[hamdani yusuf (OP)]

Interference is the result of superposition of two or more wavefronts producing maxima and minima with a spatial distribution related to the distribution of the sources. 

Yes. The defining characteristics of interference is the occurrence of points in space with cancelling wave, called destructive interference. Wave addition/superposition that doesn't produce "darker spots" isn't usually called interference. In my video, I mentioned color mixing, beat, and superposition of two light waves with perpendicular polarizations as some examples of superposition that doesn't produce interference.

[hamdani yusuf (OP)]

What you call a "conceptual difference" is what scientists (and parents) call "giving different names to different things".

Agreed.

[alancalverd]

I wrote that interference and diffraction are distinct phenomena, but both are effects of superposition.

Which is wrong. Superposition requires two sources, by definition. There is only one in diffraction.

[alancalverd]

In my video, I mentioned color mixing, beat, and superposition of two light waves with perpendicular polarizations as some examples of superposition that doesn't produce interference

Color mixing is due to physiology, not physics! If you illuminate a surface with two light beams of different frequencies, the eye may perceive one different color but a spectrometer will show the original frequencies only.

This is different from true frequency modulation or mixing in a nonlinear medium. 

"Beat" is true superposition. When one frequency is much higher than the other, it is called "amplitude modulation".

[alancalverd]

- Superposition derives from quantum theory, and cannot be described by classical physics.
- You don't need superposition to describe radio waves

Beg to differ. The superposition of radio waves can and does cause spatial modulation, entirely calculable from the classical addition of sinusoids. It was the basis of  Knickebeine and LORAN long-range precision navigation. There is no mention of quantum physics in my radio navigation textbooks, though they do talk about relativity in the GPS chapters!

[Eternal Student]

Hi.

   So having a look through all the replies, this is how I'm tempted to summarise the responses:

1.   We do not know exactly what the underlying process for diffraction or interference is.   There are some models.

2.   In those models, especially Huygens principle, it is quite common that diffraction and interference do appear to have similar explanations and could be due to the same underlying principles of physics.

3.  Is it important to distinguish between interference and diffraction because there is some very different underlying physics you (especially Hamdani Yusuf) are trying to present?   Answer:  No.

4.   Is it important to distinguish the difference just for improved communication and consistency between specialists?   Answer:  (You both claim) Yes.   Your main example is that aeroplane pilots and navigators can die otherwise.
   My view:  Sure, in some areas of work there will have to be some very precise use of terminology.  In more general science there is also some benefit in recognising that some apparently disparate things are actually closely related and the same underlying principles could apply.   For example,  Beta particles are not the same as high velocity electrons.  Beta particles can ONLY be produced by changes that take place in the nucleus of an atom.   However, it is much more useful just to recognise that they behave so much like high velocity electrons that they almost certainly are high velocity electrons.

 Best Wishes.

[Eternal Student]

Hi.
   Now just some basic tests to see if everyone really has thought this through:

When monochromatic light passes through a single slit and produces a pattern on a screen.  What do you call that pattern on the screen?
   (a)   A diffraction pattern.    (Note that Feynmann suggested diffraction applies only for multiple sources interacting.  However, light does spread out when passing through a slit, I mean this single slit thing is the text-book example of a diffraction process isn't it?).
   (b)   An interference pattern.   (Can you have interference with one source of light?  Yes you can and you do @alancalverd  whenever that light passes through an aperture).

When monochromatic light passes through two slits and produces a pattern.... What do you call that?
   (a)  A diffraction pattern.
   (b)  An interference pattern.

When ...light  ... passes.... diffraction grating... screen.  What do you call that?
    (a)  A diffraction pattern.   (I mean, it had the name "diffraction" in "diffraction grating" didn't it?)
    (b)  An interference pattern.    (Feynmann suggestion is that you really should be calling this diffraction by now).

   I think common usage of the terms would have the answers   a (for 1 slit),  b (for 2)  then back to a (for diffraction grating).  @hamdani yusuf  asked about school pupils and what they are taught in schools.  I've just spent 20 minutes going through the AQA syllabus for physics at "A" level.  They do describe the patterns with these terms.

   Meanwhile the Feynmann suggestion should give   b, b, a.

    If I've got the gist of what @hamdani yusuf  wants the definitions of diffraction and interference to be.   (One is about bending light, the other includes producing bright and dark bands in a pattern)  then the answers   are  b, b, b.

    My view is... what the figgy pudding does it matter?

Best Wishes.

[alancalverd]

2.   In those models, especially Huygens principle, it is quite common that diffraction and interference do appear to have similar explanations and could be due to the same underlying principles of physics.

3.  Is it important to distinguish between interference and diffraction because there is some very different underlying physics you (especially Hamdani Yusuf) are trying to present?   Answer:  No.

I think not.

