Why No One Has Measured The Speed Of Light

Birt 31 okt 2020
Physics students learn the speed of light, c, is the same for all inertial observers but no one has ever actually measured it in one direction. Thanks to Kiwico for sponsoring this video. For 50% off your first month of any crate, go to kiwico.com/veritasium50
Huge thanks to Destin from Smarter Every Day for always being open and willing to engage in new ideas. If you haven't subscribed already, what are you waiting for: ve42.co/SED
For an overview of the one-way speed of light check out the wiki page: ve42.co/wiki1way
The script was written in consultation with subject matter experts:
Prof. Geraint Lewis, University of Sydney ve42.co/gfl
Prof. Emeritus Allen Janis, University of Pittsburgh
Prof. Clifford M. Will, University of Florida ve42.co/cmw
The stuff that's correct is theirs. Any errors are mine.
Einstein, A. (1905). On the electrodynamics of moving bodies. Annalen der physik, 17(10), 891-921.
(English) ve42.co/E1905 (German) ve42.co/G1905
Greaves, E. D., Rodríguez, A. M., & Ruiz-Camacho, J. (2009). A one-way speed of light experiment. American Journal of Physics, 77(10), 894-896. ve42.co/Greaves09
Response to Greaves et al. paper - arxiv.org/abs/0911.3616
Finkelstein, J. (2009). One-way speed of light?. arXiv, arXiv-0911.
The Philosophy of Space and Time - Reichenbach, H. (2012). Courier Corporation.
Anderson, R., Vetharaniam, I., & Stedman, G. E. (1998). Conventionality of synchronisation, gauge dependence and test theories of relativity. Physics reports, 295(3-4), 93-180. ve42.co/Anderson98
A review article about simultaneity - Janis, Allen, "Conventionality of Simultaneity", The Stanford Encyclopedia of Philosophy (Fall 2018 Edition), Edward N. Zalta (ed.) ve42.co/janis
Will, C. M. (1992). Clock synchronization and isotropy of the one-way speed of light. Physical Review D, 45(2), 403. ve42.co/Will92
Zhang, Y. Z. (1995). Test theories of special relativity. General Relativity and Gravitation, 27(5), 475-493. ve42.co/Zhang95
Mansouri, R., & Sexl, R. U. (1977). A test theory of special relativity: I. Simultaneity and clock synchronization. General relativity and Gravitation, 8(7), 497-513. ve42.co/Sexl
Research and writing by Derek Muller and Petr Lebedev
Animations by Ivàn Tello
VFX, music, and space animations by Jonny Hyman
Filmed by Raquel Nuno
Special thanks for reviewing earlier drafts of this video to:
Dominic Walliman, Domain of Science: ve42.co/DoS
Henry Reich, Minutephysics: ve42.co/MP
My Patreon supporters
Additional music from epidemicsound.com "Observations 2"


  • i think answer is not getting stuck on synching the clocks idea at all. all you need is properly recording signal output and signal recieved times in two seperate clocks. later you can find the one-way light speed by getting these clocks together and analysing time difference between them.

  • And here all this time I thought somebody was smart enough to use a mirror and then use a highly precise atomic clock to measure the light going away and then the amount of time it would take to come back as one complete unit of measure. After all the mirror we have on the moon is how we measure the distance using a laser and the amount of time it takes for laser to go from the Earth bounce off the mirror on the moon and then return. If a suggested late had a preferred direction then such experiments as CERN which shoot photons into a loop would not work.

  • Here's an idea for y'all. Let's say that light coming from one direction of the universe traveled to earth at 0.75 c and light from the opposite direction traveled at 1.25 c. Also lets consider that when we look back to the most distant objects in the universe they appear as they were some millions of years ago. Would this not mean that we could infer the existence of this directional light speed difference from observing systematic differences when looking at distant objects in these opposite regions of space? Perhaps we could find more indicators of an early universe in one region 50 million light years way as compared to the polar opposite region. Or perhaps the irregularities found in the CMB map could also be an indicator of this.

  • Thank you very much for this very interesting video. I work on studying light and I didn't even know this! Blow my mind ! Btw, I don't know if you would be interested, but we published a work where at we consider a sequence of two light pulses at the input of a nonlinear system carrying information. At the output, whereas both pulses have experienced exactly the same optical path, there is the creation of a third pulse at the central delay that carries the information of both pulses. This means that the information carried by the last pulse arrives at the output before the pulse itself !!! Honestly, before the experimental demonstration we performed, we were just scratching our heads thinking "it is what the equations tell us, but how on earth this really occurs?". Fascinating :D If you are interested, here is the publication "Lassonde, Philippe, et al. "Information transfer via temporal convolution in nonlinear optics." Scientific reports 10.1 (2020): 1-8." Cheers

  • how about you ACTUALLY cook an egg Einstein

  • can't we use atomic clocks ?

