[Eternal Student]

Hi.

We might have lost the OP.  I'm inclined not to continue saying much until they've had a fair time to reply.   At any point you ( @varsigma ) can steer the discussion in some direction.
- - - - - - - - -
@alancalverd  has made some good points.   Thanks for spending some time to look up something about Stern-Gerlach apparatus and bothering to make a reply,  it is appreciated.   However, measurement and exactly what constitutes a measurement is just too big a topic for me to start discussing at the moment.

Best Wishes.

[alancalverd]

This is all good serious stuff and stimulates the grey cells!

The S-G experiment was part of the undergraduate syllabus 60 years ago, and, beneath the layer of dust, my notes are all about polarisation and selection, not measurement! Maybe time to grumble to Wikipedia?

I've been thinking about the 45 degree polariser. Quite simply, it comes down to the resolution of a vector. If we represent the E field of a polarised  EM wave as a unit vector in the x plane, we can resolve it as the sum of  1/√2 vectors at  ±45° to the x plane. So the (idealised) polariser at +45° will transmit about 1/√2 of the incident energy as an EM wave at 45° to the original plane. Then the third polariser does the same, resulting in about half of the original beam being transmitted. Again, this isn't "measurement"  but manipulation by absorption and retransmission through the polar medium.

[Zer0]

The discussion was soo Energetic...
The OP seems to have lost all Energy..
To Charge ahead!

[Eternal Student]

Hi.

Well... this was something I had prepared, with a couple of questions to work through, if you're bored, @Zer0 or anyone else.   But as usual it started to get too long.   It's only going to get a person half-way to a quantum model of light and how filters work.  I didn't really have the time to finish it  PLUS the common sense to assume no-one would read it.   BUT,  if you're bored then there are two questions here to work through, like a lunch-time crossword puzzle etc.   I've not checked the spelling or grammar or done much editing.

I've been thinking about the 45 degree polariser. Quite simply, it comes down to the resolution of a vector.

    Yes.   The Classical explanation of how a polarising filter works is exactly like this.  Well done and keep the dust off those old notes, they're always useful.
    Using classical models, an idealised polarising filter could be imagined as something with conduction electrons that can oscillate in only one direction (e.g. some thin wires running in one direction like a multi-wired chesse-slicer for people who want to cut mutiple slices and aren't prepared to do it one slice at a time).   For optical wavelengths of e-m radiation, the wires are really small and really closely packed together so that a crystalline substance is the best way to go because no-one would want cheese cut that thinly.  Anyway, an idealised polariser is able to absorb all the oscillation that is in the direction of its orientation when an e-m wave passes through.   Only the oscillation perpendicular to its orientation is transmitted.

     The only minor detail is that the energy in an e-m wave is proportional to the square of its amplitude.   So the amplitude was reduced exactly as you claim   (1/√2 for a polarising filer at 45 degrees) but the energy is reduced by 1/2.   (The remaining calculations you suggested then have a follow-through error).
     The next thing that can be interesting is to ask yourself what the frequency of the transmitted light must be.    If the light incident on the filter has frequency  f   and all the filter does is remove a vector component of that oscillation what is the frequency of the transmitted light?
   
     (a)   It is still  f.   (It may have a different polarisation but the greatest deflection along any axis would be precisey when the incident light was at its greatest deflection etc.)
     (b)   f_out  ≠  f_in .

[Minor aside, which you probably already know:  Photographers used to like using a pair of polarising filters and just adjusting their orientation relative to each other as a way of controlling the total brightness of what they are filming or photographing.   It reduces the amplitude of everything quite uniformly, e.g. putting the filters at 45 degrees to each other halves the intensity at all frequencies.  In particular it does not change the colour balance or favour some frequencies over another.   Using some dark coloured glass is cheaper but then the filter does knock out some colours more than others.   It doesn't matter so much with modern technology - light editing on photoshop will correct a red saturated image of a bus into a stunning photo of a young man with perfect teeth.]

   Now, assume you turn the incident light down in intensity as much as you can.   With modern technology and a laser pulse, let's say you could get only a few photons to hit the filter.
   The next  multiple choice question has to be:   What will be transmitted at the other side of the filter?  Somehow you must have a lower intensity at the other side of a filter (unless the orientation was precisely inline with the polarisation of the incident light).
(a)   Same number of photons (incident and output), with typically lower frequency on the output.  E = hf,  lower f, lower E.
(b)   Same number of photons and they also have the same frequency.   A component of the amplitude in some direction was removed, so the photons have a bit less energy.
(c)   Light is quantised.  So you get a fraction of the photons at the other side rather than fractional-photons or anything that could have satisfied (b).
(d)   Something else.   Just don't consider photons.
---------
You'll probably take option (d) though - just don't use Qunatum theory.   That's ok, it's a choice.   QM is just a model.