Whilst the Huygens wavelet model can be used to predict the downstream waveform for both diffraction and interference, you can't ignore the fact that diffraction requires only a single source and interference requires at least two sources.

Unless 1 ≥ 2, they cannot have the same explanation or be based on the same physics!

There is an additional problem with the wavelet model, to which HY's reference alluded: it predicts an equal and opposite wavefront travelling back towards the source(s), which is not observed. It is a convenient geometric trick but it ain't physics!

And sadly, beta particles are indeed high speed electrons, just as gamma and x-rays are identical. They are named differently on account of their different sources, and again the distinction has safety implications: you can switch off an electron or x-ray source, but not a beta or gamma source!  (Superpedants may object to my ignoring the betatron, but you won't get many bikes out of a cyclotron - sometimes we just have to live with the legacy). 

[alancalverd]

When monochromatic light passes through a single slit and produces a pattern on a screen.  What do you call that pattern on the screen?

If the slit is wide, most of what you see is the unobstructed primary beam with some interference where the edge-diffracted beam has superimposed on the primary beam, and some diffraction outside the slit width.

As the slit gets narrower the diffraction contribution begins to dominate until most of what you see is the interference pattern of both edge-diffracted beams.

Wikipedia discusses this in detail under "Airy disc".

What comes out of an ideal diffraction grating is a diffraction pattern, but most of what passes through most real gratings is primary beam, so you get a convolution of diffraction and interference, like the single slit but with a larger element of  diffraction.  The mathematics of x-ray diffraction from powdered solids and stretched fibers is fascinating and ends up with the structure of DNA, and hemoglobin. I have a delightfully disgusting story about x-ray diffraction which can wait for another day.

[Eternal Student]

Hi.

you can't ignore the fact that diffraction requires only a single source and interference requires at least two sources.

    I'm fairly sure you started that reply before seeing the other post....

Can you have interference with one source of light?  Yes you can and you do @alancalverd  whenever that light passes through an aperture

Here's your later reply:

If the slit is wide, most of what you see is the unobstructed primary beam with some interference...

   So you fundamentally admit and you are aware that there can be interference when there was only one source of light.    You then did some carefull tip-toeing around the issue.   There were not two or three (or more) beams that interacted.   There was only the one beam of a certain width, it all came from just one source of light.   There is some diffraction at the aperture and there is then some interference effect that gives rise to an intensity on the screen that is the usual  Sinc function.
   (You also mentioned "airy disc", which as I'm sure you know you'll only get for a 2 dimensional aperture, a round hole instead of a slit.  I'm deeply suspicious that's an attempt to sidetrack the issue.   The key issue is that ONE source can still give you an interference effect.  You do get interference from just ONE beam of coherent light whenever that beam goes through an aperture).

Here's the more geometric interpretation using rays of light, which I'm sure you've seen before:


[image from Wikimedia]

    Looking at this representation, the light rays are bent in various directions and there is a path length difference to where they converge on the screen, giving a phase difference and hence a constructive / destructive interference effect.    I'm sure you know this.  Anyway, as far as I'm concerned, you can legitimately call the single slit "diffraction pattern" you see on the screen an "interference pattern" because it is due an interference effect.

    I suppose if you wish to be pedantic, for a classical model of light, there NEVER is just one source of light.   There is always a beam of some width which we can divide and treat as separate sources or rays.   In the diagram above the beam was usefully divided into two halves.

   Now even if you slim the beam of light down as much as possible, so that you actually only had one photon,  then you are in the territory of Quantum mechanics and the usual single and double slit experiments.   You still have both the effects of diffraction (bending of a path away from a straight line) and also interference observable.   In this situation the interference is really an interference in the wave function ψ(x,t) and only becomes noticeable when the experiment is run several times and the usual interference pattern is built up on the screen.  (I'm sure you know all of that and there's no need to fill this thread with the details).   The key point is that even with just ONE photon being sent through an aperture, interference of some kind is still exhibited.    Your assertion that interference requires two or more sources is not something that I will accept without a (non physical, purely academic) fight.

Best Wishes.

[alancalverd]

So you fundamentally admit and you are aware that there can be interference when there was only one source of light.   

Oh dear, we are in danger of arguing about the meaning of "one" and "source"! In my analysis, and indeed in the Huygens model, an edge becomes a source.

Fortunately for the artist, I haven't seen this diagram previously. Look at the three rays entering from the left. The upper one turns left. The lower one, apparently diffracted from an antisymmetric edge, also turns left- why?  But most mysterious of all, the one in the middle also turns left for no reason at all! Doesn't light travel through space in straight lines any more?

