One Way Speed of Light

http://cr4.globalspec.com/blog/22/Relativity-and-Cosmology

Regards JD.

That was a quick comment, but most likely you did not follow my idea. Unlike all other attempts to synchronize clocks by the signal from A to B , I'm "rewinding time" to the instant when signal was emitted from point A and I am adjusting the time to that which will indicate the clock B when the signal from A arrive.So if we really could turn back time, the clock A (corrected) at the moment the signal was emitted would indicate t0 and clock B would show t0 -time the photon need to travel from A to B

From what I understand of your proposal, you are making three assumptions that you do not properly address:

1. The precise distance between A and B is known.

2. That the instant in time the time reference (single?) photon is transmitted and received will be precisely known with no uncertain variation from any instrument.

3. That the point of this mental experiment is to obtain absolutely precise time marks at point A and B. IMHO the point of this experiment is to show simultaneously that absolute precision is a myth but the uncertainty can be statistically known to an acceptable level.

It seems to me that the process of sending the information about B's clock back to A counts as the return trip. After all, one way of sending the information about B's clock back to A is to send a light beam from B to A. So a round-trip is still required.

Sorry, the post should be as follow:

Some people say it is impossible to synchronize the clocks to measure one way speed of light. That would be true if one would use Einstein synchronization of the clocks.

What if we use back synchronization as follow:

Let's start with point A and B separated by known distance d. At time tx as indicated on clock A we will send a signal (photon) from A to B.Let's assume the signal will arrive at B at time t0. Now we can correct the clock at A by substituting t0-d/c for tx (so when the clock at A was ticking and indicated tx+nt, it now will indicate t0-d/c+nt).

Now if we send a signal from A to B at time t1, the clock at B will indicate t1 if the speed of light is isotropic and equals c in both directions

Can anyone point an error in my reasoning?

Lorentz contraction.

You have totally confused me.

1. Send a photon from A to B and back sync your clock based on clock B.

2. Send another photon from A to B and that proves isotropic behavior?

How?

Did you mean send a photon back from B to A instead?

that only works for daylight savings applications

I think you are right. You can't synchronize two clocks using one of them as a reference. So let' modify our experiment:

Let's have 2 points A and B and point C exactly in the middle. We can indirectly synchronize the clocks at A and B by sending the signals to C and calculating back the time on A and B clocks so the signals from A and B would arrive at C at the same time. Let's assume temporarily that the speed of light is the same in both directions and equals c.

We can now synchronize the times at A and B with C knowing that at the moment signals were send towards C (assuming the signals arrived at C at the same time) the time at A and B should be same as time at C minus travel time of the photon from A (or B) to C. We assume this time to be the same in both directions and equals distance divided by c

Now we can send the signal from A and B towards C at the same time as indicated on clocks A and B. If the speed of light is indeed same in both directions, both signals will arrive at C at exactly the same time. If there were any differences, we could calculate speed of light in either direction.

You now have light traveling in two directions.

I still think this is like using a ruler to check if the same ruler is accurate.

This is because you are using a defined constant, C, to calculate the speed of light.

The results will always agree 100% because you are not actually measuring anything.

How does one at B know precisely the time difference between t0 and t1 at A? There must be communication from B back to A. This makes it always two directions, unless I still misunderstand your thought experiment.

I think the issue is that he knows the distance between A & B and then he uses the speed of light to determine the transit time of the photon.

The rub is using a defined value for C and then calculating it is bound to yield the exact same answer for C. It can be no other way.

However, are the clocks actually synchronized? The answer is no.

So far, the only way you can actually sync clocks is to have them at the same location, then slow transport one clock ( that would be by USPS down here) at a velocity that is so slow as to not compound the experiment to a known distance away and run the experiment using the two clocks.

Put another way, clock A now has a built in error. It that synchronization? Not if you move clock B.

Movement is not a problem if the velocity is slow enough that any resulting clock error is smaller than the precision by which you choose to measure.

So, if I move it slowly to the Moon, without introducing a new error, it won't matter?

"So, if I move it slowly to the Moon, without introducing a new error, it won't matter?"

In a word NO.

You still have the problem of time dilation due to the proximity to the different masses of the Earth and Moon. The end result is that it cause the clocks to run at different speeds.

No?

Yes!

Is the speed of light different on the moon than on the earth?

"Is the speed of light different on the moon than on the earth?"

No the speed of light in a vacuum is always the same regardless of where you are, which way you or how fast you or the source are moving. However, the clocks on the Moon and Earth will tick over at different rates due to the different masses of the Earth and Moon which distort space time differently and as a result the way time passes. It's called time dilation and if you follow the link it will explain it better than I can.

Who said the clocks tick?

If one is discussing the movement of photons, at the speed of light, would one use an old fashioned mechanical clock that ticks? Not I.

I don't believe the type of clock is specified as one that ticks, in the classic sense.

Are you suggesting that crystals oscillate at different rates on the earth and moon?

Yes, they do.

The passage of time will be different for a clock on the Moon versus one on the Earth.

GPS satellites experience the same phenomena due to a combination of lower gravity in orbit (runs faster) and orbital velocity (runs slower) relative to "stationary" clocks on Earth. The net effect is significant for GPS accuracy and those clocks are compensated for that shift before they are launched. The effect of orbital velocity is greater than the effect of less gravity for those satellites.

That being said, the observed speed of light (in a vacuum) will always be the same for an observer at any point in the universe regardless of gravity or velocity.

