The overly clever setup of the beam splitters, detectors and the clock reminds me of how designers of perpetual motion machines try to hide the energetic symmetry of their devices behind complex mechanics. The end result is the same: you just can't cheat your way around conservation laws and symmetries.
@@-danR 🤦♂🤦♂🤦♂ that's why I said "read my comment", because I meant MY ACTUAL EXPLANATION COMMENT ON THIS VIDEO 🤦♂🤦♂🤦♂ why would anyone think that "false, read my comment" 4 literal words would be the explanation of anything? I... I'm legit debufunged
@@raulkaap yes but not because of "cheat your way around conservation laws and symmetries", there are not "conservation laws" nor "symmetries", is just that if you measure it, it is the same, and it kinda makes sense because it would be a bit weird if it changed, but I don't necessarily see any clear reason why it couldn't, because there are no such "laws". but at the same time, your comment seemed to imply that there is no way to measure it, and in my comment I explain how there is, in fact, that is my point, that if you measure it, it is the same, because you can measure it
Somewhere, probably on PBS Spacetime, I learned that if I relabel the 'speed of light' as 'speed of causality', and light just going 'full speed', that makes the intuition slightly better. After a few years of mulling it over, I agree.
your videos have repeatedly inspired me to re-engage with scientific learning. you make scientific understanding feel within reach of anyone, not just labs with big grants. also you are precise, but also humble. thanks for publishing your videos.
learning is what makes it unique, whereas any of the many edutainment-type science channels can ultimately only offer a form of teaching or lecturing, the distinct ability this channel has is to just offer learning instead
This is a great a great video! I want to point out that you missed a simple (yet arguably not obvious) way to make opposing anisotropy locally: make the spring-mass setup MOVE! Or equivalently, make the wave generator move, so that it excites a different mass at each time step. From the resting frame of the wave generator the wave propagation is then opposing anisotropic. You might say that this makes no sense in the setting of the EM field because the field cannot have absolute motion. However, this is exactly the ether interpretation of the EM field (and spacetime itself): opposing anisotropy in the speed of light can be explained by assuming an absolute rest frame (the rest frame of the ether aka the rest frame of the CMB) and anisotropy resulting from motion relative to it.
But while this possibility doesn't suffer from the energy conservation problem, the aether has other problems, and all the old aether theories have been disproven, e.g. by the movement of the earth trough the aether. My takeaway is the following: The current theories describe most of our universe extremely well, so we often try to make small changes only. This video convincingly shows that an anisotropic one-way speed of light is not possible by making only small changes to our theories. But your comment hints that we can never consider every possibility. That is, unless we employ a much more systematic approach to the assumptions underlying our modern understanding of the universe. There are a lot of interesting ideas out there, but nowhere near enough manpower and funding to check every single one with all the consequences individually.
@@entcraft44the key thing here is that the *old* aether theories have been disproven. There is a way to have an aether field in such a way where you get the exact same equations that you get from relativity, where the speed of light is constant in the reference frame of the aether, and objects moving relative to the aether get length contracted and time dilated. This formulation gives identical math as special relativity so it is not possible to distinguish which interpretation is closer to reality through measurements
@@entcraft44 Aether has been disproved but an absolute rest frame has not. In big part because of the two-way speed of light. Once you assume that speed of light is the speed of transfer of information from one segment of space to another then you can have both the relativistic effects as well as an absolute reference frame, where the constant speed of light in any reference frame is just a result of slower processing of the thing moving close to the speed of light. (ie: imaging you have a clock that measures the passage of time by sending a short ray of light towards a mirror and waiting for a response, that clock would indeed slow down the closed you go to the speed of light, and stop completely once you reach the speed of light, since light would have to move perpendicularly to the mirror, the exact same thing happens when one particle sends our information and waits for it to come back) All the other space warping effects are also accounted for as being just a mirage caused by difference in processing speeds of information( photons in this case) that arrives perpendicular or and paralel to the motion as well as what just happens when light travels at a fixed speed.. First from the object's PoV the entire universe shrinks along direction of motion because its own processing slows (so everything just starts becoming very bright, and very blueshifted) Additionally the faster the object moves the more perpendicular to its movement the light has to be to actually hit it. I did the math and some simulations on this and it all checks out, tho I'm still unsure as to where the extra mass is coming from when you accelerate things close to the speed of light.
The best explanation of 0ermittivity and permeability ever, and the analogy with mechanics made so clear the concepts of isotopi and anisotropy of the material... literally wow, You deserve an honorary Professorship
@@farhanrejwan surprinsingly, many ppl still enjoy oldtimes TV bc they don't need to choose something, just the channel and watch watever they give youl; I have a tiny amt of doubts and also no evidence, but I believe this could be cause by mental exhaustion,(which can come from anywhere), at least for ppl who do it for long periods of time......... also could be they don't want to bother looking for a youtube channel or whattever that plays random shit, like simpsons episodes or so,, due to ignorance or also lack of mental energy, there is a possible third scenario, which involves a third party showing the subject how to look for and/or setup such tvlike experience on phone, pc or whatever, but the subject will refuse, possibly due to lack of mental energy.... yea, i know, it's the root of all problems, even mine..... what am i even doing here, all coked up, writing pretentious nonsense..... ah.....
i think people are too afraid of "unsolvable" problems like this. if our current model does not predict any effects from this hypothetical and we do not observe any effects to prove our model wrong in that regard, it truly does not matter what the answer is. its like trying to prove the existence of an unobservable god with science. i appreciate the insight on how a closely related problem is much more approachable
@@YuriyKrivosheyev not quite. russel's teapot places some arbitrary "burden of proof" on the "positive claims". in reality, neither claim inherently has more merit than the other. the proper way to look at it is that whether this teapot "exists" or "doesnt exist" cannot mean anything unless it has observable effects (which it does not). embrace the fact that there is no useful answer rather than arguing one way and demanding proof from the one arguing the other way. otherwise its like trying to prove a god *doesnt* exist with science.
Yes, but… not really? Science has its dialectic in pure idealism, the imaginary, intuitive and creative. These “brain functions” are our pathway to pure empirical knowledge - ontology and epistemology, subject and object. More culturally, we depict someone getting a bright idea (lightbulb moment in cartoons), and then the “labour” to prove or disprove, calculate vs experiment, theory vs observation. As this video also shows, Einstein knew his Maxwell, and “imagined” how time, space, mass and energy interacted - and, well it still works, empiricism backwards and forwards to now. Einstein also put a nail in time travel, so the only way forwards is - unknown and undefined. Perfect for those “unscientific” traits we have, besides logic or reasoning. This is also how “wrong ideas” emerge, but are not empirically sound if they can be disproved. There are always “ideas” in a sort of superposition - before we are able to both prove and disprove. Both are necessary. If I’m wrong in this position, we have invented or discovered “everything”, there is no forward empirical path - nothing to imagine into the field. It’s a gloomy world at this time, but not _that_ gloomy?
@@jotch_7627Ultimately you can’t prove about a generic god with science, but many religions make enough additional specific claims about their gods that the gods they describe can be assessed
once again, what a brilliant video. one way speed of light is such a complex thing to wrap my head around, when I'm thinking about it i always get lightheaded.
exactly, that is how they do it, they try to " complex thing to wrap my head around" their way so that you don't try to put yourself on it, to try to understand it, to fight it, to bother with it, and to just accept what they say, they scare you, they make you think you just can't understand it, make you feel dumb, fight that, you are not dumb, put them in their place
@@NeroDefogger they don't make me feel dumb. they're making me smarter by making my brain work until topic makes sense. whatever you are saying is not the way to learn anything. my comment had an intended pun.
@@tsraikage "is such a complex thing to wrap my head around" and "when I'm thinking about it i always get lightheaded" does not "make me feel dumb"? whatever you say... "they're making me smarter" no they are not. "until topic makes sense" the topics never make sense. "whatever you are saying" I'm pretty clear on what I'm saying, did you not understand what I said yet you claim to understand the topic of the video? "is not the way to learn anything" what? you really did not understand anything I said... "had an intended pun" ... what?
Well done. It's one thing to argue that we cannot measure the speed of light in a single direction (always requiring a reflection), but quite another to assert that the speed of light in one direction could be different from the speed of light in the other. While the E=cp explanation is much simpler, I very much enjoyed the full exploration of the possibilities of anisotropy. Recall that it was Maxwell's equations that eventually led to special relativity, where only the Lorentz transformation could account for a constant speed of light. So if Maxwell's equations are fundamentally true (omitting quantum perturbations), then even if you can postulate the speed of light being different in opposing directions, it's impossible to model it in a self-consistent way.
If we can't measure the one-way speed, does it even make sense to consider to question what the one-way speed really is? In SR one-way speed of light is just a choice of coordinates. The problem with the argument from Maxwells equation is that in their standard form you have already chosen an isotropic, Einstein synchronized coordinate system. If you choose another coordinate system, you will have equations with anistropic light speeds (and hence not the standard wavequation with c = 1/√(ε₀μ₀))
_"even if you can postulate the speed of light being different in opposing directions, it's impossible to model it in a self-consistent way."_ - it is actually the other way around. Any anisotropic model of the light speed is self-consistent, as along as the two-way speed is c. The one-way speed is theoretically impossible to determine and is a matter of definition. You can choose any one-way speed you like for your model, and you will still end up with the current consistent physics model.
You can use curved space instead of a mirror. Shoot a photon to orbit a black hole near the event horizon. It will come back to you. Just time the orbital period. You know the distance so you will know the one way speed.
@@jeffbguarino And how would you prove the act of going around the black hole didn't slow down or speed up the photon in some way? And let me be clear, it doesn't. I feel very confident that c is constant in all directions but the point is, how do you prove it?
@@wingracer1614 But the circle of space at the event horizon is all the same direction. This is how a direction is defined by light which takes the geodesic which is the shortest path. So it is one direction or the same direction or do you somehow superimpose a euclidean coordinate system on top? It is just as if the earth is at two different places in an inertial euclidean coordinate space and you are at two places at the same time. If the speed of light slows down or speeds up at different places going around a black hole then it could do this anywhere else. Like between mars and the earth speeding up and slowing down all over the place and you would never know. So that is introducing a new animal into the problem. I found the definition of a straight line on google: (which I never really knew before) The shape of spacetime requires a new definition of a "straight line". The way to define a straight line is that trajectory that is followed where object experience weightlessness. Otherwise, if you experience weight, then you must be accelerating (according to the equivalence principle) and, hence, you must be changing your direction through spacetime. Since the space station feels weightlessness, then it is travelling in the straightest possible path through spacetime. Similarly the Earth is going straight as it orbits around the Sun. We standing here on the Earth are, however, accelerating. Thus when you are at the event horizon or just above it, and fire a laser tangentially , the light will be in free fall but here the escape velocity is the speed of light, so the light can avoid falling into the black hole.
I thought about this some time and realized that no matter what I try, I always end up with a two-way speed of light in the end. And I had some really crazy ideas that would be practically impossible to do even if we had access to materials that literally have unrealistic properties.
It's actually quite easy, technically you just have to take advantage of the non-linear nature of time dilatation. Basically you take some atomic clocks, move them rapidly in two different directions and graph the change in time. The time dilation on each clock should fit the curve t₀/√(1-v²/c²) where t₀ is the time for the 'rest' reference. If they fit to the same curve for each direction, then the one way speed of light is the same. The other way is that you just use a particle accelerator to spin a proton or whatever at like 99.9999% C. If the speed of light were different then the particle would have to slow down on one side of the ring in order to maintain it's mass. (or it's mass would fluctuate on each side) in either case you'd need to adjust the timing or strength of the magnets to maintain the particle in the ring - in other words every time the large hadron collider is use the one-way speed of light is verified to be equal.
@@takanara7 _"If they fit to the same curve for each direction, then the one way speed of light is the same. "_ - they will always fit the same curve for each direction, independent of what the one-way speed of light really is. _"in either case you'd need to adjust the timing or strength of the magnets"_ - you wouldn't need to, because the timing and current in the electromagnets would be subject to the same isotropic speed of the light. The one-way speed of light is theoretically impossible to measure. Trying to come up with mechanisms that could measure it, is like coming up with mechanisms for a perpetuum mobile. It is just a matter of determining why it fails.
@@rogerphelps9939 :"Romer's experiment!"_ - That's a difficult one. The mathematics for why Rømer's experiment is not a one-way speed of light measurement is described by L. Karlov in “Does Roemer's method yield a unidirectional speed of light?” Australian Journal of Physics 23, 243-258 (1970) I have a hard time interpreting what the correct interpretation of the math is, but I think it may be the following: Rømer's experiment measured the time between eclipses of Jupiter's moon Io. When the Earth in its orbit is moving away from Jupiter, there should be more time between those eclipses than when the Earth is moving towards Jupiter. Assuming the speed of the Earth wrt. Jupiter is the same in both case (but in opposite direction), you can calculate the (one-way) speed of the light from that difference. But how do we know that those speeds are the same? We measure that by measuring angles towards the Sun. Those measurements are based on the lines of sight to the Sun and Jupiter, that is, based on the (one-way) speed of light from the Sun to the Earth. Due to that, with an anisotropic speed of light, the Earth's speed would appear to be larger on one side of the Earth's orbit than on the other. That is, the orbit of Earth would appear to be more elliptical. The difference in travel time of the light between both side of the Earth's orbit would be attributed to that difference, rather than to the lower/higher speed of light.
@@rogerphelps9939 That is a difficult one. The mathematics for why Rømer's experiment is not a one-way speed of light measurement is described by L. Karlov, in “Does Roemer's method yield a unidirectional speed of light?” Australian Journal of Physics 23, 243-258 (1970) I have a hard time interpreting the math, but I think it may be the following. Rømer's experiment measures the apparent time between eclipses of the moon Io of Jupiter. It then compares those times at a part of the orbit of Earth when it is moving away from Jupiter, with the part where Earth is moving towards Jupiter. Effectively, it is a comparison of the Doppler shift in two directions. But the calculations assume the speed of the Earth wrt Jupiter is the same for both directions. The result depends on that, but that is merely a convention. To measure the "real" speed of the Earth, you would need to measure the distance Earth travels between two points in time. But how much you measure for that speed, depends on the one-way speed of light again. That is, you are measuring the one-way speed of light in relation to the one-way speed of light.
wow i’m so glad i didn’t click away after the first half lmao first half was amazing (for the viewer that’s not familiar w the conclusion) just excellent motivational context-building but the second half was an amazing next layer deeper of thought experiment / pre-analysis
Now I am imagining that little green fempto-pulsed laser blip surfing a gravity wave and I am very happy with that. Thank you for the spring mass analogy. Love when EE concepts match up pretty well to mechanical concepts. Gives more intuition about the more abstract concepts which can be powerful. Love your videos by the way. Optics was one of my favorite courses in college and these videos solidify fundamental teaching while at the same time not shying away from more advanced concepts and topics. At the end of our course, our professor talked some about non-linear optics but treated it as dark magic. Would be awesome to explore some experiments or topics in that realm if your equipment would allow.
