Compelling evidence for in vacuo light speed anisotrophy
in [Relativity]
|
From: Tom Roberts on 17 Apr 2008 00:50 wbbrdr(a)gmail.com wrote: > On Apr 17, 6:01 am, Tom Roberts <tjroberts...(a)sbcglobal.net> wrote: >> You must READ MY PAPER. I derived an errorbar >> > by misinterpreting signal fluctuations as measurement errors. > So your errorbar is completely invalid. Not true. My errorbar is derived directly from the raw data, without any signal dependence at all. It shows that the DATA THEMSELVES, using Miller's analysis technique, are not capable of displaying any orientation dependence (the errorbars from the averaging significantly exceed the variation in the averages). READ MY PAPER to see this. > On Apr 17, 10:03 am, Tom Roberts <tjroberts...(a)sbcglobal.net> wrote: >> > Bottom line: the patterns Miller found are not real, they are >> > INSIGNIFICANT. >> > > Owing to the false premise in your paper you haven't proved that. There is no "false premise" -- what I did is to apply standard statistical techniques to the average of a set of data points to derive an errorbar on the average. There is no "premise" at all in that, except the assumption that basic mathematics applies -- if that is "false" then there is no possibility of doing any physics at all. YOU need to learn basic experimental technique. Specifically: how accurately the average of multiple data points reflects the true mean of what was measured by those points. This is known as error analysis ("error" in the sense of unknown variation, not in the sense of "mistake"). > In my opinion Miller's method of analysis was general enough to cope > with a fluctuating signal. Your "opinion" is based on complete ignorance, and you are therefore not competent to have a cogent opinion. You share Miller's ignorance -- he can be excused because he died before modern DSP techniques were discovered, and before quantitative error analyses were common. YOU have no such excuse. Either get an education or shut up about things you remain ignorant of. > Suppose we look at an FFT of the raw data of such a signal. if the > signal has an orientation dependence, then the bin with a period of > 1/2 turn should stand out. That is PRECISELY WHAT I DID, in section IV (using a DFT, not an FFT [#]). I repeat: you need to READ MY PAPER. but you also need to educate yourself on basic experimental technique so you will be able to actually understand what I wrote. [#] Here's a challenge: do you know what the difference is between an FFT and a DFT? And do you know why an FFT cannot be used on Miller's data but a DFT can be used? HINT: read my paper -- ample clues are contained in section III, _IF_ you actually know what an FFT is. > On the other hand, if the signal has no orientation dependence, or > does not exist, then there should be nothing special about the bin > with a period of 1/2 turn. Right! and after subtracting the background model [#] it is 0 in every run with reasonable stability. ZERO. With an errorbar that translates into an upper bound on "absolute motion" (using Miller's model) of 6 km/s (90% confidence). [#] Read my paper to learn what this is and why it is needed. And also observer that it is completely independent of any pre-supposed "form" of the signal. > So such a theory can be refuted by experiment. Not if the "fluctuations" are large enough to make everything I said "invalid" as you claim. In that case all DFT bins will show random values. Miller's don't -- LOOK IN MY PAPER and see the DFT presented therein. [This is going nowhere, and I probably won't respond unless you actually come up with something relevant.] Tom Roberts
From: Jerry on 17 Apr 2008 05:24 On Apr 16, 4:56 pm, wbb...(a)gmail.com wrote: > On Apr 16, 9:37 pm, Jerry <Cephalobus_alie...(a)comcast.net> wrote: > > On Apr 15, 9:51 pm, wbb...(a)gmail.com wrote: > > > Cahill's analysis of the Joos data illustrate his complete and > > total intellectual dishonesty. > > From a scientific point of view the interesting thing is his re- > analysis ot the Miller data in June 2003. No, there is no scientific interest whatsoever in a biased reassessment made with complete contempt for reasonable error analysis. > That allowed him to derive > the speed of 415km/s, that five years later proved to be so close to > the speeds of 420-450km/s required to reconcile the spacecraft earth > flyby anomalies. I currently have no time to go into the details of his recent paper, but my quick scan of his current paper shows it to be riddled with the same sorts of errors as all of his previous ones. There is no reason for me to waste time going into details. > That means that 70+ years ago Miller did indeed measure the absolute > velocity of the earth and that his data was sound. No. He had a figure that he wanted to match. My own earlier fit for the Miller's published figures, COMPLETELY IGNORING PROPER ERROR ANALYSIS, is that a better figure would have been on the order of 200 km/s and in a direction about 30 degrees greater in right ascension. You can do almost anything you want with bad data, if you are willing to ignore proper error analysis. I also note that it is EXTREMELY difficult to distinguish between absolute motion towards the constellation of Dorado, from absolute motion in the direction completely opposite. Between 1906 and 1933, Miller completely flip-flopped on this issue. The difficulty arises, as you are aware, from the fact that MMX measurements of aether drift should in theory be identical with opposite orientations of the apparatus. Only with extensive measurements made at different times of day and at different seasons of the year should these opposite possibilities be distinguishable. The flyby anomalies are indeed interesting, but unlike the Pioneer anomaly, there is unlikely to be any "new physics" involved. I'm reminded of the big flack over Kotel'nikov et al. "Measurement of delay time and Doppler correction in radar observations of Venus in 1975", Sov. Astron. 