a vacuum or in air. This behavior is not guaranteed in all media.)
posted by kiltedtaco at 12:42 PM on February 21, 2009 [1 favorite]

It's not.

Wrong. Google the term "photon-photon scattering". There is very much an amplitude for photon scattering in QED, the simplest contribution involving a photon -> electron, positron pair -> photon loop. In a sense, the vacuum has an index of refraction for sufficiently high energy photons (>10^20 GeV, if I recall. Yes, that's huge.). It is too small to be directly measured in the lab, however this effect (and many others) adds to the directly measured and confirmed magnetic moments of fundamental particles.

fatllama, you're being overly picky. Your own statement "It is too small to be directly measured in the lab" is a sign of how, in all practical, normal terms, they have no perceivable effect on each other.

And what Chunder said:
My understanding (from physics lessons half my life ago) was that the interference only has an effect specifically where the signals coincide; the interference does not disrupt the ongoing propagation of the stream.
... is roughly true. Substitute "to the degree that the two beams overlap, in the same direction, with the same polarization" for the word "coincide", and it will be more accurate. That is, beams intersecting at 90 degrees have no measurable effect; like-polarized beams that intersect at 45 degrees will have 71% (cos 45 degrees) of the effect of beams that are fully parallel; and beams that are fully parallel, coincident, and share the same polarization will either constructively or destructively add their fields, changing the output power somewhere between 0 (annhilated) and 4x the power of either separate beam (assuming they are of equal field strength).

Whew.

[IANOpticalEngineer.]
posted by IAmBroom at 12:50 PM on February 21, 2009

D'oh.

Make that, IAAOpticalEngineer.
posted by IAmBroom at 1:23 PM on February 21, 2009

One way to think about it...
The Hubble telescope takes pictures of galaxies billions of light-years from us.
The photons it captures have been traveling for billions of years, through a veritable sea of other photons traveling in all directions.
Yet the telescope sees them clearly enough to form a good image.

So. Yes, there is some scattering. But it is a tiny tiny effect. Basically, waves are unaffected by other waves traveling in other directions.

This is not true of waves on water (which have different propagation rules).
posted by hexatron at 3:47 PM on February 21, 2009