So, in the condition where clocks that are stationary relative to each other, but exist in different gravitational locals, the clocks will de-sync over time.

Applying compensation to one of the two clocks will not work in this case because for the observer at the compensated clock time passage will not agree with that clock, therefore any temporal based measurement will be different than what it would be for the observer at the uncompensated clock.

This is the crux of this real paradox and why the "one way" speed of light cannot be measured. A critical part of this paradox is most experiments and analysis permit a loose use of time variable nomenclature in our mathematics. This experiment is not one of them.

In the equation the OP refers to there should be several time measurements symbols (tA[0], tB[0], tA[n], tB[n]) for the time readings at each clock and a few Δt in the equations to show when a comparison is made at each clock. In most time based measurements of phenomena there will not be a critical difference between tA an tB so one can simply assume tA=tB and thus tA[0]=0; tB[0]=0; t[0]=0 is a valid assignment for the start of an experiment. This makes the convention of ∆t=t[n]-t[0] into the much easier to write ∆t=t[n].

If clock "A" and clock "B" are actually synchronous during the measurement then one can apparently make a "one way" measurement of the speed of light. However, this stipulation of synchronous clocks simply hides the second path in a two way measurement. The only way one knows that the two clocks are synchronous is by comparing the values of each clock with the other in their respected time reference. Communication must be established in both directions thus a "two way" measurement has already happened and the desired measurement of a "one way" measurement has been prohibited.

I applaud the OP for tackling this paradox. This is a very subtle paradox in the complexity of relativistic Physics. Many people do not grasp the paradox. Most who properly appreciate the paradox gain a better understanding of relativity by trying to find a way to solve it.

"Are you suggesting that crystals oscillate at different rates on the earth and moon?"

I'm not suggesting that they oscillate at different rates, I'm stating they oscillate at different rates. To a person on the moon they will appear to oscillate at exactly the same frequency as they always have, however if you compare them to identical crystals back on earth they will oscillate at a different frequency. It's all about relativity and how the progression of time is relative to the specific conditions at a given point in space time.

Time dilation is a huge problem when it comes to the accuracy of time measurement that is needed for GPS systems because the clocks in the GPS satellites run at a different rate to identical clocks on Earth and as a result they have to reset the clocks in the satellites on a regular basis so that the error between the network of clocks is kept within an acceptable limit.

Well, I've studied this and have to concede that both you, Anonymous Hero and others are smarter than I and therefore I must be unable to come to grips with this and will just have to accept it as one of the great mysteries of life.

So, I'll not be moving slowly to the Moon after all.

"others are smarter than I and therefore I must be unable to come to grips with this"

I wouldn't say that, we all have areas of expertise where we excel. Besides the whole relativity thing sounds like it a right barking mad idea and even when you've read up on it extensible getting your had around the whole concept is bloody near impossible. As to the mathematics behind the proof I think there are very few people on Earth that can say that they understand it completely. Besides, relativity is a work in progress theory not a law and may be wrong, although every time scientists test it experimentally it appears to be spot on.

So don't worry about not coming to grips with it, very few really do. If you do want to understand it then I suggest reading Relativity 4 Engineers by our very own Jorrie.

Yes, jdretired's link in post #1 lead me there.

It's over my head, but so is a lot of other stuff.

Since I have local, much less exotic issues, to consume my energies, I'll just accept it.

Cheers.

Well, we will miss you while you are away, so I wouldn't say it won't matter.

Is there no CR4 on the Moon, then?

The 'problem' with is: 1) How do you know the two clocks are slowly moving apart? By looking at them as you move them apart. Thus you are using 2-way communication to ensure the movement is slow.

Or put another way: 2) How do you know the clocks maintain synchronization while they are moved apart? Same answer - by looking at them as you move them apart.

Or a 3rd way: 3) How do you know the exact distance between A and B? By making sure you know the positions P1 and P2 of A and B along the 'tape measure' when the time signal is sent; i.e., again, 2-way communication between the end points.

It's a theoretical problem and the slow moving clock is an answer to that problem, in theory.

In practice you really need to be away from any gravity sources, which sort excludes our current capacity, but moving clock B away from clock A at a known velocity is not beyond our capacity and measuring the elapsed time that clock B is in motion is also well within our capacity.

So, let's do a thought experiment. If you use some form of linear accelerator to put clock B into motion at a precise velocity and sync both clocks at the point where clock B reaches terminal velocity, then all you need to do is wait a precise preset time before firing a photon from point A towards the moving clock B.

Since we know the velocity and time of clock B we can take into account its exact point in space when the photon is fired and we will know its exact position when the photon is received.

Clock B will log the reception of the photon and time-stamp the event. Clock B can transmit the results of the experiment back to point A. The return trip for the data is not a factor in the experiment, so you really have logged the one-way speed of light with a reasonable degree of certainty.

The precision of the measurement relies on a number of factors, but if you can accurately transport clock B in a vacuum where the influence of gravity is negligible, there should not be any other technical hurdles that can't be surmounted.

Please reread Usbport's post. One way is the problem.

in a vacuum?

It would be easier to ask if anyone can point to a correct assumption in your reasoning.

Less work for us.

Can anyone point an error in my reasoning?

Yes, it takes time to transmit time tx from B to A, so you are measuring the round trip speed of light, assuming you transmit time tx back to A at the speed of light. You are trying to synchronize the two clocks at a distance, so the return speed of light is a factor.