The equivalence principal tells us if we do these experiments while in freefall we will always get the same results. This is even true if the experiment is done in free fall towards the event horizon of a black hole (one large enough to ignore tidal effects). However, observers in different reference frames may see things very differently. Light will always move at the same velocity, but clocks in different reference frames won't agree. Since the velocity is constant, the wavelength will appear different.
There's an experiment that came to me today, just before this video came out, that I would love to see, but don't have the means to perform. I just recently learned that Neodymium-doped Lithium Niobate can be used as a lasing medium. I am incredibly interested in the interference and diffraction patterns produced by it, specifically when it is being used as a laser medium while also piezoelectrically oscillating. Lazing Lithium Niobate, even when not doped, is a parametric down conversion process too, producing more than typical entangled photon pairs! If you're interested at all I can link you to a few papers, and a source for Neodymium-doped Lithium Niobate on substrate. Love you so much!!
Loved the explanation in this one, so eloquently put forth. Thank you so much for making these video's, always look forward to them! Makes me wanna go back to school and study light.
The moment you said a spring with a different constant on one side than the other, I instantly realized as it vibrated, it would grow in vibration in a single direction and I was like: “Free energy machine!!!” 😂
Brilliant reasoning and video - did you consider publishing it Physical Review Letters D (I think). The left right anisotropy also reminds me of Huygen's assumptions for wavelet sources in diffraction gratings. The whole question seems to be connected with the idea of relative separation in space-time and the presence of an impedance tensor. If that were so Classical Dynamics too would be in big trouble. Very fine work and thank you.
@Huygens Optics. Please consider the following and critique. I believe it would be a way to measure the 1-way speed of light. Let a laser light source send a pulse horizontally from S to a mirror M. Between these points place a beam splitter B close to the source. Anywhere along a line below and parallel to the line S-M, place a receiver/Clock RC. For convenience, the first example assumes RC is positioned directly perpendicular to the midpoint on S-M, such that the distances B-RC and M-RC are equal. (Variations on this experiment would place RC perpendicular to points B and M to eliminate “x” or “-x” components of the down-beam velocity, and to compare results between experiments.) A pulse of light from S would partially deflect to RC from B, starting the clock. The remainder pulse from the same beam would reflect off of the mirror M and traverse to RC, stopping the clock. Since the delay from either B to RC or M to RC could be made equal (centered) or related by trigonometrically by the geometry of the triangles formed by the paths, the delay between pulses at RC would be entirely due to the distance between B & M divided by the 1-way velocity of the beam, right? If as you say in your video, the round trip (the C' paths) cannot be eliminated, I would ask it replacing them with a known transmission velocity wouldn't do the trick. For example, the speed of electric signal propagation in a wire is not the speed of light, and if the materials are consistent, the speed of the signal should also be. Alternatively, the speed of sound does not diminish with energy loss, so why not use sound waves for the paths from splitters to receiver? In this way there is only one light path, and only in one direction.
If in a Frame of Reference there is a shift in time as a function of shift in space, then light would effectively travel at different speeds depending on the direction it travels. This is exactly what happens in the equation t’ = gamma(t-vx/cc). In a moving Frame of Reference, time is shifted proportional to distance. It’s the x in the special relativity time dilation equation. Therefore the speed of light is different in the 2 directions depending on the Frame of Reference.
@@lmmortalZoddno. From a moving Frame of Reference perspective, the distance is identically contracted in both directions of the round trip journey, and time is similarly dilated. But in one direction time is also shifted forward proportional to distance and then shifted backwards in the opposite direction proportional to distance. This means in effect from the moving Frame of Reference, light seems to travel at a different speeds in the two directions to compensate for this shifting of time as a function of distance.
Light can have different speeds in two opposite directions with respect to a moving frame of reference. The law of conservation of energy is still preserved.
*Advanced Tinkering* channel just mentioned you in an absolutely lovely way in the video "Creating Ultra-Fine Details in Titanium - 20 Micron Resolution" ! 🤗
Good video. You compare the situation with an elastic medium that has different spring constants for the two directions, which leads to violation of conservation laws. But that is not the only way to achieve anisotropic propagation speed. The medium itself could be isotropic, but simply have a velocity in one direction. That would not lead to violation of conservation laws. I think the more profound observation is that there is a delicate link between the one-way speed of light, and how simultaneity of distant events is defined. The definition of simultaneity makes it theoretically impossible to measure the one way speed of light. You might want to do another video on that. Ever since I discovered this myself, I have marvelled at the following simple sentence in the introduction of Einstein's 1905 SR paper: _"We have not defined a common “time” for A and B, for the latter cannot be defined at all unless we establish _*_by definition_*_ that the “time” required by light to travel from A to B equals the “time” it requires to travel from B to A"_ Einstein already knew that what you have discovered as well: that the one-way speed of light is a matter of definition, not of measurement.
It seems like this is just the simultaneity paradox in disguise. Alice and Bob meet up, get two clocks, and synchronize them. Alice moves a distance D left, Bob moves a distance D right, and they both stand still (relative to each other). Every second on the second, Alice flashes a light. Bob sees these flashes happening periodically, with a 1-second interval between each, and thus infers Alice experiences the same rate of time as he does. Moreover, Bob sees these flashes come slightly *after* each tick of his clock, with the delay given by dt=2D/c. When Bob also flashes a light every second on the second, Alice observes the exact same period of 1 second, and the exact same delay of 2D/c. If the clocks are still synchronized after having been moved, then the one-way speed of light necessarily equals the two-way speed. The only way for the one-way speed to differ is if the clocks are *not* synchronized. But, there is already a reference frame in which the clocks are not synchronized! Consider Charlie, who is on a rocket ship moving 0.5c to the right, relative to Alice and Bob. Charlie sees light take longer to get from Bob to Alice, than it does going the other direction, simply because both Bob and Alice are moving fast left from Charlie's perspective. Relatedly, Charlie *also* sees Bob's clock being ahead of Alice's clock, i.e. the clocks are desynchronized to account for the differing times required to cross the distance in either direction. The definition of "simultaneous events" is entirely dependent on your reference frame; simultaneity only makes sense for events at the same location. When events are at different locations, "simultaneous" depends on the observer. Measuring the one-way speed requires some notion of simultaneity across distances, which makes it inherently ambiguous. Even if the one-way speed were different than the two-way speed, then there would be some speed at which Charlie could move that results in the clocks being synchronized in his reference frame. We can simply redefine "simultaneous events" to be events which are simultaneous from Charlie's reference frame, which results in the one-way speed equaling the two way speed. This is a simple change of coordinates. The "one-way speed" is not unknowable in the same sense as Russel's Teapot. Rather, it's unknowable in the same sense that "which way's left?" is unknowable. It's an entirely semantic paradox, a simple matter of definition and choosing a coordinate scheme. It doesn't even qualify as a mystery.
This is a random thought that hit me while watching: 1) We sync up two clocks very close to each other so we know the speed of light difference will have an extremely minimal effect 2) Both clocks are taken away from each other to some meaningful distance that will allow us to measure a difference in the speed of light 3) Based on a predetermined time in the future both clocks will cause a flash of light at the same instant 4) Both sides of the setup record the exact moment based on these clocks when they witness the light from the other end of the setup getting to them The problem this solves is the relative nature of light which we are able to subvert due to us taking a higher perspective. We are looking at a 'global' time for this context & creating two independent light sources & measurements based on this 'global' time. This allows us to bypass the issue of relativity completely & get a measurement of the speed of light in one direction.
@grayishcolors The problem is that the two synchronized clocks experience time dilation when you move them, even if you move them only slowly. And if the speed of light really is anisotropic they would experience different amounts of time dilation even if you move them at the exact same speed over the exact same distance in opposite directions. Edit: And there would be no way to measure the difference in time dilation unless you bring the clocks back together again, at which point you're back to having done a two-way measurement rather than a one-way.
@@blahfasel2000 I was aware of the time dilation aspect, but the thought was that it has almost no effect unless we are talking about really fast speeds. All we care about in this hypothetically is to prove that light moves at different speeds, not necessarily have a perfect measurement. I suppose you could have a perfect measurement like you said if you brought the clocks back together. I still don't see how that breaks the results though. If the time dilation is different in direction then the measurements will also see the difference which would add to the results. Perhaps there's something I am not seeing as an issue here?
To add to this: We have a CLOCK A & CLOCK B Let us assume that time dilation somehow made CLOCK A run a whole 10 seconds slower by the time it gets to position for the experiment. Light always goes at C from all reference frames meaning the time the light takes to travel won't be affected be affected. So the light will be released by CLOCK B 10 seconds faster than CLOCK A & arrive at CLOCK A 10 seconds sooner than it will create its own light. This is of course assuming light moves at a constant speed & not two separate speeds. The only way for this not to be the case is if the speed of light moves at a different speed as to exactly cancel out the time dilation. That doesn't make sense, though, because time dilation is relative & based on speed whereas the speed of light is always C from all perspectives.
@@grayishcolors "time dilation is relative & based on speed" this is true only as long as you look at uniformly moving clocks. As soon as you accelerate the clocks to start and stop them that is no longer true. This is the famous twins paradox: If twin A flies to Andromeda and back at high speed, will she be elder or younger than twin B when she returns? During flight, both twins will insist that the other twin experiences the time dilation, so both should be younger? No, because twin A had to turn around and this changes things completely. Twin A will be younger when she gets back. I have not done detailed calculations for your scenario with the two clocks, but I am convinced the effects will cancel out exactly.
I also saw many videos on the topic, included that posted on the Veritasium channel. In my opinion you’ve just found a way of proving that c is equals in all direction (@21:35), at least in a small region of the space-time, as I did more than 3 years ago but with a different approach. Here is my point Let’s imagine we have a monochromatic source of light with wavelength λ propagating toward a diffracting grating with parallel slits each separated by the distance d. We know that we observe diffraction and the first maximum (order 1 diffraction) is offset by a theta angle according to the following equation: d sin(theta) = λ (eq.1) But from the definition of velocity (space divided by time = space/time ) we also know that c = λ /T = λ v (eq.2) being v the frequency of the light. Thus combining the two eq. we have c = d sin(theta) v (eq.3) Since the frequency of light should not chance by changing the direction of propagation (we can ideally measure it in the same reference frame at rest)(*), if the velocity of light is different along the two directions, from eq. 2 we should observe different wavelengths for the two directions so that, according to the eq.1, by changing the orientation of propagation of the light , we should find the first maximum along a different angle (no more exactly theta). This simple “gedanken experiment” should be sufficient to prove that the speed of light is equal in every direction. Indeed, I suspect that the speed of light cannot be different back an forth because of geometry reason (conservation of angular momentum and so on) and is equal to c for every small region of the space-time lattice. In fact assuming that the speed of light would be different along different direction (actually ways of propagation), by imposing that the average speed (roundtrip) is always the same, there will be necessarily a particular direction in which the speed of light in the two ways is the same (assuming that a small change in the direction will not change dramatically the speed of light). (*) We can be sure that the light traveling in different directions has the same frequency by ideally measuring the frequency with a clock for both the directions of propagation. Another way of measuring the frequency is to have an emitter (emitting in all directions) and a receiver exactly tuned to the emitter frequency placed once on the right of the emitter and then on the left of the emitter and record if the signal is received equally in both cases.
Would the difference in c result in different wavelengths for the same frequency propagating in opposing directions? Those quantities feel tractable to measure, and are a lot more local of a measurement, avoiding the whole loop problem. It felt like the video was about to go there a couple of times, so maybe I missed where the idea was dismissed as not even wrong 😅
I really like the concrete visualisation of what would happen if the speed of light was directionally anisotropic. It brings home the fact that in dealing with reality, you can't just arbitrarily change/question fundamental properties without it having collateral effects on everything. You have to step back and think what your proposition actually means in the greater context. Directionally anisotropic c won't just mess with your measurements, it implies a wholly different universe.
Would it also be possible to prove a constant one way speed of light by observing that gravitational lensing effects are the same for all directions of incidence? By my understanding gravitational lensing occurs because the spacetime is accelerating towards the beam, thus introducing a curvature. The radius of the curvature would then be related to both the field strength (which is constant from all directions of motion) and the speed of light.
Like you said at the beginning, when you start mixing the timing in wires or fiberoptic cables to send back the information to the clock, you've just added yet another variable and compounded the problems and then made the problem what's the transmission inside of a cable. The fact that there's some kind of refractive indexing in space really boggles my mind!
Now i might be missing something but we know the speed of sound is constant in air so could one not use spund to trigger the clock and the delay is the distance divided by speed of sound. Thus avoiding measuing the two way speed of light.
