I have edited this question for three reasons. They are: 1. A possible duplicate of another question. (I believe the "answer" to that question is incorrect.) 2. I received two contradictory answers. 3. I have new information related to the question. See below

(Start of new question) It's safe to say that the surface of a star is a violent place. Light emitted is not coherent, yet when the starlight arrives here it is coherent. So I can now answer part of my question, "Can two separate sources of light become coherent. The answer is yes. This very short video demonstrates the point https://youtu.be/4o48J4streE. The mathematics behind this is the Van Cittert-Zernike theorem. This contradicts what has be said on this site. Not sure what I can do about that. The question that remains, however, is can two light sources become coherent without traveling great distances. Also, what is a light source? Is a star a single source, or is the light source each individual electron. (End of new question.) (Start of original question) I would like your input on whether or not I can create coherent light. Here’s what I am thinking: If I have two separate white light sources a color filter could be used to select their temporal component. Then pinholes could be used to establish their spacial coherence. I could also move the light source to match the spatial coherence as well. If necessary a polarizing filter could be used to align their polarity. The two light sources would be focused and then combined through a beam splitter and sent through a single or double slit to see if an interference pattern is present. Will it work? Any suggestions? Thanks

  • $\begingroup$ Dirac once stated, that a photon interferes only with itself. However, his statement was proofed to be incorrect. See article nature.com/articles/198255a0 $\endgroup$
    – Semoi
    Jul 28, 2019 at 19:40
  • $\begingroup$ @Semoi Dirac was still right, what happens when you get interference from two different lasers is that each photon that hits the screen could have come from either laser, so it's an interference between the two possible origins where the same photon could have come from. $\endgroup$ Jul 28, 2019 at 20:54
  • $\begingroup$ Thirteen ducks - each duck creates its own wave pattern. All of the individual wave patterns gather to create coherent waves that move away from the ducks. This demonstrates that separate wave sources can and do combine. $\endgroup$
    – Lambda
    Jul 29, 2019 at 5:11
  • $\begingroup$ Spacial coherence is achieved by moving the source, like moving a mirror in an interferometer. $\endgroup$
    – Lambda
    Jul 29, 2019 at 5:18
  • $\begingroup$ Lasers are a entirely different situation. They have no room to adjust to match with another source. $\endgroup$
    – Lambda
    Jul 29, 2019 at 5:20

2 Answers 2


@PhysicsDave probably has not tried to interfere the light from two independent lasers. In fact, the two will NOT form a visible interference pattern. Their frequencies and relative phases will never have a stationary relationship (which is necessary for forming a visible interference pattern) unless the two lasers are coupled. At any instant, there will indeed be an interference pattern, but it will be continually moving BECAUSE the two lasers are not mutually coherent.

There is indeed a way to "force" two lasers to be mutually coherent. If a small amount ("sample") of the light from each laser is split out using beamsplitters and those "sample" beams are combined in an interferometer, and a fast photodetector is placed in the (rapidly and randomly moving) interference pattern, and the output of the photodetector is used to drive one of the cavity mirrors in one of the lasers, the two lasers will become "phase locked". The interference pattern formed by the sample beams will become stationary, and the main beams from the two lasers will then be mutually coherent. But if the phase-locking loop is broken, the two lasers will immediately cease to be mutually coherent.

Each photon that enters an interferometer forms an interference pattern of sorts. That is, its wavefunction is in the form of an interference pattern (though the photon when detected will be at a random place in the interference pattern). As long as the interferometer is configured in such a way that each photon's wavefunction will form essentially the same interference pattern, you will see fringes -- regardless of where the photons come from. This is the case in "white light interferometry". But that does not mean that if those photons come from different sources they are mutually coherent. They are not mutually coherent unless the sources are mutually coherent.

Two independent white light sources cannot be made mutually coherent. Two femtosecond lasers, whose output can have a broad enough spectrum to be called "white", can be phase locked to produce two phase-locked "white" light beams that are, indeed, mutually coherent.