Actually, return speed is not important. What you need is information at what time signal has arrived at B. You can convey this information even by telephone or mail.As long as you have recorded at what time at A signal was emitted, knowing the distance between A and B you can synchronize clock A with assumption that the speed of light is direction independent. Now if you send another signal at agreed time t1, you can verify if indeed the one way speed of light is the same as 2 way speed.If it is, you will have perfectly synchronized clocks. If not, you can calculate one way speed of light.

What you have described is the standard Einstein synchrony. This cannot be used to determine the one-way speed of light, but as you said, just to synchronize clocks.

-J

I still lean toward utilizing three standard clocks. They each communicate via wireless network through computer. Once the clocks are set and phase-locked, i.e. whatever the waveform is, it remains at phase angle 0 for all three stationary clocks. Z remains at base, A and B move off with photon event detectors (ultra-fast laser pulse train most likely. Each clock will record with its own photon detector, the exact count and phase angle each pulse packet is received, along with intensity information (if important later), each in its own time frame. No greatly relativistic speed of separation is allowed, but one must define "greatly" at some later point. Once the A and B stations move off, acceleration is stopped, and constant speed measurements are made. A and B decelerate to zero relative to Z, then another series of measurements is made (the lion's share). Then as A and B return to Z at constant velocity, a third and final set of measurements is made. All three sets of data should have some particular meaning to the final result. Once clocks A and B have returned, they are compared in count and phase angle with Z clock once again, and expected to be still in perfect synchrony (within the limits of relativity).

No, not really It is a subtle change. You synchronize the clock A with the assumption that one directional speed of light is the same as two way speed of light.Than you send another signal to verify it. If the one directional speed of light is the same as 2 way speed of light, the synchronization will be equivalent to Einstein's synchronization. If not, the clock B will show the difference.

kkris1: you are making a mistake of logic here. Once you have synchronized clocks A and B using the assumption that one-way speed is the same as the known two-way speed, there is no possibility that the second signal can throw up another sync or a different speed. Within the realm of experimental errors, you have forced it to be the same.

Bottom line, you can never measure the one-way speed of light with your method.

-J

I would like to have an expert answer a couple of troubling questions:

(1) if space has expanded since the beginning when there was light, supposing the speed of light to be constant, is it possible the early universe was extremely relativistic in speed as well, and has slowed to the present condition? Did space expand faster than the speed of light, indeed space is nothing, and if it did, what makes it expand, and is this exploitable?

(2) if the speed of light were non-isotropic with respect to spatial orientation, what decides the direction of the anisotropy? Also if anisotropic, what effect does that have on the map of the universe based on radio-astromomy?

GA

Great answer! Once again you have cleared up a few things for me.

When you reference to intrinsic spacetime curvature, is this as pertains to gravity wells only or something left over from the great singularity? If this is from the original singularity, then spacetime must continue to expand until it is no longer curved, i.e. "flat". Define flat, I can't, but in a some particular constraining limit I suppose one could.

If spacetime inflation is due to dark matter, how long before we can calculate further acceleration ad infinitum, or an eventual deceleration, or even contraction?

Cosmological spacetime curvature doesn't care about 'local' gravity wells, but rather about the large scale average curvature. One of the more difficult thing to fathom is that large scale space itself may be flat overall (so 'flat' that we cannot decide on which side of 'flat' it actually is), but spacetime is definitely not flat, because we are sure that large scale space is expanding. This means that parallel lines will not stay parallel, but will diverge over time.

Now, intrinsic spacetime curvature may have been imparted by the inflationary epoch, which could be considered as part of the Big Bang (which is an ill-defined concept anyway). It does not have to start as a singularity - in fact present day cosmologists are pretty sure it did not and could have been a "bounce", perhaps a very long period of pseudo stability at a density just above the Planck density, before starting to expand exponentially due to quantum effects.

If inflation did impart a negative overall spacetime curvature, then when expansion has 'thinned out' matter and radiation energy to a large degree, the intrinsic spacetime curvature will dominate, because it is not "thinned out". If this is so (and cosmologists are not absolutely sure, but it seems more and more to be the case), then it will 'look' just like Einstein's cosmological constant, i.e. a constant vacuum energy density. This provides a negative pressure that is apparently overwhelming the normal positive pressure of gravity today. Hence the present accelerated expansion.

Bottom line, such negative spacetime curvature may never disappear, but (for all we know) cause expansion to accelerate ad infinitum. Not the "Big Rip" though, just a gradual thinning out of matter until the whole present universe is essentially a vacuum again. But, this is speculating very far and wide...

-J

OK. You can say that is definitely far and wide.

I was reading elsewhere about mass-charge analogy, where E=Mφ, such that the potential energy is the mass charge applied in the gravitational field. My question here is why does mass only have a positive charge? Is there no matter anywhere that could ever have a negative mass charge? This is separate and distinct from, but somewhat analogous to electromagnetic charge.

Suppose this negative mass charge is stuck in the "middle" of the universe (why the separation of such charge, I have no idea). Could this replusion still be effective? Doubtful? Apparently like mass charges attract, and opposite mass charges repel?

I had enough problems with Physics 101, and Modern Physics 301.

James, I haven't heard of the mass-charge analogy before. Can you provide a reference that is not behind a pay wall?

I know about some electrical-mechanical analogies, but not that one.