Yes, light propagating through space will be the same in any direction; but while the light is travelling, the receiver can move, which makes the detection at the detectors register different speeds. Your initial setup is indeed a two-way experiment... but you can change the setup: 1) 2 detectors in close vicinity, in the center, with a high precision clock that records the time when the detectors on left and right are triggered. 2) Put two emitters that have a stable clock, that will always tick at the same rate (more detail later). Each emitter will fire a short pulse (milliseconds are fine, you just mark the leading edge detection). Amazon has lasers that are good for up to 2 miles (10,000ft)... (light is approximately 1 foot per nanosecond, so over 10000ft is 10000ns or 10us.) Over the 10us, because of our motion through the universe towards VIrgo according to the redshift in the CMB, we move 370km/s or .0012 ft/ns... so it will move 12ft over 10us, and register a different speed from one side vs the other of +/-12ns... or a total difference in the speed of light of 24ns. (in the perfect arrangement).... so this would need to be aligned with the constellation virgo/cetus(opposite side constellation), so that there is a best-case... it's aligned to approximately +9degrees north; can't really just make this go any direction, or you can end up with a near null result, the orbit around the sun is 10% of the speed through the universe, and the rotation of earth is 1% of that... so +/-1.2ft or +/-0.012ft from those effects... the motion through the universe is much more on point. Mind you - the speed of light does not change based on the speed of the emitter... just the speed of the receiver - such speed is then c+v and c-v..... There may be additional skews to the stable clocks.... that once deployed they are not in exactly the same gravitational field... but this will be a constant effect, and the constant drift can be factored out. Air pressure is an insignificant factor; and since it's likely that the 4 miles the experiment covers (2 miles from one side of center and 2 miles from the other side) will likely be the same it ends up being non-measurable... and any change under like 100,000atm is barely notable... a few millibar is not going to change the experiment...same with humidity - the same amount of humidity is likely experienced across the apparatus. Another way to consider this is say you're playing catch with someone else, and every second they throw a ball at you at the same speed, if you move towards them, then the throw that happens while you are moving will be caught by you in a slightly shorter time. If you continue to stand in place, at this new distance per second, the ball will be registered as every 1 second. If you walk away during a throw, then the time it takes to catch the ball is slightly longer, again, until you stop. If the experimental apparatus is perpendicular to the velocity, that's basically initial conditions - and every pulse is received at 1 second intervals from both sides... as the apparatus aligns with the direction of motion, it's like the center detector is able to take a few steps forward, and pulses from one side will arrive in slightly shorter time or slightly longer time, whether the detector is moving towards or away from the emitters respectively. Once it reaches the maximum alignment, the pulses will still be every 1 second, but will be skewed from each other compared to where they started... if they start on every integer second, then it would be at +12ns and -12ns from the original state along a timeline. The times between each impulse registered from each side are subtracted from each subsequent sample, leaving a small delta change between each received pulse... at the end you'll have a net bias (probably) from clock skew, this can be removed by subtracting the final value from the initial value, and subtracting that sloped line from the result, biasing the beginning and and to 0. (This would mean a complete 24 hour cycle should be run... it would be less meaningful to do only 6 hours or even 12 hours - because at the 12 hour mark you're not necessarily in the same arrangement, since the apparatus is aligned with a specific point on the horizon, at 180 degrees of earths rotation, then the device is no longer in the same alignment as it started, but is tipped in a counter direction (that's not the right word but maybe you get the idea). I have been working on setting up this experiment, and if I can get the apparatus built, I would take it to the great salt lake, there are very few areas where there are 4 miles in an arbitrary direction that are entirely clear and flat, and wouldn't interfere with the line of site - could built something in the ocean maybe, but it would have to be tied to the ground. The two emitters and the center detector MUST be rigidly arranged - floating them, or launching them into space will not help. Anyway - I did setup a program for an FPGA that has a high speed clock, and two registers internally that can be latched when a signal from a light detector is received... the latched clocks can be read more slowly and stored in a computer over the next second between each pulse. The FPGA though is actually pretty slow, and although I'm almost able to get a 600ps clock, it's unstable, and a temperature corrected crystal oscillator that is more than a Ghz is expensive... That, and I don't know how many photons the photodetectors have to see, or how bright the intensity is - but I would expect actually quite a bit of noise from that receiver - plus, I don't know a good way to gate a laser pulse - for the same reason, propagation of all the voltage regulators, plus time to build up a signal to transmit are likely going to cause more noise in the experiment than +/-12ns. I've somewhat settled on using a synchronous AC motor that rotates 60RPM (once per second) and put a wheel with a 1mm slot on it - but even then, as that is actually pretty slow, it's going to uncover a fraction of the beam, and then the whole beam, and then start covering a fraction of the bean as it passes... which makes the signal leading edge not very concise... and is another point of noise. There are high precision frequency generators/counters but I find they are $16,000+ to be within a range of desired resolution, and then that is a bulky external thing, which introduces nanoseconds of delay with propagation of those signals... I just don't see it being done with off-the-shelf components. Probably have to make an ASIC with dedicated silicon for the clock and 2 latch registers (a super simple thing, though you do need about 52 bits of precision, which is quite a lot of bits, and a long chain to update.... which potentially makes the clock have grey bits in the middle while it's still counting a clock tick... the clock edge at one end will change even while the middle of the counter is still updating from the previous tick. That and the light detectors need to be pretty precise, and the module I got, the sensor is actually not JUST the sensor but has a transisitor the detector is attached to - which, again, signal jitter/noise. I don't know if maybe I got a DLP(?) sort of chip used in projectors that could gate small mirrors in two directions pretty fast... I would think those are still on the scale of 100s of microseconds... and far from the nano/picosecond gating I would want. Speaking of space, changes in space of the gravitational field only propagate at the speed of light also, which means ahead of our planet the field is somewhat compressed, and slightly more dense and is slightly weaker, so satellites will orbit just slightly further away (1 meter per kilometer roughly), and on the trailing side, the field will be elongated, and have slightly greater effect, making satellites on that side orbit slightly closer... This makes GPS signals always take the same amount of time to arrive. There is a thing in the solar system the 'Axis of evil' which is an overall alignment of the elliptical orbits of things that is always in the same direction. One could argue that GPS (multiple emitters some fixed distance away from a central detector that registers the time in the same location with 0 propagation time of the clocks signal) is the same as the above; that is, the same, except for the specification of being rigidly attached. On my channel the last few videos I did were on simulations I built (open source, links in descriptions to the demos and/or more information documents in the github repo) that are based on the one-way constant speed of light... and indeed, any two-way measurement will always be a constant time for a distance, in any frame, moving in any direction, while the individual one-way paths are neither the speed of light (as registered by a moving receiver). Light, once emitted no longer has anything to do with the source that emitted it, and it propagates in space regardless of what the source does afterward... Though the net combined effect does result in light aberration, both on transmission, and on reception of signals (see synchrotron radiation beaming effect for more about this aberration for electrons travelling close to the speed of light - the direction the light is detected is mostly all directed forward.
gps satellites can give a unique one-way predictable source of future information that could reduce need for timing data / start cues / synchronizing, right? and will be below margin of error, else you use a signal like the upcoming weird nova we expect, the arc difference will be minimal, but gps can be timed in midpoint, and as they are timestamped, reading the data means nothing then (but you can say speed of causality slows the clock?)
Aren’t the conservation of energy and momentum downstream of Noether’s theorems, which take symmetry as an assumption? Saying c must then be symmetrical is a circular argument.
Every once in a while, I think about the one-way speed of light and am convinced there must be a simple solution. Every single time, I whip out some paper and start drawing diagrams, labelled with t's and d's and light sources and detectors. Then, I measure the amount of "lefts" and "rights" a particular light pulse must take, sigh, and scrunch up my work. Only to suddenly be struck by inspiration a few months later.
I think the speed of light is the same in both directions. Because if speed of light in one way was different even by a tiny amount, the universe would look completely different in different directions. Also, if speed of light was different in different directions, things like GPS would break because the time dilation would be different than expected.
Can't we just measure the length contraction or time dilation from two similar yard sticks or clocks moving in different direction to conclude if the speed of light same or different ....
If the speed of light varied based on direction, then the observable universe would not be spherical and we would not be at its center. We'd see farther in some directions than others and thus see more stuff in those directions. We could (roughly) measure the relative speeds of light coming from different directions by counting the number of galaxies visible in those patches of the sky. I think it would also show up as an anisotropy in the size of the temperature fluctuations in the CMB. The relation between redshift and distance would be anisotropic. And probably other stuff too that I can't think of at 3 in the morning.
I thought of similar things. If there are enough stars with redshift to measure the expansion of the universe it should be possible to see a different speed of expansion in different directions if rhe speed of light would be direction dependent. The point of seeing more galaxies is brilliant as well (If mine has any merit & independent of that)
14:10 I'm ptobably overlooking something obvious, but is there an equivalent to E=mc^2 for a similar elastic medium. E=mk/p Does that actually mean something?
Best time resolution in a time-of-flight experiment with electronic triggers may be around 50 picoseconds. A 1 ppm error of c consequently can be achieved at a measuring distance of the light path of more than 15 km. Getting a stable signal transport over e.g. a coax cable of the same length looks almost impossible. However, if it may be feasible, rotating the whole setup 180 degrees might answer the question of whether the vacuum in space is anisotropic, of course only at the precision level of 1 ppm..
I'm probably talking nonsense here but: What about an apparatus similar to that shown at 3:36 but with the bottom right mirror reflecting the pulse to the left? Both reflected pulses would then hit a detector in order to mesure their flight time. If the speed of light of a reflected pulse is different from the one of the emitted pulse, the mesured two way speed of light for the two reflected pulses would be different because they both have the same path length from the laser to the beam splitter but different reflected path length.
The energy and momentum conservation argument is pretty rock solid, having a different speed of light in different directions would break so many symetries (and thus conservation laws) nothing would end up making sense. You could even use that argument to prove that the momentum transfer and thus the speed must be the same in both directions by measuring the radiation pressure exerted in opposing directions. One also has to ask what direction would the light care about, the direction of the observer, the universe, some mystery third thing? If its the universe the entire idea of relativity breaks down. I have to say that although directly measuring the speed of light in different directions may be impossible due to the arguments you and Veritasium laid out, the one way speed can be indirectly measured using shared clocks. Using a shared clock, while not a direct measurement because you must infer that the clocks are indeed shared and synced up, is a straightforward way to do it. Many research groups doing quantum networking already have timing and synroization of sub ns presistion over hundreds of km, not to mention GPS which relies on super presice shared clocks.
I don't think so. Lorentz and Poincare based their aether theory on the assumption that the speed of light is constant only in the stationary aether. This is how Lorentz created his transformations. On this basis, he concluded that the forces in the atom are electromagnetic in nature and this simply explains the mechanism of time dilation. It seems to me that even in the so-called in the ether, the speed of light is not constant - in a gravitational field it slows down in the direction of the action of gravitational forces, because gravity itself results from the imbalance in the transmission of forces at the atomic level in deformed space.
Isnt the michelson moreley experiment aimed at proving the two way speed is the same as the one way speed? The basis of their experiment was that an aether would cause light to be faster in one direction than the other as we moved through it.
One option might be to get 3 atomic clocks, 1 at the signal source and 2 for the other ends where the signal is detected. Place all clocks together in the same location to start, then sync them all up or make a note of the time, or offsets compared to one another (short wires connecting them together since they're in the same place) Physically carry 2 of the clocks away from the middle one in ether direction with a detector on each. Send a signal out in both directions making a note on a piece of paper of the time reading on the clock that's sending and the ones receiving so no wires involved, then look at the difference.
Honestly I think that it doesn't feel like that's possible, because if it were, then wouldn't we see a little bit of redshifting if we look at something in one direction vs a different direction? Since, if the speed of light is different, that would mean the waves would get a little doppler shift if you look at them from a different direction.
Were the experiments to detect 'the Ether' not similar to this: anisotropic light speed would indicate we are propagating through a medium. But light speed was found to be isotropic so the Ether model was dismissed. Anyone recall Michelson-Morley ?
Jesus Mary! Michaelson Morley's experiment proves that the one-way speed of light varies, because if it proved that it is constant, it would deny that dimensional contraction occurs in motion.
A couple points: 1) the expansion of space has the effect of slowing light in that the frequency is reduced and energy isn’t conserved. 2) the moving mass you mentioned in the end of the video you say we would not be able to detect, we do detect with LIGO. You are describing gravitational waves. I must say, I thoroughly enjoy your videos sir. Thank you and I always look forward to seeing the next! 👍🏻👍🏻
@@Holobrine You are correct, only accelerated masses in an asymmetric configuration (like two orbiting masses, but not one spinning mass) cause gravitational waves. But while a mass moving linearly with constant speed wouldn't radiate waves, it would still have a varying field that could in theory be detected locally with a LIGO-like system. In practice this particular case (e.g. caused by the sun moving relative to the earth) is absolutely impossible to detect with current designs, because the gradients in the field would be much lower, and several sources of noise like seismic noise and thermal noise, that are small in the region of interest for gravitational waves, explode in the low-frequency limit.
@@entcraft44 It would be a 'wave' but it would just have an extremely low 'frequency' If a large black hole zipped right past the earth you would probably be able to see it with ligo, lol. (and the 'frequency' would be higher if it were closer as well, if you think about it)
LIGO measures the 2-way length of a laser bouncing multiple times between two mirrors, it's the 2-way speed of light that determines the result. It uses interferometry so the laser has to go out and come back in order to work. So LIGO wouldn't be able to detect changes in the one-way thing.
Huygens Optics there is a channel called Dialect - do you know them? They are talking the same problem that you do - but they don't take for granted that the one way speed is 1/2 the 2 way speed... They allow it to be any value - even the one way being faster than the 2nd way... Thus they are trying to explain a different interpretation of relativity! I wish you could watch their videos and tell us what you think of them!
To me it seems that as long as you measure it localy, it HAS TO be different in one direction than the other because light reaches its speed limit instantly and its speed isn't affected by where it came from. Except the place where it came from isn't standing still, it's moving through the universe. Meaning that relative to where it came from, light has to travel at different speed in different directions. If you were to hypothetically travel at the speed of light and projected two light beams, one in front of yourself and one behind, relative to you the beams couldn't travel at the same speed otherwise they would travel at different speeds relative to "the universe" (because you're already traveling at the speed of light relative to "the universe" and light clearly can neither travel at twice the speed of light, nor stand still). Unless I'm misunderstanding something, this would tell me that as long as I'm moving relative to "the universe", as planets, solar systems and galaxies clearly do, relative to where I am light would be slower in on direction by the speed at which I'm moving in the other, since light doesn't accelerate and its speed doesn't change based on where it originated from. "The universe" because it is unclear what is it as a frame of reference.
Fantastic work, as always. I'm curious about the simulations you showed with opposing anisotropy. Certainly, it does not correctly model wave propagation as we observe, and violates the laws of energy conservation. But the phenomenon, more generally, of the disproportionate acceleration exerted in such a system, modeled with the springs at 20:00, does come off as one similar to gravity. I would be interested in whether gravitational fields could be accurately modeled using a modification of the anisotropic simulation at 20:39, altering the axis of anisotropy to extend radially, so that amplitude weakens outward and strengthens inward (if gravity can be said to have something like amplitude). Certainly, within an anisotropic system described by wave propagation, energy is not conserved, but if gravity is indeed an anisotropic phenomenon then such a description may still suffice, if the description accounts for energy beyond the wave propagation in some form-mass, perhaps? I also find it interesting that your proposed method of measuring the 1-way speed of light so closely resembles Eddington's eclipse observations which confirmed general relativity. Perhaps such a 1-way speed of light experiment could indeed be conducted using carefully placed spacecraft to exploit the sun or moon's gravity.
Indeed, the only circumstance where directional anisotropy makes sense is in an accelerated frame of reference. If you consider mass a very high frequency locally contained oscillation of the vacuum, it could explain why direction anisotropy accelerates mass and continuously increases its kinetic energy.
Well, if we did measure different times for different directions than it would mean that there was some additional moving gravitational potential somewhere in the vicinity of the light's trajectory and in fact we do use it cosmology when we measure the mass of black holes in a galaxy using long term gravitational lensing observations - recently covered by Anton Petrov on his channel discussing results from a Warsaw, Poland study on this exact subject. There's really no good reason to assume that light travels with different speeds in opposite directions when there's no change with respect to time in gravitational potential. If there was a difference we would not be able to infer anything about masses from any measurements, which would be contrary to experiments and day to day perception.
