"A ground glass plate attached to a variable speed, DC motor to create a pseudothermal light source to ensure the detection of Bose-Einstein statistics in the appropriate regime. The speed of rotation of this plate is inversely proportional to the correlation time of the scattered light. Thus, the speed of the plate is altered in order to alter the coherence time."

You didn't mention the B-E distribution in your initial post. How does that fit into the chaos surrounding the terminology of all these light sources?

RocketRoo, thank you for your comment, but I am not sure where you are heading with this comment. The term "Bose-Einstein statistics in the appropriate regime" is somewhat nebulous.

Usually when I think of Bose-Einstein condensates (BEC), temperatures below a degree Kelvin come to my mind, while we normally do color science at room temperature. Classical experiments with BEC are based on bosons like rubidium isotopes, which condensate into molecules. When you sweep a magnetic field you get a stable condensate due to interference among the atoms, but if the magnetic field strength is too high, you get a "bosenova" implosion, which is something I would not like to happen during a color measurement. ;-)

What is the equivalent of a molecule for a BEC of photons? Is it the slow-light of Lene Vestergaard Hau? She recently had an intriguing letter in Nature 445, 623-626 (8 February 2007) where she was storing photons in "optical dipoles". The amazing result was that two photons stored in two independent BCEs can become indistinguishable, which would be symmetric in time to entanglement.

Getting back to the room temperatures in which we normally work, do you suppose BCE can explain photon bunching? What is the important distinction of light sources with bunched photons versus light sources where the photons are stationary, i.e., have independent arrival times at the detector?

Second point to keep in mind: All light is produced by matter (except perhaps for the Big Bang, but nobody seems to remember what happened there). Therefore, all light is produced by atoms. From the standpoint of QED, that means all light is produced by the electrons in the atoms. Since the electrons are bound to a nucleus, their energies form discrete levels or shells and it is the transitions between these levels that determine the energy (finite frequency) of the emitted photons.

But wait, there's more! BE stats has to do with the NUMBER of emitted photons in a given quantum state. BE predicts that I can see more than one photon emitted at a time, which leads to super-Poisson behavior (cf. your comment about Poisson laser light) or so-called "photon bunching". Can we see this effect? Yes, but not directly. The average number of photons in a given state is the same thing as measuring the intensity of the light (number of photons arriving per unit time). In atomic transitions, the number of photons being emitted is random (or "chaotic") but correlated because of the BE effect. Therefore, the intensities measured at slightly different locations in space or time can appear (weakly) correlated. The eye is only sensitive to the intensity of light, but cannot see intensity correlations. For that you need special equipment to measure the correlations.

For various reasons, there has been an intense(pun?) interest in correlated photons, and that is the unwritten sub-text in the seemingly muddled quotes above. All light produced by atomic transitions is called "chaotic" these days (for better or worse). It also includes the effects due to atomic collisions (Lorentzian line shape) and Doppler shifting of the frequencies (Gaussian line shape).

To summarize so far, all atoms are "chaotic" BE emitters. This, however, begs the question: Why make such a redundant distinction? The reason is that so-called "thermal sources" are not BE emitters. The classic thermal source is essentially a black-body e.g., the sun. But, I hear you ask, "You just got through telling me that all light comes from atoms, and the sun is made of atoms! What gives?" Correct, and I thought you'd never ask.

Black-bodies emit according to the Planck distribution (discussed elsewhere on this blog), which is a continuous spectral distribution. In that case, because of the (nuclear) energies involved, the electrons have been stripped from the hydrogen and helium atoms and are no longer bound to their respective nuclei. Hence, the spectrum is no longer discrete. The electrons no longer act as tiny oscillators. This state of affairs also washes out any BE stats and, indeed, the intensity fluctuations from a black-body source look Poissonian.

And finally, before you put finger to keyboard in response, starlight is black-body radiation (obviously, if the sun is). However, because of Wien's law, the peak spectral wavelength is inversely proportional to the temperature. Hence, distant stars appear colored (even to the naked eye in some cases). With the appropriate filters one can capture their light as if it was "pseudo-thermal" (a term also used more or less synonymously with "chaotic") and once again, detect the BE effects. This is what allowed astronomers to go from Michelson (phase) interferometry to the much more stable and accurate intensity interferometry.