Best Wishes.

[alancalverd]

Usual problem of mixing two partially adequate but incompatible models of EM radiation.

First, I admit to the "amplitude/energy" error. It occurred to me whilst driving home yesterday but you beat me to publication!

The simplest way we can discuss polarisation of EMR is with a classical wave model, which fits neatly with our models of the electronic structure of chiral molecules and all the way down to radiofrequency polarisers and reflectors. It is less obvious why a chiral molecule or wire grid should preferentially transmit a boson with spin up or down, and even less obvious how it can rotate the spin of an incoming photon, nor what "frequency" means to a particle model - surely some energy is lost in transmission so why do the exit photons have the same energy as the incident photons (they do - your sunglasses are grey, not red!)? 

However it is clear that at very low intensities we can count individual photons, so we are back to an adage I used in some other threads: our best model of EM transmission and propagation is Maxwellian, but detection (particularly at low intensities) is distinctly a quantum phenomenon.   

Your option c is the only one that makes sense and describes the "black box" transfer function of multiple polarisers, but whilst Stern-Gerlach segregation is explicable in terms of spin/field interaction  it gets very difficult to explain what happens inside a molecular polariser box unless you use a wave model.

As for the photographic use of polarisers, it seems like a complicated way of reducing intensity, compared with simply reducing the aperture of the lens, which doesn't affect the visual impact of the image. A single polariser, however, is very useful for suppressing the polarised glare from a reflective surface and thus increasing the contrast or conspicuity of the object of interest. Excellent sunglasses for coarse fishing and gliding, but dangerous for driving!

PS - my undergraduate notes on Stern-Gerlach are from the quantum physics course, not classical electromagnetics!

[varsigma (OP)]

Interesting that the Wikipedia article, when discussing sequential S-G  systems, talks about "measuring" when a particle passes through a nonhomogeneous magnetic field. When we use magnetic fields to select regions for analysis by spin resonance (e.g. MRI) we talk about polarising or forcing, not measuring*. 

What I've seen about measurement of quantum states is that preparing a state, such as by applying an external magnetic field, is the same thing as a measurement. It's really about deciding where the inputs and the outputs are, in terms of the quantum information.

I read a sort-of interesting SciAm article which included a discussion of SG and the way you don't get a Boolean logic if you try to build a circuit, so to speak, using SG 'gates'.
The reason as I recall, is that spin doesn't distribute logically; you can't write down an expression like "spin is up in the x direction and spin is up or down in the z direction", after doing some measurements.

p.s. I've been following the discussion but haven't had a lot of opportunity to post anything. Also there are some problems with my ability to connect. I'm back now though, so, hi.

[alancalverd]

Usual pedantic grumpiness rises to the surface.

If you look at the Wikipedia article https://en.wikipedia.org/wiki/Stern%E2%80%93Gerlach_experiment  there's a very neat diagram at the bottom ("sequential experiments") of a triple S-G assembly.

You might argue that the first unit is doing some sort of measurement, though it is what I called "segregation" rather than measurement. But the second one is merely rotating the spin axis of the incident beam: there is no element of counting or segregating anything and the output is identical to the input, just rotated.

I think a lot of confusion arises from an elementary "explanation" of indeterminacy, where people talk about measuring the speed and position of a particle by bouncing a photon off it, and argue that the transfer of momentum is responsible for the "uncertainty" of the result. That wasn't what Heisenberg said at all!

[varsigma (OP)]

About the other discussion

It seems to me the application of an entity/attribute model to physics theories or experiments isn't optimal.
Instead of for example, arguing about whether energy is a real physical thing, whether you can put it in a bottle or whether it's an entity or an attribute, just say it's part of physics, it has physical units, it's physical.

The concept of work is completely ingrained into our existence, we have to do work all the time in a gravitational field. Can we put gravity in a bottle? Does that question mean anything?

What does the bottle have to be? An entity? What attributes should it have? Isn't gravity an attribute of the universe, so already "in a bottle"?
Or complex quantum probability amplitudes, what kind of bottle can you put them in?

[alancalverd]

If you can't put it in a bottle, it isn't an entity.

Gravity is a phenomenon with an attribute - it sucks. You might be tempted to call it an attribute of mass, since it is always associated with mass, but dark matter seems to confound that statement!

Work and energy are quantities.

A quantum probability amplitude is a mathematical model.

[Eternal Student]

Hi.

It occurred to me whilst driving home yesterday but you beat me to publication!