Very thought provoking. In no particular order… 1. I think the effect of moving mass in the return leg is somewhat related to how we detect gravitational waves, so perhaps not impossible to measure. 2. I wonder whether setting up a dual comb fs laser system, with the two periods walking past each other at a low (Hz or kHz rate) would enable you to decouple the effective speed of light in optical space from the effective speed of light in message passing (the wires to the clock). Instead of the time of arrival of the laser pulse being when the fs pulse appears, the time of the event is when the two fs laser pulses align at the detector. This has the effect of slowing down the speed of light by the mean repetition rate of the lasers divided by the difference in their repetition rates. For a fairly simple fs pair of lasers this might be 50MHz/500Hz, or a factor of 100,000, so instead of 300mm/ns the apparent speed of ligh would be 30mm/ms. Your wires would still be communicating at 300mm/ns, since they are only passing info that the correlation has happened. Effectively one of your two terms in your pair of equations will each have a time stretch factor in it and the other will remain purely related to c or c’. By changing the relative length of optical path and electrical path you can set up situations which should pick up if there are different speeds in the two different directions. 3. For the reasons you set out at the end of your video this would be an experiment with a very dull result. Thanks for the excellent work you do!
Regarding number 1: LIGO and other gravitational wave detectors are only sensitive enough in a small frequency band, limited from below by thermal and seismic noise and above by quantum shot noise. They would not be able to detect a mass in uniform motion since that is in the low-frequency limit. Another approach would be to use a pulsed laser system, but with my limited knowledge I doubt that the necessary precision can be achieved with current technology.
Two remarks I think are interesting: - the first part of the video actually shows that the claim "the speed of light is isotropic" is physically undecidable. That is, there exists no physical experiment that is able to discriminate if it's true or wrong - the second part, particularly at 20:25, is in fact very similar to what happens in a (uniformly and linearly) accelerated frame. Think, for example, of a rindler's observer in special relativity. Such an observer eventually disconnects himself from half of the universe behind him, but sees a massive blue shifted universe in front of him. So, in some sense, you're giving a self-consistent argument of the form "the speed of light is isotropic you're an inertial observer", which is perfectly in lign with special relativity
Third remark: What happens if you rotate the experimental device you considered by pi midway? Say, instead of having the light travel in empty-space, it travels in some fiber optic that is also assumed two-way isotropic
Indeed well spotted. All discussion is for an inertial frame of reference, which I maybe did not stress enough. Acceleration indeed can be a cause of anisotropy and therefore alway requires energy.
Well you can always modulate a timestamp on a light pulse And then decode it to get the speed relative to the receiver you point a "laser" at tx an rx both have atomic clock, and you can synchronize both using a 3rd atomic clock half the distance between Rx and tx to average out any time dilation effect... How long does the second pulse take on average from the previous detected pulse, knowing the distance traveled and the time encoded on the radio signals
You don’t make mention of length contraction/time dilation in your proposal which depends on the speed of light in a given direction. Given that these depend on c, it’s likely that these effects would cancel out the infinite energy and there would be no observable difference if the speed of light was different in different directions. It would be interesting to see a follow up taking these into account.
One possibility for a different speed of light in opposite directions is when one assumes the substrate for the light propagation, the "ether", moves in a specified direction relative to the apparatus. So if we cannot measure the speed difference according to direction what about the Michelson-Morley result ?
just put the sensor and transmitter right next to the timer, and extend the experiment ~200km to deal with precision error. if you form a loop you should be able to time the 1-way travel. then you can compare it to the two-way by reflecting it back and having the sensor on the other side instead of ending with the sensor. the big question i want answered though is does it take longer for light to travel the same distance when being reflected instead of just travelling through vaccuum. does the energy linger on the surface of the reflector? why does it seem like everyone just assume that the reflection of light is instant?
When I watched the Veritasium video, I thought of 2 experiments: 1. Examining deep field astronomical observations to see if the universe looks younger in any particular direction. 2. Take a device consisting of a laser facing an array of detectors, spin it at a constant rate, and measure the change of the deflection. Then twist the axis of rotation so as to check every 3D direction, and measure the change of the change in deflection of the beam. If the deflection remains the same for a system in constant rotation, then the 1-way speed of light is the same in all directions. Do these experiments fail, and if so how?
Clever set up. I always thought that a refractive medium with a secondary reflective surface would demonstrate the incident speed and the reflected speed would have to be the same. If they were different it wouldn't follow the law of reflection in the medium after a while. If it changed speed it would alter the secondary oscillation of the electrons in the medium, as it is time dependent, and thus change the summed slower electric field oscillation creating a different reflection angle about the normal axis.
Why isn't it setup so the sensor detects the light twice? First the light directly from the source and second the light off the mirror? Distances and time between sensor detections are known
Very good video and I think you are correct. Light speed is the same in any direction - HOWEVER what we really want to look at is if it takes the same time for light to go the same distance one way or the other. You may say that if we measure the same length both ways then of cause it should take the same time if light speed is the same. I then say but what if the whole setup is moving? That is really the thing I believe as we all know that we are moving and as the speed of that movement is unknown it has been said that any reference point is valid. I believe that if you put a movement into the setup it is clear that the time taken one way must be different to the other way but to find those time differences is THE REAL PROBLEM. Remember the distance scale is moving with the observer. When we become able to check the time taken for light to go in any direction we should then be able to calculate time dilation for every observer correctly. The way I understand that it is at the moment is that an observer traveling near the speed of light can claim to be the reference point and there fore standing still. If this observer sends a light beam in his own travel direction it obviously can't travel very fast COMPARED TO THIS OBSERVER as other ways it would break the speed of light. My view point is that all observers are probably moving at different speeds and in different directions BUT then there must be an ABSOLUTE STAND STILL. This will allow all observers to see the speed of light at their point AND THEIR TIME SPEED to calculate every body else's time speed and actual speed as well I believe. Time dilation is real I believe but that is really not much different to that we freeze meat. The frozen meat lasts longer (stays young longer) before it gets too old to eat.
17:50 Well but the point here is not that the speed of light doesn't change. It's only saying that the 2 way speed is identical for both A and B. Try moving A and B closer to the mirrors, without moving the mirrors. Then the signal from B will arrive earlier than A. We can do for this an interference experiment in the vertical direction, aligned with the gradient of the gravitational field :) this experiment has been done and the result is called the shapiro delay. The reason you don't see interference normally is because the gravitational field is pretty much identical in every point of the room you're in. 21:50 well we DO have gravitational redshift, so energy conservation does not really apply here...
If the speed were different in different directions, then you are correct it would cause uneven dispersal, but we could not observe that unevenness unless we were standing at some third perspective independent of those. Perhaps though we would see sections of the night sky that would be apparently more bright or darker.
I was thinking about a slightly different setup. Synchronize two clocks at point A and tell Alice shoot a laser pulse towards point B in one hour. Bob then takes one of the clocks to point B and measures the incoming pulse there. The clock should show 1h + (distance between AB) / (speed of light in the AB direction). Where does this experiment go wrong?
Let's say you have an energy ball at 2 jules. If the ball is heading in a direction where everything is under 2 jules, it would move past them by force, right? So what happens when the 2 jule energy ball hits something equal or bigger? Wouldnt there be some form of energy transferance that either absorbed or repelled the 2 jule energy ball? If that happened, wouldn't the 2 jule energy ball incrementally lose energy in percentages until it, on it path of travel, became disproportionately equal to any energy ball ot could meet? In any direction the ball could reflect to? And if theyre spherical, couldnt that energy bounce off anywhere, unless it hit perfectly flat to its opposite line of direction?
I am familiar with the reasoning in the Veritasium video, which is repeated (and added onto) in yours. Yet, in principal, I think the one-way speed of light can be measured. But it may be difficult in practice. Say we have two lasers opposite and parallel to each other. If we give both an extremely short and equal amount of time to reach the other side before they are occluded by some device in the middle, it will then be clear to see whether both lasers make it to the other side in time or only one of them. The device in the middle makes sure the timing is equal by way of mechanics, so without a clock. One such device may be two opaque rotating discs, each with a small hole close to the outer edge. The discs rotate extremely quickly in opposite directions, giving an extremely short amount of time where the holes align, allowing light at c to pass through but not anything slower. The holes will align at two places so the lasers can each use one of those. Because the lasers are both subject to the exact same device at the same time, equal time to pass though both holes is guaranteed as long as the discs spin at a constant rate which will be helped by the inertia of the discs. The alignment time of the holes will have to be tuned to the distance between them to require a speed of c for light to pass through. Although we could also simply keep increasing the speed of the discs until one or both lasers no longer come through. And we will have experimental evidence. Sadly this may be very hard in practice because the discs may have to rotate so fast they would be torn apart at distances we can manage on earth.
Strictly speaking, if one laser were exactly half or a fourth (etc) as fast as the other, both would still pass through. But that seems like enough margin to rule out most difference in speed.
The only issue with the conclusion is if we take the reference of an object in motion, let’s say traveling at say 1/2C then the speed of light forward of the object should appear to be slower (1/2C) while the speed of light aft of the object should appear to be faster (3/2C). This simple thought experiment seems to suggest that there MUST be a difference in 1 way speed of light for any object in motion.
A traveler moving at any speed would always measure the speed of light to be c in any direction, that's the fundamental observation that SR is based on.
The speed of light being different in different directions breaking the conservation of energy explains why a lot of people very desperately want this to be the case since it would allow for infinite energy devices.
At 2:33 onward, you talk about seeking to prove C1 = C1' and therefore get C2 = C1 + C1' but my question is: Does a mirror reflect instantaneously (Tm = 0) or is there a delay (Tm > 0)? If Tm > 0, it must still be very small, but when you're talking about the speed of light, small is not zero. I'm no physicist so I'm not sure if this is important or irrelevant, but, at the very least, C2 = C1 + C1' + Tm and so inferring the one-way speed of light by halving the two-way speed of light would still not be correct.
Around the year 2002 an interesting star was observed. It's V838 Monocerotis star which is surrounded by a thin nebula. After its "eruption" during which it brightened considerably, "an echo of light" on the nebula was observed. Since it expanded equally in all direction it is hard for me to imagine how the speed of light could be different depending on the direction. Also, in 1676 Rømer roughly measured the time the light takes to travel the Earth's orbit diameter and as far as I can understand his method, it's a one way determination of the speed of light which is the same, regardless of the Earth-Jupiter orientation in space (granted, only along roughly the plane of the ecliptic but I fail to see how that particular plane would be in any way special in the cosmic scale).
What you observe in the case of V838 would really depend on the direction of the anisotropy in the speed of light. Also, it would be hard to measure the difference in opposing directions if it is small compared to the speed of light itself. Don't ge me wrong though, I'm not advocating that the speed of light should be different, just that it is hard to prove that it can't be different.
If you used a set of springs to dampen a braking force again and again, they might not perform similarly when damping acceleration. So a system of real springs, initially with isotropic spring constants, can become directionally anisotropic in the way you described when it is used more often in one direction than the other, because of internal stress, entropy, etc, in the spring structure. Just an analogy, but could it be possible that the vacuum has a tiny anisotropy in the radial direction towards our galactic scale acceleration towards the great attractor, because all our mass is usually accelerating that way more often than not?
I will repost my comment from Veritasium's video HOW TO MEASURE SPEED OF LIGHT We can measure the time it takes light to go from point A to B in vaccum, bounce of the mirror and go back from B to A, BUT inside a glass Then repeat the experiment but in the opposite way. If the speed of light decreases by the same % going from vacuum to glass regardless of the direction, we should be able to see the difference in travel time. Furthermore, we can make 2 additional experiments (both ways in vaccum and both ways in glass), so that we have 4 measurements of time and 4 unknown velocities - easy to solve equations Great video btw :D
How are we defining speed? Don't we have to agree on what time it is at two different locations is in order to talk about speed at all? And to synchronize clocks at a distance, we need the speed of light in both directions. Is the question meaningful in the first place?
I have an idea of how it's theoretically possible to measure the one-way speed of light and I don't think it has any problems. You would send a light wave and start a clock at the same place and time. The light wave will reach a machine you have set up a certain distance away. The machine then sends some other type of wave back, e.g. a sound wave, through a carefully controlled medium, when this wave reaches the clock, the clock stops. You could use the measured time to calulate the one-way speed of light, accounting for the time the sound? wave would take to reach the clock. Can someone tell me if (and why) this wouldn't work? If it would work, I wonder why I haven't heard of this before.
Armand Hyppolyte Fizeau is one of the best names I've ever heard, holy crap P.S. how do you pronounce "Huygens"? Is part of that silent like the Vietnamese Nguyen (or are my white eyes just seeing the same letters and assuming a connection lol)?
When moving through a medium, like with a speaker and microphone that is not stationary in air, the one-way speeds can be different relative to your apparatus (but consistent relative to the medium, in the case of air). We now know that there is no aether and the speed of light is independent of the frame of reference, but this had to be established. So I would think the one-way speed of light has more to do with moving reference frames than with anisotropic physics.
The overly clever setup of the beam splitters, detectors and the clock reminds me of how designers of perpetual motion machines try to hide the energetic symmetry of their devices behind complex mechanics. The end result is the same: you just can't cheat your way around conservation laws and symmetries.
false, read my comment
@@NeroDefogger
"false, read my comment"
@@-danR 🤦♂🤦♂🤦♂ that's why I said "read my comment", because I meant MY ACTUAL EXPLANATION COMMENT ON THIS VIDEO 🤦♂🤦♂🤦♂ why would anyone think that "false, read my comment" 4 literal words would be the explanation of anything? I... I'm legit debufunged
I read the comment and the comment seemed to agree that the speed of light is the same in both directions.
@@raulkaap yes but not because of "cheat your way around conservation laws and symmetries", there are not "conservation laws" nor "symmetries", is just that if you measure it, it is the same, and it kinda makes sense because it would be a bit weird if it changed, but I don't necessarily see any clear reason why it couldn't, because there are no such "laws". but at the same time, your comment seemed to imply that there is no way to measure it, and in my comment I explain how there is, in fact, that is my point, that if you measure it, it is the same, because you can measure it
Holy smokes it's another masterpiece. Light is such a tricky subject to understand.
photons are downright fucking weird
the kid is a natural
no it is not
@@NeroDefogger yes it is
@@harriehausenman8623 maybe to you, not everyone is you
Somewhere, probably on PBS Spacetime, I learned that if I relabel the 'speed of light' as 'speed of causality', and light just going 'full speed', that makes the intuition slightly better. After a few years of mulling it over, I agree.
Matt has saved us again 🤗
your videos have repeatedly inspired me to re-engage with scientific learning. you make scientific understanding feel within reach of anyone, not just labs with big grants. also you are precise, but also humble. thanks for publishing your videos.
learning is what makes it unique, whereas any of the many edutainment-type science channels can ultimately only offer a form of teaching or lecturing, the distinct ability this channel has is to just offer learning instead
Don't make me cry
This is a great a great video! I want to point out that you missed a simple (yet arguably not obvious) way to make opposing anisotropy locally: make the spring-mass setup MOVE! Or equivalently, make the wave generator move, so that it excites a different mass at each time step. From the resting frame of the wave generator the wave propagation is then opposing anisotropic.