  • What if you start with 2 clocks in the Middel move them do the test en move them back to the middle

  • Sounds like a political activist arguing about science🦖

  • Why are people surprised by this? We don't even have a real picture of the Earth from space.

  • Don't we use 'redshift' to understand the constant?

  • Four Clocks moved from the center in East, West, North and South, and four rays of light shot from each direction. The difference in time dilation in the East-West axis would not be visible in the North-South axis and vice versa.

  • Couldn't you use a system that utilizes a qbit(Entangled Pair) as the message transmission tool. Entangled pairs appear to communicate instantaneously over any distance so you have one of the entangled pairs at the detector which flips on detection of the light source. This then resolves the other member of the pair which acts as the "stop" button for the clock. Has this been tried?

  • Wouldn't universe in some directions look significantly younger than in other directions if the speed of light was influenced by direction?

  • I'm not an expert. god forbids. but what about this experiment: at the start we send, at regular intervals of 1 second (or whatever time you agree), pulses/signals/whatever not only at speed of light but at a known, and verified, speed. for example each second is sent a signal at 20,000,000 m/s, simultaneously we also send another signal at 30,000,000 m/s, another at 50,000,000 m/s and another at the speed of light. at the destination, we start counting the signals as soon as we get simultaneously one of each and we stop the next time all pulses arrive at the same time. now we can count how many light speed signals we get between the start and the end of the period of time we measure at destination. there's no need to use mirrors or sincronized clocks. note: we use multiple non-light signals for the sake of redundance. the more signals the better. so long as they are all send simultaneously and periodically.

  • If we ever make the Alcubierre-drive (aka. a Warp Drive, which moves an object along space by moving space itself, which (according to theorists) would permit true FTL-travel) a reality, we could send a signal through the unwarped space, reach the signal’s destination (trying not to accidentally hit and destroy the signal) before its expected arrival (taking with us a clock ticking at precisely 1 second per second), then see if the signal arrives at that point and when. If, in any direction, the signals would take exactly the expected time, then we can be certain that the one-way speed of light is a single constant in all measurable directions. If, however, the signal takes *different* amounts of time depending on direction, we can be certain that the one-way speed of light is a variable, based on the exact angle that the light travels. And *if* the signal never arrives (which means we missed the signal), we can be equally certain that the one-way speed of light in *that* particular direction is greater than the average speed that our Warp Drive-powered spaceship flew at. This obviously assumes that the Alcubierre-drive completely isolates us from the unwarped space in such a way as to make time dilation non-existant within the space bubble (such that time will flow at the same absolute speed regardless of the “relative” speed of the spaceship itself), which may still be up for debate. I believe that *if* the one-way speed of light is a variable, it’s because of the constant (and accelerating) expansion of the universe. In that case, I believe that the difference will be minor (but still measurable).

  • I have a solution how to prove that it is same speed at both directions (or disprove), by using calculated speed of light (average from 2-way) you can check if its the same on one way and another, by having at first end a light emitter and light detector, but on the over side have some kind of device which could only reflect light one time per time which light travels one way ("average speed of light on one-way") the distance you made beetween device and light emiting source. Main problem would be synhcronising it, but after that, if you detect light on detector, that means light is same speed in any direction

  • As a thought experiment you could have a force scale that is infinitely sensitive in a steral vacume in a room that is infinitely dark

  • So what I'm getting from this is technically light could be a finite speed or infant

  • If I were to vote for the best video made on ISchats. This one would be it.

  • 10:01 if we find out that the clocks are not synchronized then we know that the speed of light is not the same in all directions. and then we can find the ratios of both the speed with the help of the difference in time in the clocks and then equate the ratios with the speed of light and find out the speed of light in all directions. I am not sure if this is right.. just an idea.

  • I'm not a scientist, I don't even have a degree in anything... but here is a question from layman's point of view: If light travelled at different speed depending on the direction to which it's travelling, wouldn't that be observed from other observations, eg the expansion of the universe, or in measuring the distance between us and other stars beyond our solar system?