    Since it's a friendly forum, it's not as if it matters a lot.  I only made the comment since there may have been 1 other person reading it and getting confused for a day.

Your option c is the only one that makes sense and describes the "black box" transfer function of multiple polarisers,...

    Yes absolutely.   A "black box" approach is all I would have tried for if I had continued developing the argument for how a polarising filter works when a photon approaches it.   I wasn't entirely sure exactly how I was going to finish that development and make plausible arguments for polarisation being thought of more as a blend of answers "yes" and "no" for whether a photon would pass a filter of a given orientation etc.  I was just fairly sure that going straight for a mathematical representation or any argument that polarisation is spin would not be useful for many readers.   (I don't know why I worry about that too much, even on a busy day there will be only 2 readers plus a moderator or two who felt obliged to read it).

It seems to me the application of an entity/attribute model to physics theories or experiments isn't optimal.

    Which seems like a very reasonable statement to me.   It's quite different to what was suggested in some of your earlier posts.  In much of QM it is very unclear how attributes behave and should be associated with an entity.
      However, I'm not claiming to have a better description of what physics is or how you should think about it.  Physics is an attempt to explain or model what we (human beings) observe in our surroundings and it can be done in various ways.  If it works and is useful, it's OK and you can call it physics.

Best Wishes.

[Eternal Student]

Hi again.

As for the photographic use of polarisers, it seems like a complicated way of reducing intensity, compared with simply reducing the aperture of the lens,....

       Aperture may be limited in adjustment (1,2,3,4,  units etc.),  while polarisers aren't - you can spin them to any angle.
       Setting the aperture as wide as possible will help to make sure only what you have manually focused on will be in focus,  stuff in the foreground and background gets some natural blur.   This allows some artistry in photography.    If you had to cut the brightness down by reducing the aperture then everyting ends up in focus.

Best Wishes.

[varsigma (OP)]

If you can't put it in a bottle, it isn't an entity

Well, again, what kind of bottle?
What kind of material is the bottle made of?
And, is the universe an entity? If it is, what kind of bottle does it go in?

[Bored chemist]

Excellent sunglasses for coarse fishing and gliding, but dangerous for driving!

Why?

[alancalverd]

And, is the universe an entity? If it is, what kind of bottle does it go in?

A really big one. Or, if you are of a mathematical turn of mind, a Klein bottle, which resolves the dispute between Big Bang and Continuous Creation.

[alancalverd]

Excellent sunglasses for coarse fishing and gliding, but dangerous for driving!

Why?

1. You can see below the reflecting surface of still water.

2. It improves the conspicuity of cloud structures, for the same reason.

3. Car windscreens have polarising stress patterns and some instrument panels have anti-glare polarised surfaces, giving you blind spots. Glider canopies in contrast are annealed Perspex with very few stresses, and analog aircraft instruments have plain glass fronts - but beware of LCD displays on all GPS systems.

[paul cotter]

Hi ES. I worked for a short time in the video industry and all the lens assemblies we used had continually variable irises. Just for the record I read all your posts and those of Alancalverd, BC, Origin, Kryptid, Evan_au, Halc in the hope of improving my grasp of physics( i'm sure I left some contributors out!, didn't mean to ). Quick question concerning the nature of energy: work can definitely be described as a boundary phenomenon, can energy also be categorised as such? ( WAG question ).

[paul cotter]

Alancalverd, I have been using polarised sunglasses for driving for years and years and I find the glare reduction to greatly improve my discernment of hazards. The stress patterns you mention is something I commonly noticed in the past but don't seem to be an issue anymore( maybe i'm too blind to be driving!! ).

[alancalverd]

Could be that you have a posh new car with an annealed, laminated windscreen, and gentlemanly paper maps rather than that newfangled satellite thingy. The old toughened glass screens were highly stressed and some laminates didn't take kindly to being hot-glued in to the vehicle frame or being treated with heated resin  to repair minor cracks.

[alancalverd]

Quick question concerning the nature of energy: work can definitely be described as a boundary phenomenon, can energy also be categorised as such? ( WAG question ).

The conversion between energy and work involves at least a hypothetical boundary, and in most engineering embodiments there is a material boundary between the energy source and the object we want to change, so some energy is "lost" in heating  or moving the boundary. I would term that loss as a boundary phenomenon, but stick to considering work and energy as quantities, not phenomena.

[paul cotter]

As a matter of fact I drive 19year old Honda civic on which I do all maintenance work myself. This has not been an economic choice, I just like the car. Two years ago I bought a new clio hybrid for the "boss" and I find all the bells and whistles and stupid alarms infuriating.