You might say that this makes no sense in the setting of the EM field because the field cannot have absolute motion. However, this is exactly the ether interpretation of the EM field (and spacetime itself): opposing anisotropy in the speed of light can be explained by assuming an absolute rest frame (the rest frame of the ether aka the rest frame of the CMB) and anisotropy resulting from motion relative to it.
Like a wave on a river
But while this possibility doesn't suffer from the energy conservation problem, the aether has other problems, and all the old aether theories have been disproven, e.g. by the movement of the earth trough the aether.
My takeaway is the following: The current theories describe most of our universe extremely well, so we often try to make small changes only. This video convincingly shows that an anisotropic one-way speed of light is not possible by making only small changes to our theories. But your comment hints that we can never consider every possibility.
That is, unless we employ a much more systematic approach to the assumptions underlying our modern understanding of the universe. There are a lot of interesting ideas out there, but nowhere near enough manpower and funding to check every single one with all the consequences individually.
@@entcraft44the key thing here is that the *old* aether theories have been disproven. There is a way to have an aether field in such a way where you get the exact same equations that you get from relativity, where the speed of light is constant in the reference frame of the aether, and objects moving relative to the aether get length contracted and time dilated. This formulation gives identical math as special relativity so it is not possible to distinguish which interpretation is closer to reality through measurements
@@entcraft44 Aether has been disproved but an absolute rest frame has not. In big part because of the two-way speed of light.
Once you assume that speed of light is the speed of transfer of information from one segment of space to another then you can have both the relativistic effects as well as an absolute reference frame, where the constant speed of light in any reference frame is just a result of slower processing of the thing moving close to the speed of light.
(ie: imaging you have a clock that measures the passage of time by sending a short ray of light towards a mirror and waiting for a response, that clock would indeed slow down the closed you go to the speed of light, and stop completely once you reach the speed of light, since light would have to move perpendicularly to the mirror, the exact same thing happens when one particle sends our information and waits for it to come back)
All the other space warping effects are also accounted for as being just a mirage caused by difference in processing speeds of information( photons in this case) that arrives perpendicular or and paralel to the motion as well as what just happens when light travels at a fixed speed..
First from the object's PoV the entire universe shrinks along direction of motion because its own processing slows (so everything just starts becoming very bright, and very blueshifted) Additionally the faster the object moves the more perpendicular to its movement the light has to be to actually hit it.
I did the math and some simulations on this and it all checks out, tho I'm still unsure as to where the extra mass is coming from when you accelerate things close to the speed of light.
The best explanation of 0ermittivity and permeability ever, and the analogy with mechanics made so clear the concepts of isotopi and anisotropy of the material... literally wow, You deserve an honorary Professorship
Nothing but garbage on TV tonight, what a brilliant timing!
you still watch TV?
@@farhanrejwan It's my excuse for laying on the couch.
@@Shokkwavez i think i'd still prefer my phone or my pc on a couch 😂
@@farhanrejwan surprinsingly, many ppl still enjoy oldtimes TV bc they don't need to choose something, just the channel and watch watever they give youl; I have a tiny amt of doubts and also no evidence, but I believe this could be cause by mental exhaustion,(which can come from anywhere), at least for ppl who do it for long periods of time......... also could be they don't want to bother looking for a youtube channel or whattever that plays random shit, like simpsons episodes or so,, due to ignorance or also lack of mental energy, there is a possible third scenario, which involves a third party showing the subject how to look for and/or setup such tvlike experience on phone, pc or whatever, but the subject will refuse, possibly due to lack of mental energy.... yea, i know, it's the root of all problems, even mine..... what am i even doing here, all coked up, writing pretentious nonsense..... ah.....
i think people are too afraid of "unsolvable" problems like this. if our current model does not predict any effects from this hypothetical and we do not observe any effects to prove our model wrong in that regard, it truly does not matter what the answer is. its like trying to prove the existence of an unobservable god with science.
i appreciate the insight on how a closely related problem is much more approachable
Exactly, Russel’s teapot
@@YuriyKrivosheyev not quite. russel's teapot places some arbitrary "burden of proof" on the "positive claims". in reality, neither claim inherently has more merit than the other. the proper way to look at it is that whether this teapot "exists" or "doesnt exist" cannot mean anything unless it has observable effects (which it does not). embrace the fact that there is no useful answer rather than arguing one way and demanding proof from the one arguing the other way. otherwise its like trying to prove a god *doesnt* exist with science.
Yes, but… not really?
Science has its dialectic in pure idealism, the imaginary, intuitive and creative. These “brain functions” are our pathway to pure empirical knowledge - ontology and epistemology, subject and object.
More culturally, we depict someone getting a bright idea (lightbulb moment in cartoons), and then the “labour” to prove or disprove, calculate vs experiment, theory vs observation. As this video also shows, Einstein knew his Maxwell, and “imagined” how time, space, mass and energy interacted - and, well it still works, empiricism backwards and forwards to now.
Einstein also put a nail in time travel, so the only way forwards is - unknown and undefined. Perfect for those “unscientific” traits we have, besides logic or reasoning.
This is also how “wrong ideas” emerge, but are not empirically sound if they can be disproved. There are always “ideas” in a sort of superposition - before we are able to both prove and disprove. Both are necessary.
If I’m wrong in this position, we have invented or discovered “everything”, there is no forward empirical path - nothing to imagine into the field. It’s a gloomy world at this time, but not _that_ gloomy?
@@jotch_7627Ultimately you can’t prove about a generic god with science, but many religions make enough additional specific claims about their gods that the gods they describe can be assessed
@@Holobrine and no proof for their gods
Fascinating! The examination of the asymmetric spring analogy was very intuitive
once again, what a brilliant video. one way speed of light is such a complex thing to wrap my head around, when I'm thinking about it i always get lightheaded.
exactly, that is how they do it, they try to " complex thing to wrap my head around" their way so that you don't try to put yourself on it, to try to understand it, to fight it, to bother with it, and to just accept what they say, they scare you, they make you think you just can't understand it, make you feel dumb, fight that, you are not dumb, put them in their place
@@NeroDefogger they don't make me feel dumb. they're making me smarter by making my brain work until topic makes sense. whatever you are saying is not the way to learn anything. my comment had an intended pun.
@@tsraikage "is such a complex thing to wrap my head around" and "when I'm thinking about it i always get lightheaded" does not "make me feel dumb"? whatever you say... "they're making me smarter" no they are not. "until topic makes sense" the topics never make sense. "whatever you are saying" I'm pretty clear on what I'm saying, did you not understand what I said yet you claim to understand the topic of the video? "is not the way to learn anything" what? you really did not understand anything I said... "had an intended pun" ... what?
Well done. It's one thing to argue that we cannot measure the speed of light in a single direction (always requiring a reflection), but quite another to assert that the speed of light in one direction could be different from the speed of light in the other. While the E=cp explanation is much simpler, I very much enjoyed the full exploration of the possibilities of anisotropy. Recall that it was Maxwell's equations that eventually led to special relativity, where only the Lorentz transformation could account for a constant speed of light. So if Maxwell's equations are fundamentally true (omitting quantum perturbations), then even if you can postulate the speed of light being different in opposing directions, it's impossible to model it in a self-consistent way.
If we can't measure the one-way speed, does it even make sense to consider to question what the one-way speed really is? In SR one-way speed of light is just a choice of coordinates. The problem with the argument from Maxwells equation is that in their standard form you have already chosen an isotropic, Einstein synchronized coordinate system. If you choose another coordinate system, you will have equations with anistropic light speeds (and hence not the standard wavequation with c = 1/√(ε₀μ₀))
_"even if you can postulate the speed of light being different in opposing directions, it's impossible to model it in a self-consistent way."_ - it is actually the other way around. Any anisotropic model of the light speed is self-consistent, as along as the two-way speed is c. The one-way speed is theoretically impossible to determine and is a matter of definition. You can choose any one-way speed you like for your model, and you will still end up with the current consistent physics model.
You can use curved space instead of a mirror. Shoot a photon to orbit a black hole near the event horizon. It will come back to you. Just time the orbital period. You know the distance so you will know the one way speed.
@@jeffbguarino And how would you prove the act of going around the black hole didn't slow down or speed up the photon in some way? And let me be clear, it doesn't. I feel very confident that c is constant in all directions but the point is, how do you prove it?
@@wingracer1614 But the circle of space at the event horizon is all the same direction. This is how a direction is defined by light which takes the geodesic which is the shortest path. So it is one direction or the same direction or do you somehow superimpose a euclidean coordinate system on top? It is just as if the earth is at two different places in an inertial euclidean coordinate space and you are at two places at the same time.
If the speed of light slows down or speeds up at different places going around a black hole then it could do this anywhere else. Like between mars and the earth speeding up and slowing down all over the place and you would never know. So that is introducing a new animal into the problem.
I found the definition of a straight line on google: (which I never really knew before) The shape of spacetime requires a new definition of a "straight line". The way to define a straight line is that trajectory that is followed where object experience weightlessness. Otherwise, if you experience weight, then you must be accelerating (according to the equivalence principle) and, hence, you must be changing your direction through spacetime.
Since the space station feels weightlessness, then it is travelling in the straightest possible path through spacetime. Similarly the Earth is going straight as it orbits around the Sun. We standing here on the Earth are, however, accelerating.
Thus when you are at the event horizon or just above it, and fire a laser tangentially , the light will be in free fall but here the escape velocity is the speed of light, so the light can avoid falling into the black hole.
And this, ladies and gentlemen, is how you explain in a easy way something hard to understand. What a great video!!!
I love this approach! Finding contradictions in conservation laws and the inconsistent consequences if one assumes anisotropic speed of light.
I thought about this some time and realized that no matter what I try, I always end up with a two-way speed of light in the end. And I had some really crazy ideas that would be practically impossible to do even if we had access to materials that literally have unrealistic properties.
It's actually quite easy, technically you just have to take advantage of the non-linear nature of time dilatation. Basically you take some atomic clocks, move them rapidly in two different directions and graph the change in time. The time dilation on each clock should fit the curve t₀/√(1-v²/c²) where t₀ is the time for the 'rest' reference. If they fit to the same curve for each direction, then the one way speed of light is the same.
The other way is that you just use a particle accelerator to spin a proton or whatever at like 99.9999% C. If the speed of light were different then the particle would have to slow down on one side of the ring in order to maintain it's mass. (or it's mass would fluctuate on each side) in either case you'd need to adjust the timing or strength of the magnets to maintain the particle in the ring - in other words every time the large hadron collider is use the one-way speed of light is verified to be equal.
@@takanara7 _"If they fit to the same curve for each direction, then the one way speed of light is the same. "_ - they will always fit the same curve for each direction, independent of what the one-way speed of light really is.
_"in either case you'd need to adjust the timing or strength of the magnets"_ - you wouldn't need to, because the timing and current in the electromagnets would be subject to the same isotropic speed of the light.
The one-way speed of light is theoretically impossible to measure. Trying to come up with mechanisms that could measure it, is like coming up with mechanisms for a perpetuum mobile. It is just a matter of determining why it fails.
@@renedekker9806 Romer's experiment!
@@rogerphelps9939 :"Romer's experiment!"_ - That's a difficult one.
The mathematics for why Rømer's experiment is not a one-way speed of light measurement is described by L. Karlov in “Does Roemer's method yield a unidirectional speed of light?” Australian Journal of Physics 23, 243-258 (1970)
I have a hard time interpreting what the correct interpretation of the math is, but I think it may be the following:
Rømer's experiment measured the time between eclipses of Jupiter's moon Io. When the Earth in its orbit is moving away from Jupiter, there should be more time between those eclipses than when the Earth is moving towards Jupiter. Assuming the speed of the Earth wrt. Jupiter is the same in both case (but in opposite direction), you can calculate the (one-way) speed of the light from that difference.
But how do we know that those speeds are the same? We measure that by measuring angles towards the Sun. Those measurements are based on the lines of sight to the Sun and Jupiter, that is, based on the (one-way) speed of light from the Sun to the Earth.
Due to that, with an anisotropic speed of light, the Earth's speed would appear to be larger on one side of the Earth's orbit than on the other. That is, the orbit of Earth would appear to be more elliptical. The difference in travel time of the light between both side of the Earth's orbit would be attributed to that difference, rather than to the lower/higher speed of light.
@@rogerphelps9939 That is a difficult one. The mathematics for why Rømer's experiment is not a one-way speed of light measurement is described by L. Karlov, in “Does Roemer's method yield a unidirectional speed of light?” Australian Journal of Physics 23, 243-258 (1970)
I have a hard time interpreting the math, but I think it may be the following.
Rømer's experiment measures the apparent time between eclipses of the moon Io of Jupiter. It then compares those times at a part of the orbit of Earth when it is moving away from Jupiter, with the part where Earth is moving towards Jupiter.
Effectively, it is a comparison of the Doppler shift in two directions.
But the calculations assume the speed of the Earth wrt Jupiter is the same for both directions. The result depends on that, but that is merely a convention. To measure the "real" speed of the Earth, you would need to measure the distance Earth travels between two points in time. But how much you measure for that speed, depends on the one-way speed of light again. That is, you are measuring the one-way speed of light in relation to the one-way speed of light.
Wow! I love hove how concise yet clearly and detailed you describe such complex topics. You are great educator.
Fantastic video, I'm glad UA-cam recommended your channel!
wow i’m so glad i didn’t click away after the first half lmao
first half was amazing (for the viewer that’s not familiar w the conclusion)
just excellent motivational context-building
but the second half was an amazing next layer deeper of thought experiment / pre-analysis
I'm glad I've already watched the videos you referenced.. what a great time to be on the internet!
Now I am imagining that little green fempto-pulsed laser blip surfing a gravity wave and I am very happy with that.
Thank you for the spring mass analogy. Love when EE concepts match up pretty well to mechanical concepts. Gives more intuition about the more abstract concepts which can be powerful.
Love your videos by the way. Optics was one of my favorite courses in college and these videos solidify fundamental teaching while at the same time not shying away from more advanced concepts and topics. At the end of our course, our professor talked some about non-linear optics but treated it as dark magic. Would be awesome to explore some experiments or topics in that realm if your equipment would allow.
The equivalence principal tells us if we do these experiments while in freefall we will always get the same results. This is even true if the experiment is done in free fall towards the event horizon of a black hole (one large enough to ignore tidal effects).
However, observers in different reference frames may see things very differently. Light will always move at the same velocity, but clocks in different reference frames won't agree. Since the velocity is constant, the wavelength will appear different.
Very good video as usual. Thanks for the time spent on the videos of this channel, we all appreciate it.
