Quantum mystics

That's false! Entanglement and interference can easily be understood in terms of waves and quantized detectors. It's the QM view, with its imagined photon particle flying everywhere at once that is confusing.

What size do you imagine a photon to be?

Jeroen Belleman

Do you have any friends?

But the wave phenomena in some experiments (aka wavefunction) can belong to comparatively heavy objects that we would normally think of as classical particles. Indeed we can even image the molecules used at atomic level with scanning tunnelling microscopes.

I'm pretty sure they have diffracted buckyballs through Young's slits. I think the record for complexity is still held by a fluorinated porphyrin ~10k amu 800+ atoms and efforts are underway to diffract a small virus.

More info on Arxiv here :

Physical intuition tends to break down when you have a superposition of quantum states involved. Attempting to know which slit a particle actually went through destroys the interference pattern and experiments using ultra low flux levels with just a single photon in at any one time still show a diffraction pattern. QM is decidedly counter intuitive.

Explores all available paths mathematics gets the right results but I can't help feeling that there is a way to avoid the action at a distance implied by quantum entanglement when we get all of the physics correct.

I didn't think his talk was all that outrageous. A bit over simplified perhaps but then avoiding almost all of the maths that is inevitable.

But that world view is backed up by experiments.

Particles can behave as waves and waves can behave as particles depending on the experiment. The particle isn't "everywhere at once" either it is trapped in a spherical shell radius vt expanding around its point of origin with the amplitude of the wavefunction representing the chances of finding it at any particular position.

Depends on the wavelength of the photon but to have a well defined frequency the amplitude envelope has to be a good few wavelengths long and to agree with causality the leading edge must be zero until sufficient time has passed from its emission to reaching its target. I expect that there is a canonical shape for a photon amplitude envelope for given df/f but I don't know what it is or if it has ever been computed.

This aspect of size of a photon always seemed very awkward to me when working at 21cm neutral hydrogen and measuring what are essentially tiny correlations in narrowband random noise from extremely remote mostly point sources over a large number of different antenna pairs. What is pretty clear is that the correlations of such signals are good enough even on planetary dimensions for VLBI to work!

Sticking with the semiclassical picture of photodetection is good, because it avoids almost all of the blunders made by the photons-as-billiard-balls folk, but it doesn’t get you out of the mystery.

The really mysterious thing about photodetection is that a given photon (*)

incident on a large lossless detector gives rise to exactly one detection event, with probability spatialy and temporally weighted by E**2.

Doesn’t seem so bad yet, but consider this: If the detector is large compared with the pulse width/c, distant points on the detector are separated by a spacelike interval.

That means that so when point A detects it, there is no way for the information reach point B before the end of the pulse, when E drops to zero, and yet experimentally point B doesn’t detect it.

(*) a quantized excitation of a harmonic oscillator mode of the EM field in a given set of boundary conditions)

Cheers

Phil Hobbs

(Edited for clarity—posting from my phone)

He certainly doesn't have a job.

How often do you use real, symbolic calculus?

Solving differential equations?

It's unlimited. You can have an interferometer with different arm lengths and still get single-photon interferance.

I noticed that on a lithium niobate Mach-Zender e/o modulator. The interfering path lengths are different by thousands of wavelengths.

I'm not sure that it *is* the teaching. It is more a cultural thing. It is OK to be "bad at maths" but not OK to have not read Shakespeare.

I blame the teaching for the appallingly high proportion of electronics engineers that refuse to accept Special or General Relativity though.

Not quite daily but more than once a week. These days I tend to throw a lot of it at Maxima or Mathematica rather than do grunt work by hand. Even so mechanical tools need guiding towards the right answer.

Both have some annoying features/quirks that you have to work around.

Less frequently. Mostly I'm involved in finding faster rational approximations to awkward non-linear equations or linearising things that don't really want to play ball. The idea is to have a fast approximation that is good enough to act as a seed value for a NR/Halley/higher refinement step to always converge.

Much of that relies on calculus of variations which is another step above the sort of basic calculus taught in schools. It allows you to compute real world things like how a cable will hang between two poles.

Likewise for tensor differential analysis working in non-Euclidean curved coordinate frames (I haven't really used that in anger for decades now). OTOH being exposed to it broadens the mind.

Exactly! The path length difference is limited only by the coherence length of the light source. This is all quite natural when thinking in terms of waves. When you think of it in terms of photons, it stops making any sense.

Jeroen Belleman

We don't have single-photon-on-demand sources, nor perfect detectors. Both sources and detectors are probabilistic. I'd like to see how this argument fares using energy resolving detectors like TESs.

I do not expect the probability of a detection event in one spot to be affected instantly by a detection event somewhere else. The collapse of the wave function is an attempt to apply statistical reasoning to a single event.

Jeroen Belleman

A single photon has an infinite coherence length.

What's weird is that I can pulse a superfast laser and hit a detector with picosecond time delay jitter, even though another experiment shows that each photon is very long.

It's apparently easy for you to accept that light is made of waves until it's detected, at which time it turns into particles.

That's the part that's magical to me.

Higher energy photons, like gamma rays, can be detected with 100% probability. They pack a lot of energy.

University of Vienna, Zeilinger's fief, again. I'll have a closer look to see what tricks they played.

Over-simplified to the point of being devoid of meaning, indeed.

Jeroen Belleman

Not all that often. For most functions I encounter, I already know the solution.

From time to time, using Laplace for the continuous domain, and Z-transforms for discrete-time things.

Jeroen Belleman

I wouldn't say it like that. I'd say that the incident wave causes a detection event. I'd never say that *light* is a particle. Where matter and waves interact, quantization occurs.

Jeroen Belleman

That is the problem with popular science lectures about QM. This one - a Nobel prize lecture by Serge Heroche from 2012 is a lot more meaty and the experimental techniques they used and perfected are breathtakingly cunning. The audience has quite a few famous physicists in it.

He is wonderfully self effacing and shares the credit for the success of his experiments very generously with his co-winner many collaborators, his team and graduate students.

Non destructive sensing of single atom quantum states is incredibly impressive! I didn't know until I saw that talk that the Schrodinger's cat wavefunction has been experimentally verified.

Basically he has constructed a real life particle in a box experiment!

It took ultracold superconducting hyper polished mirrors to realise it.

No they can't. It isn't called penetrating radiation for nothing. Most of it goes straight through all but the densest of targets. The odd one gets lucky and hits something and then we see scintillation.

Some of the solid state NaI(Tl) detectors are getting pretty close to

Likewise higher energy X-rays goes straight through most matter like it wasn't there. You are always playing a numbers game of detections being a fraction of the flux passing through (unless the target is optically dense) which Earth's atmosphere is for most ionising radiation.

Ultra high energy cosmic ray particles generate an airshower of less energetic secondary particles most of which wouldn't reach the ground except for the effects of relativity. Some of them have been estimated from the total yield to be carrying as much energy as a cricket ball. Not bad for a single Fe56 nucleus!

Ground based detectors can measure the secondaries, timings and spatial distribution and make a reasonable guess about the energy it had.