-J

I did a couple searches and came up with these:

http://phys.org/news/2012-01-repulsive-gravity-alternative-dark-energy_1.html

http://resonance.is/wp-content/uploads/2013/05/1367405491-Haramein342013PRRI3363.pdf (I did not follow the concept of holographic radius, etc.)

http://nige.wordpress.com/quantum-gravity-and-the-mainstream-misunderstanding-of-electroweak-gauge-theory/ about gauge bosons, I think

http://www.sciforums.com/showthread.php?109287-Gravitational-Charge I think this is where I got the initial information

http://science1.nasa.gov/astrophysics/focus-areas/what-is-dark-energy/ the link Jorrie sent me where I went down the rabbit hole from

http://hubblesite.org/newscenter/archive/releases/2012/10/full/ galaxies collide! OH NO!! dark matter involved?

http://www.americanantigravity.com/news/space/eugene-podkletnov-on-antigravity.html This author claims a reduction of weight during test on item similar to "Searle effect generator" of up to 20%, either outstanding or a complete hoax? I don't know.

This is far beyond my training, so I basically just throwing this out there.

One flaw is using the measured quantity to determine the measurement, which I believe the Heisenberg Uncertainty Principle addresses. The act of determining the position of a particle affects it momentum, which in turn affects the measurement, which in turn affects its momentum, ad infinitum.

The other is saying "...it is impossible to synchronize the clocks..." when in fact the GPS system provides a very precise way of synchronizing two very distant unconnected timers/clocks.

"Now we can correct the clock at A by substituting t0-d/c for tx (so when the clock at A was ticking and indicated tx+nt, it now will indicate t0-d/c+nt)"

I think I know what you're trying to do and it is fundamentally flawed because in the calculations above that you are using to synchronise the clocks you are using the seed of light which is what you are trying to measure and is therefore a variable not a known factor so you can't use it to synchronise the clocks because you don't yet know its value.

At least that's the way I read what you are trying to do.

The error in your reasoning is that you assumed you had communicated clearly what you want to do and why.

How can you correct the clock at A. The Photon arrives at B. How do you know when it gets there. Any signal to say it arrived must make the return journey to you at A and this takes the same time as going to B from A. The information must come back to you before you can set the clock to the time it arrived at B. Hence it is not possible to use this method to measure the one way speed of light.

Ok, what your saying is clock A sends a signal (photon) 12:00:00 the speed of light (in Vacuum) takes 00:00:02 to cover the known distance d. Clock B is set 12:00:02 As it receives signal (photon) it starts running then both clocks should be showing/running at the same time. Is this correct?

Or what I think you are saying, clock A sends a signal (photon) 12:00:00 It takes 00:00:02 cover the distance d. Clock B is set at 12:00:00 it receives the signal (photon) and starts running. THEN clock A resets to 12:00:00 instead of showing 12:00:02.

Not withstanding time dilations, or if clock B received the signal correctly or any other issue with circuit propagation delays. Without some sort of feedback you would never know if either clock were synched together or were running at all.

Yes, you need to communicate back the time at which the signal has arrived at B. You can do it using the phone, skype or even walk back with the message. If you record the time at A at the moment you send the signal and your clock is still ticking, you can adjust the time at A to be (at the instant the signal was emitted) same as the time at B when signal arrived minus assumed travel time from A to B for the photon with the conventional 2 way speed of light =c.

Let's suppose you have emitted the signal when clock at A was showing 11:00:00.

At B the clock was showing (at the time signal has arrived) 12:00:02. You new that the light needed 4s to go from A to B and back. Somebody has rang you at 14:30:00 (on your clock at A) and told you that the signal has arrived at B at 12:00:02. Assuming that the speed of light is c in both directions, you adjusted the clock at A to be 15:30:00. Now if you send the signal from A at 16:00:00, the clock at B should show 16:00:02 if the speed of light does't depend on direction and equals c. If one direction speed of light is different to c, you will have different reading at B than 16:00:02.

To learn the distance "d", you would have to send a photon from A to A, via a mirror at B, and calculate the distance by multiplying C by the time of travel. All this before your experiment could start

Or a long Lufkin tape measure.

Here is another twist to the way of thinking about it.

From the perspective of a photon neither time nor distance exist.

For an observer traveling at C, time appears to stop. If time stops, distance is also meaningless.

Perhaps in a Newtonian Universe but not in an Einsteinian Universe.

Look at this a different way, one having nothing to do with clocks:

Assuming the speed of light were not isotropic, that its speed has a preferred direction, what implications would this have for the Universe as a whole? How would this manifest in terms of the value of α, the fine structure constant. What spectrographic evidence would we see as function of, say, the motions of galaxies moving directions parallel and perpendicular to this preferred direction? What would be the consequence of variations in the values of ξ0 and μ0 as a function of orientation, velocity and acceleration with respect to this preferred direction?

I can tell you one thing: if the speed of light were not isotropic we'd see a very different Universe altogether.

Instead of timing things with clocks, get out your Michaelson-Morley interferometer and see what it tells you.

Just to pose the purely theoretical possibility that a proper devil's advocate should....

If the speed of light preferred an orientation outside of our solar system, how would we know this? Might the deviations from "norm" of the COBE satellite observations be manifestations of how the speed of light in a vacuum does get distorted in a vacuum?

Did you hover your mouse over it? ...

"Wait till they notice the faint reflection of Michael Jordan and Bugs Bunny in the E-mode."

ROFL!