  • How about using timmer in both clock on same time

  • What if you stood two individuals in such a way that the sunrise was exactly in between them. If light travels at different speeds in different directions, wouldn't one individual see the sunrise before the other? If we say light travels at c/2 in one direction and instantaneously in the opposite direction, wouldn't one individual see the sunrise approximately 9 minutes before the other? Even though the light from the sunrise that reaches the individuals is not completely opposite (imagine it like a triangle with the sun at the top and the two individuals make the two bottom corners) wouldn't there still be some difference in speed since the light each individual sees follows a unique path to the other?

  • 2:45 how about move those two clocks opposite direction exactly 500m each, so that they are 1km apart, and still in sync?

  • Another convention: That light always travels at the same speed in one direction.

  • what if you moved two clocks in opposite directions and had 2 people stop them after motion has ceased when they reached a given time as they were still on the machine that moves the clocks further apart. then you have 2 people waiting at the end of the given distance to record the time that their clocks reached that given time which can only be reached after they have stopped motion.

  • What would be the problem with having the light trigger other sources of light to begin as it moved towards the end point? For example, if we were "shooting" the light 100m, then we would have clocks stationed every 10m along the path that were capable of "shooting light" as well. When our original photon hit one of those clocks, it would trigger and shoot light back at the starting point. Would that not give us the one-way speed by adding up the time from all those secondary shots? What would keep that setup from working?

  • That would be pretty unusual if you shine a light in one direction it travels a different speed than if you shine it in another...

  • I bet Rupert Sheldrake likes this video

  • just use a ruler lmao

  • What if you were to take two quantum entangled particles and in some way have the light react with one particle and then the other at a distance, the state from one would affect the other and then the light would ultimately hit the second particle changing both particles state I suck at explaining and please correct me if I am wrong or got anything confused or wrong lol but light hits particle A changes to A+ and changes particle B to B- and then light travels hits B- turns it to B+ turning A+ to A- does this work can entangled particles “spin” be changed ? And if I am wrong please tell me I know very little if anything at all about this but I try lol this would be the measure of one way light speed wouldn’t it ? but can quantum particles be used in such a manner ?

  • Could you not use quantum entangled clocks?

  • Can't we record the one way and 2 way speeds of light through denser mediums (i.e. water and other substances) and determine whether light travels at a different speed on its way back?

  • Can you just shoot 2 light guns and see where they hit???

  • Wait, what about blackholes? Can't you see the back of your head because the light travels around it? Can't you measure it that way?

  • What if you send the 2nd clock out and back 1km, divide the difference in time of the 2 clocks by 2 then preemptively change the time on the 2nd clock so that when you send it out 1km it can match up with the 1st clock.

  • Sinchronise both clocks in the center in the center. Move them back at the same time, same speed and them shoot each other a measure how fast the opposite shoot reach each other.

  • What if you star in the center, move both clocks from the center back , at the same distance, same speed same same hight and then shoot each other, and then measure how fast the opposite shoots reach each other.

  • The inference that the return message from earth is absurd. The Information is inherently different and for it to return to earth instantly implies the violation of the speed.

  • in your experiment (12:34 until 14 :15) were you claim various directions you are actualy experimenting various distances wich by the way confirms Einsteins theory that light travels the same in all directions but riches differently in verious points in space

  • The Apollo space craft did not travel instantly from the moon back to earth!

  • What if you set a clock and at exactly 10 seconds you have it go at a consistent speed from point A to point B and at the exact second it gets to point B you stop the clock and then do the exact same in reverse.

  • I swear this is one of the most interesting video I've ever seen

  • Light from a specific type of star at a specific distance is the same in all directions. Twice as far it is still the same in all directions. The speed of light HAS been measured very closely.

    • And the difference between the two is also the same in all directions.

  • It was just a little cringe how you looked away from the camera at the end. I like you videos.

  • Using light to measure the speed of light feels like using a ruler to measure the length of a ruler and concluding that an inch does in fact equal an inch… my brain hurts 😂

  • Why doesn’t someone just test it all the ways you can to find out and in different directions

  • If it is true that light travels at a different speed from different directions, there should be applications where you can prove this hypothesis?

  • I think light would travel differently in darkness than in light/daylight!

  • doesnt black hole would be bright if so?

  • Wait what about redshifting I know I'm perfectly hundredth person to say this but what about red and blue shifting?

  • The speed of Light can Not be Instantaneous Because If it instantly Going from point B to Point A It is so called Infinite Speed And If its instant it could be possible that it would seem like it has stretched into infinity so it is physically impossible to know how fast light is going

  • It's like trying to measure the length of an infinitely long wall with a finitely long measuring tape.

  • If this were true, the universe would look younger in one direction and older in the opposite direction. It would not be a uniform looking universe.