20, 718-724. Radar ranging measurements of Venus in 1975 showed that the published orbital parameters for the planet were slightly off, and insufficiently accurate for interplanetary navigation of space probes. Somehow or other, crackpots latched upon this paper as somehow or other disproving relativity. > Your comments about the Joos data are quite irrelevant and fall into > the realm of ad hominem. > http://en.wikipedia.org/wiki/Ad_hominem Look in a mirror, Surfer. You hoped not to be recognized by adopting a different posting style, but a quick "whois" check shows that both you and Surfer hail from Queensland. Back to my original question, which you don't want to answer, for some reason: Suppose I have a theory that the "true curve" should be a sine curve half the amplitude of the one suggested by the Miller data, but 90 degrees out of phase relative to Figure 16 in Cahill's Aperion paper, i.e. exhibiting a peak where the Miller data suggests a trough: http://redshift.vif.com/JournalFiles/V11NO1PDF/V11N1CA2.pdf There are four rotations from the Joos data that I can average together to yield a sine curve of the stated characteristics. See a reproduction of Joos's curves in the following: http://allais.maurice.free.fr/EtudeFuerxer.pdf Am I justified in selecting these four runs as representing "good data" and rejecting the remaining 18 runs, because the average of these four match my theory? Jerry
From: Dono on 17 Apr 2008 10:03 On Apr 17, 2:24 am, Jerry <Cephalobus_alie...(a)comcast.net> wrote: > > I also note that it is EXTREMELY difficult to distinguish > between absolute motion towards the constellation of Dorado, > from absolute motion in the direction completely opposite. :-) :-) > Between 1906 and 1933, Miller completely flip-flopped on > this issue. The difficulty arises, as you are aware, from the > fact that MMX measurements of aether drift should in theory > be identical with opposite orientations of the apparatus. :-) :-) > Only with extensive measurements made at different times of > day and at different seasons of the year should these opposite > possibilities be distinguishable. > :-) :-)
From: Tom Roberts on 17 Apr 2008 11:23 wbbrdr(a)gmail.com wrote: > Also at the end of the first section on page 220 I found, > "These experiments proved that under the conditions of actual > observation, the periodic displacements could not possibly be produced > by temperature effects." Bravo! At least you're reading. But you should also read the literature on this experiment. Shankland et al, using statistical techniques MUCH better than Miller's anecdotal approach, showed a strong correlation between temerature VARIATION and "signal". Remarkably, Miller was correct -- the "periodic displacements" were generated by his ANALYSIS TECHNIQUE, and appear in all runs regardless of temperature effects (though their magnitude varies greatly). This is something that Miller could not possibly have known, as Shannon's sampling theorem was several decades in the future, and modern DSP techniques even further in the future. But today we KNOW Miller's results are bogus. Except for Cahill (and his sycophants in this newsgroup). Tom Roberts
From: wbbrdr on 17 Apr 2008 20:08
On Apr 18, 12:23 am, Tom Roberts <tjroberts...(a)sbcglobal.net> wrote: > wbb...(a)gmail.com wrote: > > Also at the end of the first section on page 220 I found, > > "These experiments proved that under the conditions of actual > > observation, the periodic displacements could not possibly be produced > > by temperature effects." > > Bravo! At least you're reading. > > But you should also read the literature on this experiment. Shankland et > al, using statistical techniques MUCH better than Miller's anecdotal > approach, showed a strong correlation between temerature VARIATION and > "signal". > During the above mentioned experiments, Miller used heaters to study the effect of abnormally large temperature changes. Under such test conditions there would clearly be a strong correlation between temperature variation and signal. Shankland has been accused of misrepresenting data collected during such test conditions as representative of what would happen under normal conditions. His claims are not reliable. > > Remarkably, Miller was correct -- Agreed !!! > >the "periodic displacements" were > generated by his ANALYSIS TECHNIQUE, > From the data in your own paper I can prove thats wrong. But it is so interesting I will start a new thread on it. |