Sorry to be 'Johnny-come-lately', but some very interesting perspectives and questions here so far, e.g. Redfred's

"If the speed of light preferred an orientation outside of our solar system, how would we know this? Might the deviations from "norm" of the COBE satellite observations be manifestations of how the speed of light in a vacuum does get distorted in a vacuum?

Over large scales we actually know that light propagation is not isotropic and depends on the gravitational potential of the source and the observer respectively (and even the potential wells in-between). We can however not measure that anisotropy, because we do not have clocks synchronized at the two points. We simply believe Einstein on that.

Because of the observed large scale homogeneity and isotropy of matter (normal and dark), we make an assumption that the effects on light propagation largely cancel out for things like the CMB, when viewed from different directions. Then, because we can measure the (two-way) speed of light locally, we assume that it closely resembles the average speed of light over the whole large scale region and in all directions.

As Europium said, we have indications from cosmic observations that this is a reasonable assumption, but we cannot prove it to be so.

-J

Now that is some very heavy duty processing power.

You should notice that it simulates but does not emulate or implement a change in the speed of light. Then again if it did implement a change in speed, how would we know?

Sprint does it, at the extreme low end of the range. I know, because I called them just last week. They answered this afternoon, right after the third ring.

Alternate experiment: clocks and accelerometers good to 1 ppb (1/1,000,000,000). Clocks all synchronized in frequency and phase at time =0. Little or no local gravity (deep space). Three sub-sections separate at time 0, Z is static relative frame, A is left-handed travel, B is right-handed travel (180 degrees opposing), and acceleration is identical, constant, however velocity is not constant. A and B emit periodic signals of light that are linked in frequency and phase to the standard clocks aboard. Any relativistic time dilation is back calculated by computer a Z, using the data on "isotropic" speed of light gleaned by increasing time delay of reception of each signal. Information on the status of crafts A and B is coded in the light signals (modulation has no effect on experiment), can be a separate signal for purists. By using the laws of motion, and watch for phase drift between the actual signal arrivals, then after careful evaluation of transmitted operational data of each craft, and the data obtained, a conclusion will be reached regarding any measured "anisotropic" light speed.

The catch lurks in the statement "Any relativistic time dilation is back calculated by computer a Z, using the data on "isotropic" speed of light gleaned by increasing time delay of reception of each signal."

In the end you will get what you assumed, i.e. isotropic light propagation. I can understand why you want to use (accelerometer-measured) acceleration, because that's one of the dynamics of motion that is not influenced by a choice of coordinates.

The problem is that in order to translate that to a "one-way speed of light", you need a theory that is built on the assumption that it is isotropic in flat spacetime. And if that theory has withstood every test ever thrown at it so far, it is hard not to think that it was a darn good assumption...

-J

I will thoroughly agree with you, and perhaps any time dilation correction can be ignored in the limits that A and B are at "classical" speeds, and not "relativistic" speeds, but there again, this assumes something in the limits of precision?

Also, I still think a correction can be calculated to an extent that both A and B are totally identical in performance, that each accelerates away with the same precise acceleration (to be held constant by the motor output). Even with the pulses/waveforms arriving with more and more delta t, there should be a symmetic appearance (i.e. - the waveforms are superimposed to within the ultimate criterion of phase lock), that is each showing a steady increase (the same increase) in phase.

Again, it like using the same ruler to measure itself to determine if that ruler is accurate.

We know the ruler is accurate and precise, we just want to re-assure that it reads the same way, regardless of orientation.

If you base measurements on physics principles, there always has to be agreement on the standards of measurement, and how they are to be employed in the course of the experiments. This is nothing new.

Once you commit to using a standard value of C you have corrupted the experiment. You are not flipping the ruler, but attempting to measure the one-way velocity of light.

You can't do that if you use the value for the two-way velocity of light as part of the experiment, at least not in the context everyone is postulating.

Is it possible that one is being slightly arbitrary in stating that? Anyway, as long as relativity is irrelevant in the limit of v->0, then all is well.

What about the approach Ole Roemer stumbled upon? Is this a way to measure the speed of light traveling in only one direction?

Ole Roemer's measurements still used 'two synchronized clocks', albeit it is one clock transported roughly halfway around the Sun and with the measurements made about six months apart.

It is the same as leaving one synchronized clock at the lab and slowly transport the other one to a distance D. Then send a time-stamped light pulse from B to A (and/or A to B) and so determine the propagation time and hence the one-way speed (and isotropy) of light, right?

The problem is that in the limit, this procedure of having synchronized clocks is equivalent to Einstein's sync method, which we know is a 'self-fulfilling prediction'. A philosophical question that may be asked is this: is the slow-transport method another indication that nature somehow prefers light propagation to be isotropic and the 'c' in all inertial frames?

-J

I still do not see the second light path in this measurement but I do understand the 'self fulfilling prediction' aspect. I suspected my idea wouldn't work but could not see why initially.

Thanks again Jorrie.

There do not have to be a "second light path in this measurement", it could be a clock on the return path. The only way to be sure that 'slow clock transport' yields synchronized clocks at the start and end positions is to return the transported clock back to its original position and see to what degree it still agrees with the master clock. It could never be perfect, but in the limit, it should be good enough as an experimental procedure, which is never perfect anyway.

The better method is to send a light signal in one direction to set the second clock and then a return signal to verify the synchronization with the first clock. Which is essentially what Einstein said...