  • You could align clocks directly connected to each other for the whole 1 km distance, no matter how many clocks it takes to fill the space. All the clocks would then be synchronized, you would shoot light in one direction, each clock capturing the time when light hits it only once. You will then have a progression of time stamps, that data could be used to measure the speed of light in one direction.

  • Use 3 mirrors and measure the 3 way speed of light using different types of triangles.

  • Nice quality video. I didn't know this. What if your consciousness is light & everything radiates away/forward from you in the form of fluctuations & functions collapses???

  • this is just me, but what if you connect the two clocks at each end with a Y connection switch, where the lines are exactly the same length from the clocks, and the trigger for the light is connected directly to the first clock. In other words both clocks and light all start at the same time, then its just a matter of when the light hits the 2nd clock to indicate the stop time.

  • I’m no physicist by any stretch of the imagination but maybe it’s assumed that the speed of light is how it’s currently measured not because it’s easier but because it’s an accepted measure of the numerical value we currently have. Maybe your right and it’s not 100 percent dead on but if no other value exists then what else do you have?

  • But if the light from the universe was instantanious the formations we see in the deep universe wouldnt make sense. Gasses would have time to concense into stars and planets

  • Have we actually mesured the one way speed of anything ?

  • But why we are the ones that see the light instantly?

  • Quantum entanglement, why not use it?

  • So, if the moving clock vs the stationary clock is an issue, move both clocks in different directions for the same subjective time and distance. This should keep the clocks synchronized. If you feel that somehow the preferred direction affects this, move the clocks at right angles. Do left and right handed tests and compare the results to ensure you don't have more preferred directions. Doesn't seem that complicated to me.

    • Speaking of, that is pretty much how Ligo works.

    • Oh, and after an agreed upon duration, move them back together again to compare the results.

  • what if we look at a star before it becomes supernova and record the time it became supernova and after supernova we can look at the ruins of it and calculate how long ago it became supernova ...... i may have described very bad

  • All of a sudden my brain hurts.

  • what about using an object so massive that you measure a one-way trip that simply bends in circle? you just need a black hole :)

  • I haven't read if anyone else has commented this before. But light had assymetric speed during its travel in vacuum you would be able to note CE frequency difference between say "leaving direction" and "going back direction". Meaning you would see interference pattern in a light measured on a "mirror bounce" riig... So it is excluding that the light has a assymetric speed in different directiond

  • No problems.... 1)(a1+b1+a2)/2 (2km.) constant 1905 ears. Mi version 1) distant #1) a1+c1(=1Km) sinhronisation a1 end c1 =(t1).... distant #2) a1+b1+c1 (=2km) (a1+b1+c1(2Km))-(a1+c1(1Km))=(a1+b1+a2)/2 ??????

  • What if we put put the clocks 1 km apart and hang a rod from the middle of length 1 km (just one point in the middle) and the clocks at the end. Then we release the middle joint and the rod falls on both of the clocks at the exact same time and starting the timer? (Just hypothetical)

  • Mind blown

  • Wow

  • Is it possible to understand the speed of light by realising the time lag between the clocks if we try to synchronise them from the middle...?we will eventually know if the speeds are different and if so by how much! PS:correct me if I m wrong If right contact me thro insta @abhinav_08

  • How about this: We calculate a path for a rocket to take from Earth to Mars. We know exactly how long it should take and even put a person and a supercomputer on the rocket to guarantee that the plan was followed exactly. As soon as they touch the ground, they send a message to Earth saying they have landed. Of course we could never measure this so accurately and be able to account for the effect of special relativity but this would at least guarantee that light speed isn't more than a few percent off what we measure it as. Of course, that only covers the exact path that the light used to communicate with us took to come back to us, any other path could theoretically be different.

  • What if you could observe two clocks from far enough away to see both in the same reference frame?

  • This absolutely broke my brain🤯🤯

  • Why do the clocks need to be synched? Just looking for elapsed time, not absolute time.

    • Never mind, that was really stupid.

  • If the speed of light was different in both directions wouldn't our light horizon also be different in both directions? For example if it was infinite in one direction our light horizon would also be infinite in that direction and we would be able to see stars beyond our "light horizon" and therefore we would see substantially more stars in the infinite direction vs the other? IDK not a physicist.

  • How about building the clocks on site and synching with them with something like a pulsar 10 gazillion light years away.

  • What if you used two phones that is set to the exact same time zone and then go offline when they get into position?

  • What an AMAZING video! Man, this video truly inspired me! Thank you a LOT!