There's an experiment that came to me today, just before this video came out, that I would love to see, but don't have the means to perform. I just recently learned that Neodymium-doped Lithium Niobate can be used as a lasing medium. I am incredibly interested in the interference and diffraction patterns produced by it, specifically when it is being used as a laser medium while also piezoelectrically oscillating. Lazing Lithium Niobate, even when not doped, is a parametric down conversion process too, producing more than typical entangled photon pairs! If you're interested at all I can link you to a few papers, and a source for Neodymium-doped Lithium Niobate on substrate.
Love you so much!!
I haven't been this excited for a video in a long time
Loved the explanation in this one, so eloquently put forth.
Thank you so much for making these video's, always look forward to them! Makes me wanna go back to school and study light.
The moment you said a spring with a different constant on one side than the other, I instantly realized as it vibrated, it would grow in vibration in a single direction and I was like:
“Free energy machine!!!” 😂
Brilliant reasoning and video - did you consider publishing it Physical Review Letters D (I think). The left right anisotropy also reminds me of Huygen's assumptions for wavelet sources in diffraction gratings. The whole question seems to be connected with the idea of relative separation in space-time and the presence of an impedance tensor. If that were so Classical Dynamics too would be in big trouble. Very fine work and thank you.
@Huygens Optics. Please consider the following and critique. I believe it would be a way to measure the 1-way speed of light.
Let a laser light source send a pulse horizontally from S to a mirror M. Between these points place a beam splitter B close to the source. Anywhere along a line below and parallel to the line S-M, place a receiver/Clock RC. For convenience, the first example assumes RC is positioned directly perpendicular to the midpoint on S-M, such that the distances B-RC and M-RC are equal. (Variations on this experiment would place RC perpendicular to points B and M to eliminate “x” or “-x” components of the down-beam velocity, and to compare results between experiments.)
A pulse of light from S would partially deflect to RC from B, starting the clock. The remainder pulse from the same beam would reflect off of the mirror M and traverse to RC, stopping the clock. Since the delay from either B to RC or M to RC could be made equal (centered) or related by trigonometrically by the geometry of the triangles formed by the paths, the delay between pulses at RC would be entirely due to the distance between B & M divided by the 1-way velocity of the beam, right?
If as you say in your video, the round trip (the C' paths) cannot be eliminated, I would ask it replacing them with a known transmission velocity wouldn't do the trick. For example, the speed of electric signal propagation in a wire is not the speed of light, and if the materials are consistent, the speed of the signal should also be. Alternatively, the speed of sound does not diminish with energy loss, so why not use sound waves for the paths from splitters to receiver? In this way there is only one light path, and only in one direction.
If in a Frame of Reference there is a shift in time as a function of shift in space, then light would effectively travel at different speeds depending on the direction it travels. This is exactly what happens in the equation t’ = gamma(t-vx/cc). In a moving Frame of Reference, time is shifted proportional to distance. It’s the x in the special relativity time dilation equation. Therefore the speed of light is different in the 2 directions depending on the Frame of Reference.
Does that mean we could trick conservation of energy with a clever application of relativity?
@@lmmortalZoddno. From a moving Frame of Reference perspective, the distance is identically contracted in both directions of the round trip journey, and time is similarly dilated. But in one direction time is also shifted forward proportional to distance and then shifted backwards in the opposite direction proportional to distance. This means in effect from the moving Frame of Reference, light seems to travel at a different speeds in the two directions to compensate for this shifting of time as a function of distance.
Light can have different speeds in two opposite directions with respect to a moving frame of reference. The law of conservation of energy is still preserved.
Huygens Optics Video: My favourite kind of photons 🤗
*Advanced Tinkering* channel just mentioned you in an absolutely lovely way in the video "Creating Ultra-Fine Details in Titanium - 20 Micron Resolution" ! 🤗
Thanks for actually adding to the discussion.
Good video.
You compare the situation with an elastic medium that has different spring constants for the two directions, which leads to violation of conservation laws. But that is not the only way to achieve anisotropic propagation speed. The medium itself could be isotropic, but simply have a velocity in one direction. That would not lead to violation of conservation laws.
I think the more profound observation is that there is a delicate link between the one-way speed of light, and how simultaneity of distant events is defined. The definition of simultaneity makes it theoretically impossible to measure the one way speed of light. You might want to do another video on that.
Ever since I discovered this myself, I have marvelled at the following simple sentence in the introduction of Einstein's 1905 SR paper:
_"We have not defined a common “time” for A and B, for the latter cannot be defined at all unless we establish _*_by definition_*_ that the “time” required by light to travel from A to B equals the “time” it requires to travel from B to A"_
Einstein already knew that what you have discovered as well: that the one-way speed of light is a matter of definition, not of measurement.
My day just got a whole lot better!
It seems like this is just the simultaneity paradox in disguise. Alice and Bob meet up, get two clocks, and synchronize them. Alice moves a distance D left, Bob moves a distance D right, and they both stand still (relative to each other). Every second on the second, Alice flashes a light. Bob sees these flashes happening periodically, with a 1-second interval between each, and thus infers Alice experiences the same rate of time as he does. Moreover, Bob sees these flashes come slightly *after* each tick of his clock, with the delay given by dt=2D/c. When Bob also flashes a light every second on the second, Alice observes the exact same period of 1 second, and the exact same delay of 2D/c.
If the clocks are still synchronized after having been moved, then the one-way speed of light necessarily equals the two-way speed. The only way for the one-way speed to differ is if the clocks are *not* synchronized. But, there is already a reference frame in which the clocks are not synchronized! Consider Charlie, who is on a rocket ship moving 0.5c to the right, relative to Alice and Bob. Charlie sees light take longer to get from Bob to Alice, than it does going the other direction, simply because both Bob and Alice are moving fast left from Charlie's perspective. Relatedly, Charlie *also* sees Bob's clock being ahead of Alice's clock, i.e. the clocks are desynchronized to account for the differing times required to cross the distance in either direction.
The definition of "simultaneous events" is entirely dependent on your reference frame; simultaneity only makes sense for events at the same location. When events are at different locations, "simultaneous" depends on the observer. Measuring the one-way speed requires some notion of simultaneity across distances, which makes it inherently ambiguous. Even if the one-way speed were different than the two-way speed, then there would be some speed at which Charlie could move that results in the clocks being synchronized in his reference frame. We can simply redefine "simultaneous events" to be events which are simultaneous from Charlie's reference frame, which results in the one-way speed equaling the two way speed. This is a simple change of coordinates. The "one-way speed" is not unknowable in the same sense as Russel's Teapot. Rather, it's unknowable in the same sense that "which way's left?" is unknowable. It's an entirely semantic paradox, a simple matter of definition and choosing a coordinate scheme. It doesn't even qualify as a mystery.
I'm no scientologist but I always heard that read as "C1" and "C1 prime", not dash.
Probably I screwed up again, as a non-native speaker....
C1 dash is normal in the Britland afaik
@@HuygensOpticsYou're doing just fine! 👍
Thank you for all the time you put into these fascinating subjects. 🙏
The mea culpa, even when you are a non-native speaker speaks volumes about your scientific thought!
"scientologist"??
This is a random thought that hit me while watching:
1) We sync up two clocks very close to each other so we know the speed of light difference will have an extremely minimal effect
2) Both clocks are taken away from each other to some meaningful distance that will allow us to measure a difference in the speed of light
3) Based on a predetermined time in the future both clocks will cause a flash of light at the same instant
4) Both sides of the setup record the exact moment based on these clocks when they witness the light from the other end of the setup getting to them
The problem this solves is the relative nature of light which we are able to subvert due to us taking a higher perspective. We are looking at a 'global' time for this context & creating two independent light sources & measurements based on this 'global' time. This allows us to bypass the issue of relativity completely & get a measurement of the speed of light in one direction.
@grayishcolors The problem is that the two synchronized clocks experience time dilation when you move them, even if you move them only slowly. And if the speed of light really is anisotropic they would experience different amounts of time dilation even if you move them at the exact same speed over the exact same distance in opposite directions. Edit: And there would be no way to measure the difference in time dilation unless you bring the clocks back together again, at which point you're back to having done a two-way measurement rather than a one-way.
@@blahfasel2000 I was aware of the time dilation aspect, but the thought was that it has almost no effect unless we are talking about really fast speeds. All we care about in this hypothetically is to prove that light moves at different speeds, not necessarily have a perfect measurement.
I suppose you could have a perfect measurement like you said if you brought the clocks back together. I still don't see how that breaks the results though. If the time dilation is different in direction then the measurements will also see the difference which would add to the results.
Perhaps there's something I am not seeing as an issue here?
To add to this:
We have a CLOCK A & CLOCK B
Let us assume that time dilation somehow made CLOCK A run a whole 10 seconds slower by the time it gets to position for the experiment.
Light always goes at C from all reference frames meaning the time the light takes to travel won't be affected be affected. So the light will be released by CLOCK B 10 seconds faster than CLOCK A & arrive at CLOCK A 10 seconds sooner than it will create its own light.
This is of course assuming light moves at a constant speed & not two separate speeds.
The only way for this not to be the case is if the speed of light moves at a different speed as to exactly cancel out the time dilation. That doesn't make sense, though, because time dilation is relative & based on speed whereas the speed of light is always C from all perspectives.
@@grayishcolorsUnfortunately, for something as fast as light, even such minute difference is very important. U can't ignore the time dilation.
@@grayishcolors "time dilation is relative & based on speed" this is true only as long as you look at uniformly moving clocks. As soon as you accelerate the clocks to start and stop them that is no longer true. This is the famous twins paradox: If twin A flies to Andromeda and back at high speed, will she be elder or younger than twin B when she returns? During flight, both twins will insist that the other twin experiences the time dilation, so both should be younger? No, because twin A had to turn around and this changes things completely. Twin A will be younger when she gets back.
I have not done detailed calculations for your scenario with the two clocks, but I am convinced the effects will cancel out exactly.
I also saw many videos on the topic, included that posted on the Veritasium channel. In my opinion you’ve just found a way of proving that c is equals in all direction (@21:35), at least in a small region of the space-time, as I did more than 3 years ago but with a different approach.
Here is my point
Let’s imagine we have a monochromatic source of light with wavelength λ propagating toward a diffracting grating with parallel slits each separated by the distance d.
We know that we observe diffraction and the first maximum (order 1 diffraction) is offset by a theta angle according to the following equation:
d sin(theta) = λ (eq.1)
But from the definition of velocity (space divided by time = space/time ) we also know that
c = λ /T = λ v (eq.2)
being v the frequency of the light. Thus combining the two eq. we have
c = d sin(theta) v (eq.3)
Since the frequency of light should not chance by changing the direction of propagation (we can ideally measure it in the same reference frame at rest)(*), if the velocity of light is different along the two directions, from eq. 2 we should observe different wavelengths for the two directions so that, according to the eq.1, by changing the orientation of propagation of the light , we should find the first maximum along a different angle (no more exactly theta).
This simple “gedanken experiment” should be sufficient to prove that the speed of light is equal in every direction.
Indeed, I suspect that the speed of light cannot be different back an forth because of geometry reason (conservation of angular momentum and so on) and is equal to c for every small region of the space-time lattice.
In fact assuming that the speed of light would be different along different direction (actually ways of propagation), by imposing that the average speed (roundtrip) is always the same, there will be necessarily a particular direction in which the speed of light in the two ways is the same (assuming that a small change in the direction will not change dramatically the speed of light).
(*) We can be sure that the light traveling in different directions has the same frequency by ideally measuring the frequency with a clock for both the directions of propagation. Another way of measuring the frequency is to have an emitter (emitting in all directions) and a receiver exactly tuned to the emitter frequency placed once on the right of the emitter and then on the left of the emitter and record if the signal is received equally in both cases.
Would the difference in c result in different wavelengths for the same frequency propagating in opposing directions? Those quantities feel tractable to measure, and are a lot more local of a measurement, avoiding the whole loop problem.
It felt like the video was about to go there a couple of times, so maybe I missed where the idea was dismissed as not even wrong 😅
@@joshmyer9 And how would you measure it? Since you have the same frequency, you can assume that the speed is the same.
I really like the concrete visualisation of what would happen if the speed of light was directionally anisotropic. It brings home the fact that in dealing with reality, you can't just arbitrarily change/question fundamental properties without it having collateral effects on everything. You have to step back and think what your proposition actually means in the greater context. Directionally anisotropic c won't just mess with your measurements, it implies a wholly different universe.
If you were flowing with a river (but didn't know it), could you measure the one-way speed of sound in the water? Is this a valid analogy?
such an awesome channel for someone who did a bit of physics like me
Was just writing a comment about the AlphaPhoenix video when you mentioned it. Another under-rated channel.
I’ve often heard that the speed of light “pops out” of Maxwell’s equations, but this is the first time I’ve seen an explanation of how. Thanks!
Would it also be possible to prove a constant one way speed of light by observing that gravitational lensing effects are the same for all directions of incidence? By my understanding gravitational lensing occurs because the spacetime is accelerating towards the beam, thus introducing a curvature. The radius of the curvature would then be related to both the field strength (which is constant from all directions of motion) and the speed of light.
Like you said at the beginning, when you start mixing the timing in wires or fiberoptic cables to send back the information to the clock, you've just added yet another variable and compounded the problems and then made the problem what's the transmission inside of a cable. The fact that there's some kind of refractive indexing in space really boggles my mind!
Now i might be missing something but we know the speed of sound is constant in air so could one not use spund to trigger the clock and the delay is the distance divided by speed of sound. Thus avoiding measuing the two way speed of light.
Yes, light propagating through space will be the same in any direction; but while the light is travelling, the receiver can move, which makes the detection at the detectors register different speeds. Your initial setup is indeed a two-way experiment... but you can change the setup: 1) 2 detectors in close vicinity, in the center, with a high precision clock that records the time when the detectors on left and right are triggered. 2) Put two emitters that have a stable clock, that will always tick at the same rate (more detail later). Each emitter will fire a short pulse (milliseconds are fine, you just mark the leading edge detection). Amazon has lasers that are good for up to 2 miles (10,000ft)... (light is approximately 1 foot per nanosecond, so over 10000ft is 10000ns or 10us.) Over the 10us, because of our motion through the universe towards VIrgo according to the redshift in the CMB, we move 370km/s or .0012 ft/ns... so it will move 12ft over 10us, and register a different speed from one side vs the other of +/-12ns... or a total difference in the speed of light of 24ns. (in the perfect arrangement).... so this would need to be aligned with the constellation virgo/cetus(opposite side constellation), so that there is a best-case... it's aligned to approximately +9degrees north; can't really just make this go any direction, or you can end up with a near null result, the orbit around the sun is 10% of the speed through the universe, and the rotation of earth is 1% of that... so +/-1.2ft or +/-0.012ft from those effects... the motion through the universe is much more on point.