-J

Agreed. Do you think there is a problem with measurement when the observers are traveling with acceleration away from the source origin? As long as this is rectilinear travel, is this still considered a non-inertial frame of reference? If so, then a special case of the experiment would be objects fired off from Z with identical impulse, or even using the same, symmetric impulse. At time t>>0, there is no residual impulse, and v = constant. Time base signal is always being transmitted A-Z, B-Z to verify the correct periodicity of the time base, but if there is some doppler correction to the time base, we have once again employed the "speed of light" to measure the speed of light.

This is why stated the clocks must be identical, must "magically" keep phase lock within 10^-18 of c, and must be returned to origin at end of experiment to verify that lock still exists. Recent French experiments with electro-magnetic fields in nitrogen found something like 1 part in 10¹⁸ change of c with respect to imposed field versus no imposed field. And the field strengths were incredibly high, as I recall. To me, it is almost astonishing that the speed of light, c, is known this accurately and precisely to begin with.

I gave up on trying to measure it one-way only. But here is a link to measure the speed of light using chocolate:

https://www.google.com/search?hl=en&source=hp&q=measuring+the+speed+of+light&gbv=2&oq=measuring+the+speed+of+light&gs_l=heirloom-hp.3..0l10.1516.6047.0.6985.28.13.0.15.15.0.93.937.13.13.0....0...1ac.1.34.heirloom-hp..0.28.1203.dowwjzLggoU

Have not tried, but I suggest using only the best flat chocolate for this. You might actually have to mold your own chocolate into absolutely flat sheets to get the most accurate result. Make sure you have the frequency down correctly, and remember to include that pesky factor of two (measurement is 1/2 wave).

kkris1:

I have been reading this post with interest; your approach refers to an action that would logically read one way speed of light, I emphasize the word action! While the action is logical, you have not stated the necessary preparation to make that action logical?

Regards JD.

Hi JD.

Thanks for your comments.

I was thinking a bit more about the challenge to measure one way speed of light.

I think the preparations you are talking about is measurement of the distance between A and B. That can be relatively easy. You can measure the time the light travel from A to B and back.2-way speed of light is known with high precision, so the distance can be calculated quite accurately.

Than you send the signal from A to B starting the clocks at A and B. Now knowing the distance between A and B and the time the light would travel with "conventional" 2-way speed=c you can delay clock at B to be d/c seconds later (or advance the clock at A to be d/c seconds earlier) than set by our initial signal. So we could assume that both clock are synchronized if one way speed of light is equal to 2 way. Now we can verify this assumption by sending another signal from A to B. If one way speed of light is equal to 2 way, the reading on clock B will be the same as at A minus (d/c). If not, the reading will be different and it will be possible to calculate one way speed of light.

Hi JD,

I was thinking about the experiment, and came to a new idea:

Let's have 5 points A, B, C, D, E equally spaced (distance AB=BC =CD=DE. Let the point C be exactly in the middle. We can indirectly synchronize the clocks at B and D by sending the signals from B and D towards C and calculating back the time on B and D clocks so the signals from B and D would arrive at C at the same time. The clocks at B and D would be perfectly synchronized if the speed of light were the same in both directions. If not, there would be a difference.

Now we can send the signal from B to A and from D towards E at the same time as indicated on clocks B and D. Regardless of the unidirectional speed of light, the signals from B to A and from D to E will arrive at exactly the same time. So now we will have perfectly synchronized clocks at A and E. Measuring unidirectional speed of light now would be quite trivial.

Hi kk, the way that you described, synchronize the clocks according to the standard convention; but this means the (free space) one-way speed of light is guaranteed to be c in whatever direction you decide to measure it by means of these clocks.

The closest one can come, avoiding discrete clocks, are described in this post: http://cr4.globalspec.com/blogentry/21742/One-way-Speed-of-Light.

But event then, as you will find by reading the comments (especially reply #37), one cannot get away from the fact that it is a nifty "two-way" measurement. By establishing a simultaneity scheme for the two ends, one has already used the one-way assumption indirectly. This guarantees the outcome to be c, so nothing gained.

-J

Several things come to mind: (1) how does the experimenter know with a certainty, that extending the cable in a particular direction, or a random direction (other than coiled up in the floor) will not change the effective speed of electrons in the wire?

(2)If not actually measuring the distance (precisely), then how do you know that when extended in direction A, the distance is "identical" to the distance in direction B?

Why not use some sort of quantum effect that has a really short t_(1/2) and the equivalent of a large integrating sphere? Light should pulse more or less incoherently at some wavelength (or at least be scattered isotropically in all directions). The sphere is a very precisely made object with a spherical accuracy that is arbitrarily perfect within the limits of the time delay expected, (i.e. perfect to within 1ppb of the radius), not that I could fabricate such a sphere. The sphere can exist as two hemispheres with a slice missing to allow for insertion of the equipment. The two hemispheres rotate at an arbitrary speed that completes many rotations during the total period of observation. At the inside surface of the sphere is the chromophore agent that fluoresces a new photon energy, but has a really short lifetime, can even be Raman scattered so it is considered to be instaneous. Arguably, if there is anisotropy to the speed of light, then there should be a "time-smearing" or broadening of the received light at the detector. The inside of the "integrating sphere" should be such that the initial wavelength is almost completely absorbed by a layer beneath the active layer, and that the dectector can and should be located both at origin, and at the radius in an array. After turning the sphere in one direction, it can be reversed, then the whole assembly rotated to a new axis, 90 ° from initial axis.