  • what if you set up three clocks in a row to be synchronized with both the middle clock and first clock with 2 stoppers with there being a mirror behind the third to reflect the light beam back and the middle one would have the beam pass through therefore acting as something the be compared with therefore the time it takes to hit the middle clock the first time from the middle the end should be the same time as the reflected light back through the middle and to the first? or would that still just be measuring two way light?

  • If we assume different speed of light values for the "forward" and the "backwards" direction with respect to a certain direction, we should measure different 2-way-speed-of-light values in directions which are perpendicular to the first measurement, shouldn't we? The one way speed of light would be a property of space, kind of a "vector field": For each point in space this field would specify the one way speed of light for each travelling direction light passing this point may have. As far as I can imagine, there is only one configuration of this "vector field" that allows different forward and backward speeds not to be detectable by different measurement directions, this is if this field is spherical-symmetrical centered at the observer, hence always changing with the observer. Even if we assume that this is possible, doing the same experiment (measure the 2-way-speed-of-light in perpendicular directions) at different locations at the same time would unmask the configuration. Okay, there is the "at the same time" problem again .... Anyway: I believe in Occam's razor!

  • Ok wait doesn't this theory of instantaneous light transmission basically mean that objects beyond the cosmic horizon can send their light to us? Doesn't this make the definition of the cosmic horizon obsolete?

  • By measuring the delta between the speed of light and the speed of sound, we'd get different values for each different direction if c varies.

  • A way that u could prove this is to have the person on Mars travel back to earth without retuning the clock and see what the time desync would be (we don’t have to worry about time dialation beacause according to you the speed of light traveling to earth is infinity so thus we don’t need to account for time dialation)

  • You may do a one way measurement at a black hole's event horizon where light can travel around the hole in a circle. The tricky part would be to share the results from there though...

  • If there is no way to measure the one-way speed of light, how is it that distances between objects in space are accurately calculated?

  • My simple question is this; Can there be light without a heat source? If not then what should we be measuring the speed of; LIGHT or HEAT?

  • 4:06 haah that VSauce Vibe

    • i was finding this comment.

  • Our current understanding of the universe is that we look back in time when we observe the light from stars. What are the implications to our understanding of the Big Bang and the expansion of the universe if the light we see is of the instantaneous kind in our direction and slower on the return? That somehow makes me feel like earth is the center of the universe again. I am just trying to ponder on what you’ve taught. If we don’t go with the old conventional thinking, what else would we discover has also been assumed?

  • What about the doppler shift in waves reaching us from all sides?

  • I was watching this video and thought how about checking the mass of light with the atmosphere since the atmosphere bends light. see how much force the atmosphere is causing upon the light and determine a speed that would correlate with the mass or the range of deformation to the atmosphere

  • 6:10, if you are wondering, based on the look, the microphone the guy was using is the blue yeti.

  • Set up a 'sensor' and the points of origin and reception in an Isosceles triangle. The source emits a laser signal pointed at the sensor when the light is fired, and the receiver emits a signal pointed at the sensor when the light is received. The speed that light travels is measured by the difference in the two signals that the sensor receives. Since it is equal distance from both the source and the receiver, the 'delay' in receiving the signals is irrelevant. This works in both directions! OF COURSE we can measure the speed of light!

  • I am not strong in science, but I remember an experiment was carried out with a photon and its 2 states, as a particle and as a wave, and if you do not take measurements, then it behaves like a wave, and when a measurement occurs, then it changes its shape to a particle in the same moment as the freeze occurred and it was called time travel, but what if this is the inverse speed of light?

  • Я не силен в науке, но я помню был проведен эксперимент с фотоном и его 2мя состояниями, как частица и как волна, и если не делать замеры то он ведёт себя как волна, а когда происходит замер, то он меняет свою форму на частицу в тот же момент как произошёл замер и это назвали путешествием во времени, но что если это и есть обратная скорость света?

  • How about this: look at the universe in all directions. If light from one side arrives at a different speed, then, as you say in the video, you see a "younger" universe on one side than on the other. If the universe has the same age in all directions, then light speed should be the same in all directions.

  • What if you blocked light from the sun heading to the earth and then unblocked the light and time it then do the same at opposite times of the year so as to get light traveling in both directions and compare them to measure the margins of difference?

  • but if you say that the speed of light is infinite/faster in one direction shouldn't the observable universe be shaped like an ellipse or an infinitely long sausage-ish thing? and we could see a bigger number of galaxies and more developed galaxies in only one direction. if this theory is correct we would be seeing a much brighter part of the sky with a red tone and the other part will be twice less bright than now.