Mind you - the speed of light does not change based on the speed of the emitter... just the speed of the receiver - such speed is then c+v and c-v..... There may be additional skews to the stable clocks.... that once deployed they are not in exactly the same gravitational field... but this will be a constant effect, and the constant drift can be factored out. Air pressure is an insignificant factor; and since it's likely that the 4 miles the experiment covers (2 miles from one side of center and 2 miles from the other side) will likely be the same it ends up being non-measurable... and any change under like 100,000atm is barely notable... a few millibar is not going to change the experiment...same with humidity - the same amount of humidity is likely experienced across the apparatus.
Another way to consider this is say you're playing catch with someone else, and every second they throw a ball at you at the same speed, if you move towards them, then the throw that happens while you are moving will be caught by you in a slightly shorter time. If you continue to stand in place, at this new distance per second, the ball will be registered as every 1 second. If you walk away during a throw, then the time it takes to catch the ball is slightly longer, again, until you stop. If the experimental apparatus is perpendicular to the velocity, that's basically initial conditions - and every pulse is received at 1 second intervals from both sides... as the apparatus aligns with the direction of motion, it's like the center detector is able to take a few steps forward, and pulses from one side will arrive in slightly shorter time or slightly longer time, whether the detector is moving towards or away from the emitters respectively. Once it reaches the maximum alignment, the pulses will still be every 1 second, but will be skewed from each other compared to where they started... if they start on every integer second, then it would be at +12ns and -12ns from the original state along a timeline. The times between each impulse registered from each side are subtracted from each subsequent sample, leaving a small delta change between each received pulse... at the end you'll have a net bias (probably) from clock skew, this can be removed by subtracting the final value from the initial value, and subtracting that sloped line from the result, biasing the beginning and and to 0. (This would mean a complete 24 hour cycle should be run... it would be less meaningful to do only 6 hours or even 12 hours - because at the 12 hour mark you're not necessarily in the same arrangement, since the apparatus is aligned with a specific point on the horizon, at 180 degrees of earths rotation, then the device is no longer in the same alignment as it started, but is tipped in a counter direction (that's not the right word but maybe you get the idea).
I have been working on setting up this experiment, and if I can get the apparatus built, I would take it to the great salt lake, there are very few areas where there are 4 miles in an arbitrary direction that are entirely clear and flat, and wouldn't interfere with the line of site - could built something in the ocean maybe, but it would have to be tied to the ground. The two emitters and the center detector MUST be rigidly arranged - floating them, or launching them into space will not help. Anyway - I did setup a program for an FPGA that has a high speed clock, and two registers internally that can be latched when a signal from a light detector is received... the latched clocks can be read more slowly and stored in a computer over the next second between each pulse. The FPGA though is actually pretty slow, and although I'm almost able to get a 600ps clock, it's unstable, and a temperature corrected crystal oscillator that is more than a Ghz is expensive... That, and I don't know how many photons the photodetectors have to see, or how bright the intensity is - but I would expect actually quite a bit of noise from that receiver - plus, I don't know a good way to gate a laser pulse - for the same reason, propagation of all the voltage regulators, plus time to build up a signal to transmit are likely going to cause more noise in the experiment than +/-12ns. I've somewhat settled on using a synchronous AC motor that rotates 60RPM (once per second) and put a wheel with a 1mm slot on it - but even then, as that is actually pretty slow, it's going to uncover a fraction of the beam, and then the whole beam, and then start covering a fraction of the bean as it passes... which makes the signal leading edge not very concise... and is another point of noise. There are high precision frequency generators/counters but I find they are $16,000+ to be within a range of desired resolution, and then that is a bulky external thing, which introduces nanoseconds of delay with propagation of those signals... I just don't see it being done with off-the-shelf components. Probably have to make an ASIC with dedicated silicon for the clock and 2 latch registers (a super simple thing, though you do need about 52 bits of precision, which is quite a lot of bits, and a long chain to update.... which potentially makes the clock have grey bits in the middle while it's still counting a clock tick... the clock edge at one end will change even while the middle of the counter is still updating from the previous tick. That and the light detectors need to be pretty precise, and the module I got, the sensor is actually not JUST the sensor but has a transisitor the detector is attached to - which, again, signal jitter/noise. I don't know if maybe I got a DLP(?) sort of chip used in projectors that could gate small mirrors in two directions pretty fast... I would think those are still on the scale of 100s of microseconds... and far from the nano/picosecond gating I would want.
Speaking of space, changes in space of the gravitational field only propagate at the speed of light also, which means ahead of our planet the field is somewhat compressed, and slightly more dense and is slightly weaker, so satellites will orbit just slightly further away (1 meter per kilometer roughly), and on the trailing side, the field will be elongated, and have slightly greater effect, making satellites on that side orbit slightly closer... This makes GPS signals always take the same amount of time to arrive. There is a thing in the solar system the 'Axis of evil' which is an overall alignment of the elliptical orbits of things that is always in the same direction. One could argue that GPS (multiple emitters some fixed distance away from a central detector that registers the time in the same location with 0 propagation time of the clocks signal) is the same as the above; that is, the same, except for the specification of being rigidly attached.
On my channel the last few videos I did were on simulations I built (open source, links in descriptions to the demos and/or more information documents in the github repo) that are based on the one-way constant speed of light... and indeed, any two-way measurement will always be a constant time for a distance, in any frame, moving in any direction, while the individual one-way paths are neither the speed of light (as registered by a moving receiver). Light, once emitted no longer has anything to do with the source that emitted it, and it propagates in space regardless of what the source does afterward... Though the net combined effect does result in light aberration, both on transmission, and on reception of signals (see synchrotron radiation beaming effect for more about this aberration for electrons travelling close to the speed of light - the direction the light is detected is mostly all directed forward.
Mind you this does not measure the one-way speed of light. It only measures the differences in the one way speed of light
gps satellites can give a unique one-way predictable source of future information that could reduce need for timing data / start cues / synchronizing, right? and will be below margin of error, else you use a signal like the upcoming weird nova we expect, the arc difference will be minimal, but gps can be timed in midpoint, and as they are timestamped, reading the data means nothing then (but you can say speed of causality slows the clock?)
Aren’t the conservation of energy and momentum downstream of Noether’s theorems, which take symmetry as an assumption? Saying c must then be symmetrical is a circular argument.
Every once in a while, I think about the one-way speed of light and am convinced there must be a simple solution. Every single time, I whip out some paper and start drawing diagrams, labelled with t's and d's and light sources and detectors. Then, I measure the amount of "lefts" and "rights" a particular light pulse must take, sigh, and scrunch up my work. Only to suddenly be struck by inspiration a few months later.
Hello from Thought Emporium. Looking forward to some great content.
I think the speed of light is the same in both directions. Because if speed of light in one way was different even by a tiny amount, the universe would look completely different in different directions.
Also, if speed of light was different in different directions, things like GPS would break because the time dilation would be different than expected.
Can't we just measure the length contraction or time dilation from two similar yard sticks or clocks moving in different direction to conclude if the speed of light same or different ....
If the speed of light varied based on direction, then the observable universe would not be spherical and we would not be at its center. We'd see farther in some directions than others and thus see more stuff in those directions. We could (roughly) measure the relative speeds of light coming from different directions by counting the number of galaxies visible in those patches of the sky. I think it would also show up as an anisotropy in the size of the temperature fluctuations in the CMB. The relation between redshift and distance would be anisotropic. And probably other stuff too that I can't think of at 3 in the morning.
I thought of similar things. If there are enough stars with redshift to measure the expansion of the universe it should be possible to see a different speed of expansion in different directions if rhe speed of light would be direction dependent. The point of seeing more galaxies is brilliant as well (If mine has any merit & independent of that)
I enjoyed quite a lot this video and the way this problem is presented.
14:10 I'm ptobably overlooking something obvious, but is there an equivalent to E=mc^2 for a similar elastic medium.
E=mk/p
Does that actually mean something?
Best time resolution in a time-of-flight experiment with electronic triggers may be around 50 picoseconds. A 1 ppm error of c consequently can be achieved at a measuring distance of the light path of more than 15 km. Getting a stable signal transport over e.g. a coax cable of the same length looks almost impossible.
However, if it may be feasible, rotating the whole setup 180 degrees might answer the question of whether the vacuum in space is anisotropic, of course only at the precision level of 1 ppm..
I'm probably talking nonsense here but:
What about an apparatus similar to that shown at 3:36 but with the bottom right mirror reflecting the pulse to the left?
Both reflected pulses would then hit a detector in order to mesure their flight time.
If the speed of light of a reflected pulse is different from the one of the emitted pulse, the mesured two way speed of light for the two reflected pulses would be different because they both have the same path length from the laser to the beam splitter but different reflected path length.
The energy and momentum conservation argument is pretty rock solid, having a different speed of light in different directions would break so many symetries (and thus conservation laws) nothing would end up making sense. You could even use that argument to prove that the momentum transfer and thus the speed must be the same in both directions by measuring the radiation pressure exerted in opposing directions.
One also has to ask what direction would the light care about, the direction of the observer, the universe, some mystery third thing? If its the universe the entire idea of relativity breaks down.
I have to say that although directly measuring the speed of light in different directions may be impossible due to the arguments you and Veritasium laid out, the one way speed can be indirectly measured using shared clocks. Using a shared clock, while not a direct measurement because you must infer that the clocks are indeed shared and synced up, is a straightforward way to do it. Many research groups doing quantum networking already have timing and synroization of sub ns presistion over hundreds of km, not to mention GPS which relies on super presice shared clocks.
I don't think so. Lorentz and Poincare based their aether theory on the assumption that the speed of light is constant only in the stationary aether. This is how Lorentz created his transformations. On this basis, he concluded that the forces in the atom are electromagnetic in nature and this simply explains the mechanism of time dilation. It seems to me that even in the so-called in the ether, the speed of light is not constant - in a gravitational field it slows down in the direction of the action of gravitational forces, because gravity itself results from the imbalance in the transmission of forces at the atomic level in deformed space.
Isnt the michelson moreley experiment aimed at proving the two way speed is the same as the one way speed?
The basis of their experiment was that an aether would cause light to be faster in one direction than the other as we moved through it.
One option might be to get 3 atomic clocks, 1 at the signal source and 2 for the other ends where the signal is detected.
Place all clocks together in the same location to start, then sync them all up or make a note of the time, or offsets compared to one another (short wires connecting them together since they're in the same place)
Physically carry 2 of the clocks away from the middle one in ether direction with a detector on each.
Send a signal out in both directions making a note on a piece of paper of the time reading on the clock that's sending and the ones receiving so no wires involved, then look at the difference.
Honestly I think that it doesn't feel like that's possible, because if it were, then wouldn't we see a little bit of redshifting if we look at something in one direction vs a different direction? Since, if the speed of light is different, that would mean the waves would get a little doppler shift if you look at them from a different direction.
Were the experiments to detect 'the Ether' not similar to this: anisotropic light speed would indicate we are propagating through a medium. But light speed was found to be isotropic so the Ether model was dismissed. Anyone recall Michelson-Morley ?
Jesus Mary! Michaelson Morley's experiment proves that the one-way speed of light varies, because if it proved that it is constant, it would deny that dimensional contraction occurs in motion.
A couple points: 1) the expansion of space has the effect of slowing light in that the frequency is reduced and energy isn’t conserved. 2) the moving mass you mentioned in the end of the video you say we would not be able to detect, we do detect with LIGO. You are describing gravitational waves. I must say, I thoroughly enjoy your videos sir. Thank you and I always look forward to seeing the next! 👍🏻👍🏻
Does a mass at constant velocity produce gravitational waves? I thought it was only masses in orbits with angular acceleration
@@Holobrine You are correct, only accelerated masses in an asymmetric configuration (like two orbiting masses, but not one spinning mass) cause gravitational waves. But while a mass moving linearly with constant speed wouldn't radiate waves, it would still have a varying field that could in theory be detected locally with a LIGO-like system.
In practice this particular case (e.g. caused by the sun moving relative to the earth) is absolutely impossible to detect with current designs, because the gradients in the field would be much lower, and several sources of noise like seismic noise and thermal noise, that are small in the region of interest for gravitational waves, explode in the low-frequency limit.
@@entcraft44 It would be a 'wave' but it would just have an extremely low 'frequency' If a large black hole zipped right past the earth you would probably be able to see it with ligo, lol. (and the 'frequency' would be higher if it were closer as well, if you think about it)
LIGO measures the 2-way length of a laser bouncing multiple times between two mirrors, it's the 2-way speed of light that determines the result. It uses interferometry so the laser has to go out and come back in order to work. So LIGO wouldn't be able to detect changes in the one-way thing.
@@takanara7 you are mistaken
Huygens Optics there is a channel called Dialect - do you know them? They are talking the same problem that you do - but they don't take for granted that the one way speed is 1/2 the 2 way speed... They allow it to be any value - even the one way being faster than the 2nd way... Thus they are trying to explain a different interpretation of relativity! I wish you could watch their videos and tell us what you think of them!
To me it seems that as long as you measure it localy, it HAS TO be different in one direction than the other because light reaches its speed limit instantly and its speed isn't affected by where it came from. Except the place where it came from isn't standing still, it's moving through the universe. Meaning that relative to where it came from, light has to travel at different speed in different directions.
If you were to hypothetically travel at the speed of light and projected two light beams, one in front of yourself and one behind, relative to you the beams couldn't travel at the same speed otherwise they would travel at different speeds relative to "the universe" (because you're already traveling at the speed of light relative to "the universe" and light clearly can neither travel at twice the speed of light, nor stand still). Unless I'm misunderstanding something, this would tell me that as long as I'm moving relative to "the universe", as planets, solar systems and galaxies clearly do, relative to where I am light would be slower in on direction by the speed at which I'm moving in the other, since light doesn't accelerate and its speed doesn't change based on where it originated from. "The universe" because it is unclear what is it as a frame of reference.
Fantastic work, as always.
I'm curious about the simulations you showed with opposing anisotropy. Certainly, it does not correctly model wave propagation as we observe, and violates the laws of energy conservation.
But the phenomenon, more generally, of the disproportionate acceleration exerted in such a system, modeled with the springs at 20:00, does come off as one similar to gravity. I would be interested in whether gravitational fields could be accurately modeled using a modification of the anisotropic simulation at 20:39, altering the axis of anisotropy to extend radially, so that amplitude weakens outward and strengthens inward (if gravity can be said to have something like amplitude).
Certainly, within an anisotropic system described by wave propagation, energy is not conserved, but if gravity is indeed an anisotropic phenomenon then such a description may still suffice, if the description accounts for energy beyond the wave propagation in some form-mass, perhaps?
I also find it interesting that your proposed method of measuring the 1-way speed of light so closely resembles Eddington's eclipse observations which confirmed general relativity. Perhaps such a 1-way speed of light experiment could indeed be conducted using carefully placed spacecraft to exploit the sun or moon's gravity.