I expect that after running numerous experiments, nothing new will be gained, except to learn that there is no way to synchronize the detector, unless it is a single detector at the origin. Temporal broadening may indirectly reveal something about the isotropic nature of the space where the experiment takes place, but if measured elsewhere, might not be consistent?

"(1) how does the experimenter know with a certainty, that extending the cable in a particular direction, or a random direction (other than coiled up in the floor) will not change the effective speed of electrons in the wire?"

I suppose by, after 'tuning' the two wires for simultaneous pulses, laying them out in various configurations and checking that the simultaneity is maintained.

"(2)If not actually measuring the distance (precisely), then how do you know that when extended in direction A, the distance is "identical" to the distance in direction B?"

In the final determination of the speed of light, we obviously have to know the precise distance that the light pulse has to travel. A perfect tape measure should suffice, not so?

As you remarked, I also do not see how using "quantum distances" will yield anything better than Don Lincoln's method. In fact, he said he and his colleagues use the method as a matter of routine for measuring the speed of various particles...

-J

Perhaps Maxwell's demon needs to get around a bit more, and move a whole lot faster?

Hi Jorrie,

I think you have misinterpreted my idea:

Let's have 5 points A, B, C, D, E equally spaced (distance AB=BC =CD=DE. Let the point C be exactly in the middle. We can indirectly synchronize the clocks at B and D by sending the signals from B and D towards C and calculating back the time on B and D clocks so the signals from B and D would arrive at C at the same time. The clocks at B and D would be perfectly synchronized if the speed of light were the same in both directions. If not, there would be a difference. But the difference will be nullified if we send the light the same distance but in opposite direction.

So let's send the signal from B to A and from D towards E at the same time as indicated on clocks B and D. Regardless of the unidirectional speed of light, the signals from B to A and from D to E will arrive at exactly the same time. So now we will have perfectly synchronized clocks at A and E. Measuring unidirectional speed of light now would be quite trivial.

Let us assume for example that the light needs 1s to travel from B to C and 3s from D to C. If we synchronize the clocks at B and D to be 0 at the moment the light was emitted (if the light signals arrive at C at the same time), the real time at the time tC as indicated on the clock at C will be as follow:

tB =tC+1s;

tD=tC+3s

Now if we send the signals from B to A and from D to E (at the same time as indicated on clocks at B and D) we will have the time on the clocks at A and E as follows:

tA=tB+3s=(tC+1s)+3s =tC+4s

tE=tD+1s = (tC+3s)+1s=tC+4s

So the clocks at A and E will be perfectly synchronized.

Regards,

KK

KK, your sketch is not visible on my side, just a blank block.

BTW, there is a bit of technical inconsistency when you say that you "back-calculate" the time on B and D clocks, but I will wait for your diagram to become visible, before taking this further.

I'm not sure what you are trying to achieve; the one-way speed of light in any inertial frame is set to be the same as the two-way speed by the Einstein convention for synchronizing clocks. It is then also guaranteed to be the same in all directions/angles. Every scheme that claimed to "measure the one-way speed of light" in an inertial frame has been refuted over and over, because it always boils down to a "self-fulfilling prophecy".

-J

Hi Jorrie,

My sketch is quite simple diagram in paint. I do not know why it is not visible, but is not that crucial.

When I'm saing "back calculate" the time on B and D clocks, I mean that if the signal from B arrives 2s later than from D, we can delay clock at D by 2s, so the signals from B and D ( if send at the same time as indicated by clocks B and D) would arrive at C at the same time.Of course, if the unidirectional speed of light was the same as 2-way speed, the clocks at B and D would be perfectly synchronized. If not and speed from B to C would be faster, the clock at B would have to be delayed to have signals from B and reached C at the same time. To compensate for that, you need to send the signals in opposite direction. Therefore clocks at A and E will be perfectly synchronized regardless if the speed of light is isotropic or not.

Regards,

KK

OK, as I understand you have 5 equidistant points A,B, C D, E, linearly placed in an inertial frame (i.e. non-accelerating, non-rotating). Then you synchronize B, C and D using light signals, i.e. using distance and the speed of light, c.

Take note that there can now be no violation if the isotropy of light speed, simply by the method that you have used and that you now cannot use those clocks to measure the one-way speed of light, because you have defined it to be c.

When you say two or more clocks are "perfectly synchronized", it implies that you automatically get light speed c in any direction; otherwise they are not "perfectly synchronized". As an example, if you would either accelerated your frame in the direction A to E, or rotate it around any of the points, the clocks cannot be perfectly synchronized at all. Whatever you do, they will drift out of sync and light propagation will not be measurable as c.

Bottom line, your method does nothing to alleviate the fact that the one-way speed of light is defined and therefore it is meaningless to "measure" it. All we can do is to detect anisotropy in one-way speed due to rotation, acceleration and gravity. The most precise accelerometers and angular rate gyros make use of this anisotropy.

-J

Jorrie,

This is a theoretical problem: some people (specially young earth creationists) argue that it is impossible to synchronize two distant clocks and therefore to measure one way speed of light, which could be nearly 1/2c in one direction and nearly infinite in opposite direction and still would be c for the round trip. All proposed methods (including Einstein's) of synchronizing the distant clocks would postulate that one way speed of light is the same as 2 way. If not, the distant clocks would not be synchronized.