  • $\begingroup$ What if one could produce individual photons at definite times (with an accuracy higher than a quarter cycle relative to their frequency) from two separate sources connected upward to a common oscillator, with such a regular and predictable emission process would interference appear between those two beams? $\endgroup$
    – Winston
    Jul 28, 2019 at 23:11
  • $\begingroup$ Looks like nature has a way of creating coherence out of chaos. youtu.be/4o48J4streE $\endgroup$
    – Lambda
    Jul 28, 2019 at 23:19
  • $\begingroup$ @S.McGrew - First the question is not about lasers, just light sources created with color filters and pinholes, so maybe you could expand your answer? Secondly what you call interference or photons destroying themselves is not possible, 1900s thinking carried over into 2019?. Also how do explain multi-color DSI which can easily be created from 2 coloured sources? Finally you don't know anything about me and what lasers I've worked with or how, so please keep those narrow views and judgements to yourself, it's more appropriate for an open forum website. $\endgroup$ Jul 29, 2019 at 0:22
  • $\begingroup$ I said nothing about photons destroying themselves. Your ideas about interference are distinctly outside the scope of generally accepted physics. That doesn't make them wrong, but they should be presented as opinion, not as if they were fact. If you have done the experiment I described, you would understand my point. In a two-source DSI, remove either source and you will still get interference fringes- which proves that the fringes are not due to interference between light from the $\endgroup$
    – S. McGrew
    Jul 29, 2019 at 1:24
  • $\begingroup$ two sources. Instead, the multicolor fringes are a simple superpositon of separate interference patterns, each due to a single source. Light of one color is NEVER coherent with respect to light of another color in the ordinary sense of the word "coherent". Careful reading of papers on stellar interferometry can help build an understanding of the fact that independent sources produce independent fringe patterns. $\endgroup$
    – S. McGrew
    Jul 29, 2019 at 1:39

Yes there will still be an interference pattern from the 2 combined sources. Your approach would be the same as if we had combined 2 separate interference patterns from 2 sources. The separate patterns are identical because they are created with the same type of initial source, same color filter, same pinhole, distance to screen, and slit size and spacing.

Your question tries to get at the issue of how coherent is the light from a single source verse 2 combined separate sources of similar quality. And the answer is will there is slightly less coherence to the 2 separate sources but it is not enough to be noticed in a double slit experiment.

Another way to approach this is to use 2 different colors and you will still observe the interference pattern although not as sharp (or blurred) as the 2 colours do not exactly overlap.

In your prior post about the reflected pattern another aspect of "interference" is seen. A better way to understand DSI is to employ the modern "photon wave function" method taught by Feynman, this method initially stated that the sum of all possible paths of light need to be considered, later he realized only the shortest path need be considered as long as it was integer multiples of the photon wavelength. In layman terms light will only travel in allowed paths that are integer multiples of the photon wavelength. This makes sense, for example in a laser cavity we know photons do not propagate when the cavity length is upset. The term interference is from the early 1900s and was successful because waves (like water) interfere therefore light was waves, a major discovery. But on closer study we know photons never cancel each other out, they just pass thru each other, exactly like water waves, i.e in the DSI experiment there is never destroyed photons, bright bands are where all the photons are, in dark bands there are no photons as the paths are the wrong dimension (length). In turns out the classical math for DSI based on "interference" works out pretty much the same as the Maxwell or quantum or Feynman calculation as both calculations rely heavily on wavelength and path distance!

  • $\begingroup$ Thanks for your quick response to my question. Attached is a YouTube video I did called Feynman’s Mirror. The video is not that great. I put it up one night with no edits. You may find it interesting though. youtu.be/wNRxMMABJQI $\endgroup$
    – Lambda
    Jul 28, 2019 at 21:19
  • $\begingroup$ @Lambda Yes Feynman did a lot of work trying to understand light and "interference", your video shows his work in tracking phase with path length. Eventually he concluded the shortest path was choosen. $\endgroup$ Jul 29, 2019 at 0:07
  • $\begingroup$ Coherence is a complex subject and is sometimes best handled with statistics. When we look at starlight we are sampling 100s or 1000s of photons and when we funnel them into an interferometer we may be dealing with even less. In theory they are random in phase (but of same frequency), the signal can interfere with itself of course. $\endgroup$ Jul 29, 2019 at 14:46
  • $\begingroup$ A double slit is a very coarse way of showing interference so your experiment will work but .... I do not know the extent of the coherence you generate from the 2 sources, the 2 sources are independent and certainly do not influence each other. It just becomes a mathematical/statistical exercise .... of how coherent they are. $\endgroup$ Jul 29, 2019 at 16:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.