Indeed, the only circumstance where directional anisotropy makes sense is in an accelerated frame of reference. If you consider mass a very high frequency locally contained oscillation of the vacuum, it could explain why direction anisotropy accelerates mass and continuously increases its kinetic energy.
Well, if we did measure different times for different directions than it would mean that there was some additional moving gravitational potential somewhere in the vicinity of the light's trajectory and in fact we do use it cosmology when we measure the mass of black holes in a galaxy using long term gravitational lensing observations - recently covered by Anton Petrov on his channel discussing results from a Warsaw, Poland study on this exact subject.
There's really no good reason to assume that light travels with different speeds in opposite directions when there's no change with respect to time in gravitational potential.
If there was a difference we would not be able to infer anything about masses from any measurements, which would be contrary to experiments and day to day perception.
Very thought provoking. In no particular order…
1. I think the effect of moving mass in the return leg is somewhat related to how we detect gravitational waves, so perhaps not impossible to measure.
2. I wonder whether setting up a dual comb fs laser system, with the two periods walking past each other at a low (Hz or kHz rate) would enable you to decouple the effective speed of light in optical space from the effective speed of light in message passing (the wires to the clock). Instead of the time of arrival of the laser pulse being when the fs pulse appears, the time of the event is when the two fs laser pulses align at the detector. This has the effect of slowing down the speed of light by the mean repetition rate of the lasers divided by the difference in their repetition rates. For a fairly simple fs pair of lasers this might be 50MHz/500Hz, or a factor of 100,000, so instead of 300mm/ns the apparent speed of ligh would be 30mm/ms. Your wires would still be communicating at 300mm/ns, since they are only passing info that the correlation has happened. Effectively one of your two terms in your pair of equations will each have a time stretch factor in it and the other will remain purely related to c or c’. By changing the relative length of optical path and electrical path you can set up situations which should pick up if there are different speeds in the two different directions.
3. For the reasons you set out at the end of your video this would be an experiment with a very dull result.
Thanks for the excellent work you do!
Regarding number 1: LIGO and other gravitational wave detectors are only sensitive enough in a small frequency band, limited from below by thermal and seismic noise and above by quantum shot noise. They would not be able to detect a mass in uniform motion since that is in the low-frequency limit.
Another approach would be to use a pulsed laser system, but with my limited knowledge I doubt that the necessary precision can be achieved with current technology.
Two remarks I think are interesting:
- the first part of the video actually shows that the claim "the speed of light is isotropic" is physically undecidable. That is, there exists no physical experiment that is able to discriminate if it's true or wrong
- the second part, particularly at 20:25, is in fact very similar to what happens in a (uniformly and linearly) accelerated frame. Think, for example, of a rindler's observer in special relativity. Such an observer eventually disconnects himself from half of the universe behind him, but sees a massive blue shifted universe in front of him. So, in some sense, you're giving a self-consistent argument of the form "the speed of light is isotropic you're an inertial observer", which is perfectly in lign with special relativity
Third remark:
What happens if you rotate the experimental device you considered by pi midway? Say, instead of having the light travel in empty-space, it travels in some fiber optic that is also assumed two-way isotropic
Indeed well spotted. All discussion is for an inertial frame of reference, which I maybe did not stress enough. Acceleration indeed can be a cause of anisotropy and therefore alway requires energy.
This would require moving physical objects with a speed larger than the speed of light.
@@HuygensOptics ah you're right, my bad, so you can't save yourself like that either. So this is indeed undecidable, that's amazing
This was cool even if in the end I'm not convinced. Love this channel!
Well you can always modulate a timestamp on a light pulse
And then decode it to get the speed relative to the receiver you point a "laser" at tx an rx both have atomic clock, and you can synchronize both using a 3rd atomic clock half the distance between Rx and tx to average out any time dilation effect...
How long does the second pulse take on average from the previous detected pulse, knowing the distance traveled and the time encoded on the radio signals
You don’t make mention of length contraction/time dilation in your proposal which depends on the speed of light in a given direction.
Given that these depend on c, it’s likely that these effects would cancel out the infinite energy and there would be no observable difference if the speed of light was different in different directions. It would be interesting to see a follow up taking these into account.
One possibility for a different speed of light in opposite directions is when one assumes the substrate for the light propagation, the "ether", moves in a specified direction relative to the apparatus. So if we cannot measure the speed difference according to direction what about the Michelson-Morley result ?
19:24 What is that on the bottom right supposed to be?
slip of the tongue: replace string by spring
@@HuygensOptics Oh I see, a spring string bottom, I’m not native and had to look that up. 😅
just put the sensor and transmitter right next to the timer, and extend the experiment ~200km to deal with precision error. if you form a loop you should be able to time the 1-way travel. then you can compare it to the two-way by reflecting it back and having the sensor on the other side instead of ending with the sensor. the big question i want answered though is does it take longer for light to travel the same distance when being reflected instead of just travelling through vaccuum. does the energy linger on the surface of the reflector? why does it seem like everyone just assume that the reflection of light is instant?
When I watched the Veritasium video, I thought of 2 experiments:
1. Examining deep field astronomical observations to see if the universe looks younger in any particular direction.
2. Take a device consisting of a laser facing an array of detectors, spin it at a constant rate, and measure the change of the deflection. Then twist the axis of rotation so as to check every 3D direction, and measure the change of the change in deflection of the beam. If the deflection remains the same for a system in constant rotation, then the 1-way speed of light is the same in all directions.
Do these experiments fail, and if so how?
Clever set up.
I always thought that a refractive medium with a secondary reflective surface would demonstrate the incident speed and the reflected speed would have to be the same.
If they were different it wouldn't follow the law of reflection in the medium after a while. If it changed speed it would alter the secondary oscillation of the electrons in the medium, as it is time dependent, and thus change the summed slower electric field oscillation creating a different reflection angle about the normal axis.
Why isn't it setup so the sensor detects the light twice? First the light directly from the source and second the light off the mirror? Distances and time between sensor detections are known
Very good video and I think you are correct. Light speed is the same in any direction - HOWEVER what we really want to look at is if it takes the same time for light to go the same distance one way or the other. You may say that if we measure the same length both ways then of cause it should take the same time if light speed is the same. I then say but what if the whole setup is moving? That is really the thing I believe as we all know that we are moving and as the speed of that movement is unknown it has been said that any reference point is valid. I believe that if you put a movement into the setup it is clear that the time taken one way must be different to the other way but to find those time differences is THE REAL PROBLEM. Remember the distance scale is moving with the observer.
When we become able to check the time taken for light to go in any direction we should then be able to calculate time dilation for every observer correctly.
The way I understand that it is at the moment is that an observer traveling near the speed of light can claim to be the reference point and there fore standing still. If this observer sends a light beam in his own travel direction it obviously can't travel very fast COMPARED TO THIS OBSERVER as other ways it would break the speed of light.
My view point is that all observers are probably moving at different speeds and in different directions BUT then there must be an ABSOLUTE STAND STILL. This will allow all observers to see the speed of light at their point AND THEIR TIME SPEED to calculate every body else's time speed and actual speed as well I believe.
Time dilation is real I believe but that is really not much different to that we freeze meat. The frozen meat lasts longer (stays young longer) before it gets too old to eat.
17:50 Well but the point here is not that the speed of light doesn't change. It's only saying that the 2 way speed is identical for both A and B. Try moving A and B closer to the mirrors, without moving the mirrors. Then the signal from B will arrive earlier than A.
We can do for this an interference experiment in the vertical direction, aligned with the gradient of the gravitational field :) this experiment has been done and the result is called the shapiro delay.
The reason you don't see interference normally is because the gravitational field is pretty much identical in every point of the room you're in.
21:50 well we DO have gravitational redshift, so energy conservation does not really apply here...
Can we measure the point in space the two light beams meet to prove/disprove the spatially anisotropic case?
If the speed were different in different directions, then you are correct it would cause uneven dispersal, but we could not observe that unevenness unless we were standing at some third perspective independent of those. Perhaps though we would see sections of the night sky that would be apparently more bright or darker.
I was thinking about a slightly different setup. Synchronize two clocks at point A and tell Alice shoot a laser pulse towards point B in one hour. Bob then takes one of the clocks to point B and measures the incoming pulse there. The clock should show 1h + (distance between AB) / (speed of light in the AB direction).
Where does this experiment go wrong?
Let's say you have an energy ball at 2 jules. If the ball is heading in a direction where everything is under 2 jules, it would move past them by force, right? So what happens when the 2 jule energy ball hits something equal or bigger? Wouldnt there be some form of energy transferance that either absorbed or repelled the 2 jule energy ball? If that happened, wouldn't the 2 jule energy ball incrementally lose energy in percentages until it, on it path of travel, became disproportionately equal to any energy ball ot could meet? In any direction the ball could reflect to? And if theyre spherical, couldnt that energy bounce off anywhere, unless it hit perfectly flat to its opposite line of direction?
I am familiar with the reasoning in the Veritasium video, which is repeated (and added onto) in yours. Yet, in principal, I think the one-way speed of light can be measured. But it may be difficult in practice.
Say we have two lasers opposite and parallel to each other. If we give both an extremely short and equal amount of time to reach the other side before they are occluded by some device in the middle, it will then be clear to see whether both lasers make it to the other side in time or only one of them.
The device in the middle makes sure the timing is equal by way of mechanics, so without a clock. One such device may be two opaque rotating discs, each with a small hole close to the outer edge. The discs rotate extremely quickly in opposite directions, giving an extremely short amount of time where the holes align, allowing light at c to pass through but not anything slower. The holes will align at two places so the lasers can each use one of those. Because the lasers are both subject to the exact same device at the same time, equal time to pass though both holes is guaranteed as long as the discs spin at a constant rate which will be helped by the inertia of the discs.
The alignment time of the holes will have to be tuned to the distance between them to require a speed of c for light to pass through. Although we could also simply keep increasing the speed of the discs until one or both lasers no longer come through. And we will have experimental evidence.
Sadly this may be very hard in practice because the discs may have to rotate so fast they would be torn apart at distances we can manage on earth.
Strictly speaking, if one laser were exactly half or a fourth (etc) as fast as the other, both would still pass through. But that seems like enough margin to rule out most difference in speed.
Wouldn't the lasers have to fire synchronously for this to work?
The only issue with the conclusion is if we take the reference of an object in motion, let’s say traveling at say 1/2C then the speed of light forward of the object should appear to be slower (1/2C) while the speed of light aft of the object should appear to be faster (3/2C). This simple thought experiment seems to suggest that there MUST be a difference in 1 way speed of light for any object in motion.
A traveler moving at any speed would always measure the speed of light to be c in any direction, that's the fundamental observation that SR is based on.
The speed of light being different in different directions breaking the conservation of energy explains why a lot of people very desperately want this to be the case since it would allow for infinite energy devices.
At 2:33 onward, you talk about seeking to prove C1 = C1' and therefore get C2 = C1 + C1' but my question is: Does a mirror reflect instantaneously (Tm = 0) or is there a delay (Tm > 0)? If Tm > 0, it must still be very small, but when you're talking about the speed of light, small is not zero.
I'm no physicist so I'm not sure if this is important or irrelevant, but, at the very least, C2 = C1 + C1' + Tm and so inferring the one-way speed of light by halving the two-way speed of light would still not be correct.
23:15 isn't that pretty close to what gravity wave experiments are trying to measure? Length contraction vs speed of light seem similar here
Around the year 2002 an interesting star was observed. It's V838 Monocerotis star which is surrounded by a thin nebula. After its "eruption" during which it brightened considerably, "an echo of light" on the nebula was observed. Since it expanded equally in all direction it is hard for me to imagine how the speed of light could be different depending on the direction.
Also, in 1676 Rømer roughly measured the time the light takes to travel the Earth's orbit diameter and as far as I can understand his method, it's a one way determination of the speed of light which is the same, regardless of the Earth-Jupiter orientation in space (granted, only along roughly the plane of the ecliptic but I fail to see how that particular plane would be in any way special in the cosmic scale).
What you observe in the case of V838 would really depend on the direction of the anisotropy in the speed of light. Also, it would be hard to measure the difference in opposing directions if it is small compared to the speed of light itself. Don't ge me wrong though, I'm not advocating that the speed of light should be different, just that it is hard to prove that it can't be different.
If you used a set of springs to dampen a braking force again and again, they might not perform similarly when damping acceleration. So a system of real springs, initially with isotropic spring constants, can become directionally anisotropic in the way you described when it is used more often in one direction than the other, because of internal stress, entropy, etc, in the spring structure. Just an analogy, but could it be possible that the vacuum has a tiny anisotropy in the radial direction towards our galactic scale acceleration towards the great attractor, because all our mass is usually accelerating that way more often than not?
Effect on energy would be observable and measurable?
I will repost my comment from Veritasium's video
HOW TO MEASURE SPEED OF LIGHT
We can measure the time it takes light to go from point A to B in vaccum, bounce of the mirror and go back from B to A, BUT inside a glass
Then repeat the experiment but in the opposite way. If the speed of light decreases by the same % going from vacuum to glass regardless of the direction, we should be able to see the difference in travel time.
Furthermore, we can make 2 additional experiments (both ways in vaccum and both ways in glass), so that we have 4 measurements of time and 4 unknown velocities - easy to solve equations
Great video btw :D
How are we defining speed? Don't we have to agree on what time it is at two different locations is in order to talk about speed at all? And to synchronize clocks at a distance, we need the speed of light in both directions. Is the question meaningful in the first place?
wait isnt this difference in propagation speed caused by local gravitational fluctuations the working principe behind a gravitational wave detector?
I might be misunderstood but that sounds to me like we can measure it
Yes but because it is based on interference, always in a two-way way.
I have an idea of how it's theoretically possible to measure the one-way speed of light and I don't think it has any problems.
You would send a light wave and start a clock at the same place and time. The light wave will reach a machine you have set up a certain distance away. The machine then sends some other type of wave back, e.g. a sound wave, through a carefully controlled medium, when this wave reaches the clock, the clock stops. You could use the measured time to calulate the one-way speed of light, accounting for the time the sound? wave would take to reach the clock.
Can someone tell me if (and why) this wouldn't work? If it would work, I wonder why I haven't heard of this before.
Armand Hyppolyte Fizeau is one of the best names I've ever heard, holy crap
P.S. how do you pronounce "Huygens"? Is part of that silent like the Vietnamese Nguyen (or are my white eyes just seeing the same letters and assuming a connection lol)?
kind of like HOY-guns
When moving through a medium, like with a speaker and microphone that is not stationary in air, the one-way speeds can be different relative to your apparatus (but consistent relative to the medium, in the case of air).
We now know that there is no aether and the speed of light is independent of the frame of reference, but this had to be established.
So I would think the one-way speed of light has more to do with moving reference frames than with anisotropic physics.
There is another way to have a difference in the opposite direction speeds. You can check the last couple videos of Dialect