In my proposal I have shown that the distant clocks can be synchronized regardless if the speed of light is isotropic or not. By sending the signal one direction (from B to C and then opposite (from B to A) I have compensated for any possible variation in the speed of light in opposite directions. I have shown that the clocks at A and E are delayed by the same amount of time in comparison to clock C and therefore synchronized.

Now if you send the signal from A , it will arrive at E at time tA+ the time needed for the light to travel from A to E.

Please try to show me otherwise using my method (you can try to analyse the situation for any speed between 1/2c and infinity)

Regards,

KK

All of the clocks to be utilized MUST be synchronized at the same point in time, and same "point" (meaning observable vicinity of a few feet to allow for practicality) - C.

Two of the clocks are then slowly transported carefully to A, B, D, E stations. Then signal transmission begins (preferably of some slowly oscillating sine wave, or set of waves that is encoded with time stamp). A phase lock amplifier at C that is a four channel instrument then picks up the total signal, one in each channel, deconvolutes the time stamp, and also measures phase angle of the light signal. Voila, if there is any phase drift over time between the four channels that is non-attribute of the measuring instrument, then perhaps you have a "reasonable" measurement that proves that C(Θ) is an isotropic constant, or that C(Θ) is not a constant but exihibts spatial variance. This seems entirely simple and intuitive to me, am I wrong? IF the speed of light is variant of space, then even though the frequency of the {laser} signal is a constant, the wavelenth detected a Θ=0 will be slightly (or greatly) different from wavelength detected at Θ(180), and there will obviously be an accumulating phase drift proportional to difference of wavelength, hence difference in speed of light.

James, I'm not sure I get the "drift" of your comment. If the clocks are slow-transported to their new positions and they then remain static relative to the original clock (C), there will be no phase shift between them in gravity-free space.

In the limit, slow transport is exactly equivalent to Einstein's sync method, so ideal clocks will remain sync'd, period. Real clocks are obviously never "ideal", so there will be some de-sync over time, but for this argument, it can be calibrated out, I guess...

-J

If the signal is continuous "laser" signal with time stamp encoded digitally, and the entire is also amplitude modulated at a given low frequency (but not so low as to be in the low frequency noise problem), then if there is an an-isotropic speed of light, signals being detected in a particular orientation will experience phase drift, plain and simple. I see this in my mind's eye much easier than I can draw it on here. Sorry that I do not have animation capabilities, or I would make us all a nice cartoon with Felix the cat (formerly living in Shroedinger's place).

"... then if there is an an-isotropic speed of light, signals being detected in a particular orientation will experience phase drift, plain and simple."

James, I don't understand your one-way phase drift; compared to what reference? The clocks A, B, D, E? If there is a frequency shift, it implies Doppler shift and this implies that they are not stationary relative to clock C (at the source of the signal, I presume). I fail to see how it has anything to do with the one-way speed of light.

In any case, accurate phase shift measurement (interferometry) is only possible over closed paths, not one-way.

-J

OK here we go: modulated light signal arrives from A to C (known distance, no gravitational perturbation, etc.) and is observed to contain the following information:

(1) the phase angle of this signal is Φ=0, and is the reference signal, (2) it has a constant frequency of x/2π. (3) It has a constant rms amplitude within 0.5%. (4)Time encoded in this is Φ=0 at distance L.

Simultaneously, a similar signal is being emitted from B to C (not necessarily 180 orientation away from A), and also contains similarly coded information, and is not necessarily at the same distance L from C as is A.

From the perspective of the two-channel input phase lock amplifier, if the speed of light C is invariant, then Φ_A-B will be a constant, and non-wavering within the limits of the qualification of the observing instrument, and the signal generators at A and B (which are known to contain identical time base crystals). Clearly, the statistical quantity, average(Φ_A-B) will be a constant. As time approaches infinity, variance(Φ_A-B) will approach a minimum limit.

On the other hand, if speed of light C is a spatial variant, then one signal will have a slightly longer or shorter wavelength than the other one (although the frequency of modulation is the same at both sources), and Φ_A-B will exhibit an observable time-dependence related to C₀-C_× , thus if the difference in light speeds is 1x 10^-6 v, and v is 1 MHz, the change in phase angle of the experiment will exhibit a new frequency of 1 Hz. This is because the modulated signal takes different amounts of time to traverse a similar (but non-identical distance).

I can readily see with a spreadsheet that two sine waves with a wavelength ratio of 0.995 will exhibit a marked phase shift relative to each other after only 15-16 periods.

James, yes such a type of experiment can detect anisotropy of propagation. Not quite by phase lock-in amplifiers (they are apparently not accurate enough), but it has been done using Mössbauer spectroscopy.

AFAIK, the only time such tests have shown up anisotropy in light propagation is where either acceleration or gravitational potential difference were present. A gravitational example is the Pound-Rebka experiment.

-J

thank you for continuing the discussion past the point where I almost assumed it to be a dead horse. I do not know that gamma rays would be the best source to try for this, maybe the maser source would work.

http://www.extremetech.com/extreme/184879-einsteinian-error-the-25-year-old-supernova-that-could-change-the-speed-of-light-forever?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+ziffdavis%2Fextremetech+%28Extremetech%29

This seems interesting? If neutrinos are unaffected (per se) by gravitational wells, and photons are, that seems a bit confusing. Not to mention the proposed idea that positron/electron pairs spontaneously form and disintegrate, and that is enough to slow light down.