The single-edge diffraction pattern is the effect of a semi-infinite aperture. In my universe a light ray passing more than a millimeter or so from the edge has no reason to deflect - a millimeter is "infinite" at optical wavelengths! But a ray grazing the edge will be diffracted, so the secondary wavefront can interfere with the primary ray  in the geometrically illuminated region, but the primary ray does not interfere with the diffracted ray in the geometric shadow region.

The Airy disc is of practical importance because most useful lenses are circular and we are usually concerned with the resolution of 2D sources, but the mathematics is no more than a function of the wavelength of the light divided by the aperture diameter. The geometry is irrelevant: it just happens that circular symmetry obviously gives you a circularly symmetric pattern of fringes.

 

[alancalverd]

I must take exception to your interpretation of the double-slit experiment.

If you count single photons arriving at the target, they turn up as "whole photons" with the same energy as the source, separated in time and with their spatial distribution determined by the probability function, or indeed a Huygens superposition function. But interference by definition requires the simultaneous presence of two entities, A interfering with B, or X superposed on Y. In the single-photon experiment A and B are never present at the same time so they can't interfere.

Let us suppose for a moment that a photon interfered with itself. Constructive interference would produce....a photon with twice the energy?  Not observed - we start with a red laser and end up with a red pattern.Two photons with the same energy? Not observed - one in, one out. Pity, that would give us free energy for ever! Destructive interference - where does the energy go?  You can't cheat by integrating over time because I am using a really, really weak source and you don't know when the next photon will arrive, if ever.

So whilst an interference model describes the result, it doesn't explain or represent the actual events that lead to it.

[hamdani yusuf (OP)]

I wrote that interference and diffraction are distinct phenomena, but both are effects of superposition.

Which is wrong. Superposition requires two sources, by definition. There is only one in diffraction.

Which is wrong. Superposition requires at least two sources, which can be more.
How many oscillating electrically charged particles in a diffracting obstacle?

[hamdani yusuf (OP)]

In my video, I mentioned color mixing, beat, and superposition of two light waves with perpendicular polarizations as some examples of superposition that doesn't produce interference

Color mixing is due to physiology, not physics! If you illuminate a surface with two light beams of different frequencies, the eye may perceive one different color but a spectrometer will show the original frequencies only.

This is different from true frequency modulation or mixing in a nonlinear medium. 

"Beat" is true superposition. When one frequency is much higher than the other, it is called "amplitude modulation".

A broadband light detector, such as LDR, detects combined light intensity. It has nothing to do with physiology.

Beat is a result of wave addition, while amplitude modulation is a result of wave multiplication.

[hamdani yusuf (OP)]

1.   We do not know exactly what the underlying process for diffraction or interference is.   There are some models.

The fact that there is a disagreement doesn't necessarily mean that no one knows.
FYI, there's still disagreement about the shape of the earth, or how its rotational axis precesses.

[hamdani yusuf (OP)]

2.   In those models, especially Huygens principle, it is quite common that diffraction and interference do appear to have similar explanations and could be due to the same underlying principles of physics.

Demonstration of electromagnetic phenomena often presents analogy from mechanical phenomena, because they are easier to observe with simpler tools. Wave diffraction and interference can be convincingly shown using surface water wave. But many mechanical analogies to explain electromagnetic wave are known to be misleading, such as rope and fence analogy to demonstrate polarization or a car moving from concrete to sand to demonstrate refraction.
The best way to learn about the difference between diffraction and interference is by studying both phenomena independently through experiments that only involve one of them and not the other.

Regarding Huygen's principle,

If you think that single slit interference pattern is produced by the space between edges of the slit which act as wave sources, you’ll be in trouble explaining the same pattern produced by a thin wire.

[hamdani yusuf (OP)]

3.  Is it important to distinguish between interference and diffraction because there is some very different underlying physics you (especially Hamdani Yusuf) are trying to present?   Answer:  No.

Yes, as I and Alan explained earlier. They are two distinct phenomena. Just because some experiments involve both of them, it doesn't mean that they are the same thing and inseparable.

[hamdani yusuf (OP)]

My view:  Sure, in some areas of work there will have to be some very precise use of terminology.  In more general science there is also some benefit in recognising that some apparently disparate things are actually closely related and the same underlying principles could apply.   For example,  Beta particles are not the same as high velocity electrons.  Beta particles can ONLY be produced by changes that take place in the nucleus of an atom.   However, it is much more useful just to recognise that they behave so much like high velocity electrons that they almost certainly are high velocity electrons.

They were originally called Becquerel's rays.
Here's some history.

[alancalverd]

How many oscillating electrically charged particles in a diffracting obstacle?

According to Huygens, none. According to your microwave experiments, lots and lots of free electrons or none at all.

[hamdani yusuf (OP)]

How many oscillating electrically charged particles in a diffracting obstacle?

According to Huygens, none. According to your microwave experiments, lots and lots of free electrons or none at all.

Obviously, not just one.