Why are the 'Fixed Stars' so FIXED?

On Sun, 13 May 2007 15:13:25 +0100, "George Dishman"
wrote:


"Henri Wilson" HW@.... wrote in message
.. .
On Sat, 12 May 2007 12:26:55 +0100, "George Dishman"

lambda_r = D * sin(phi) / N

apples universally.

It applies...but it is of no value if the diffracted rays move at speeds
other
than c.

Congratulations, you just discovered one of the problems
of BaTh, gratings don't tell you the incoming wavelength.

George, you are not showing much intelligence here.

If a grating is being used to measure the speed of a star wrt Earth, my
equation is perfectly adequate. The absolute wavelength a particular
spectral
line is known.

Henry, stop and think and try to understand. I have
said your equation is correct but it doesn't tell
the whole story. The terms relating to v, u and the
incident wavelength simplify to e the reflected
wavelength. That can be found from the simpler
equation I gave you above. Think of that as the first
step in a process.

Now to find the source motion, you need to measure
lambda_i.

No you don't. For any common spectral line, it is absolute, universal and
known.

That plus an assumption of the source
wavelength gives you the Doppler shift. The trouble
there is that you have no way to _measure_ v or u.
You can make assumptions of course, such as v=u=0
if you attach the grating to a refractor and then
you get a result, but they remain assumptions.

What we were talking about though was the general
"grating equation", not just its specific application
to astronomy. In that context, the general equation is

lambda_r = D * sin(phi) / N

I would prefer to say "A general equation" not "THE general equation".
Mine is a more general one.

and the reason is fairly obvious when you look at the
diagram, there needs to be a whole number of waves in
the section marked "N lambda_r"

http://www.georgedishman.f2s.com/Hen...ic_grating.gif

If the grating is being used to measure an unknown wavelength, then so
long as
the source is at rest wrt the grating, the procedure will be exactly the
same
as the standard one, using the standard equation.

The fact remains that the _general_ equation is what
I said, the requirement is a whole number of waves on
the _reflected_ beam, not the incident light.

Mine is better. It includes factors that are not permitted in your theory.

Obviously the frequency cannot change on reflection so
you can use the second equation

f = (c+v)/lambda_i = (c+u)/lambda_r

but don't confuse that with the grating equation. The
latter is a general equation in ballistic theory
relating to reflection from any surface.

That must be correct.

Just let u=v and my equation iss the same as the classical one.

Now you are trying to use a constant speed model, u is
in general not equal to v in BaTh so you cannot make
that assumption.

How do you know what u is.

That's my point, you don't. You do know lambda_r
however because it can be found from D and phi.

You do indeed.
.......and since you know Lambda_i, you can now calculate c+u/c+v.

You have been claiming u=v to make out the BaTh fails to explain sagnac.
Have
you changed your mind George?

No, I have been pointing out that v=0 in Sagnac therefore
u=0 whether the speed is the same or rest to c or something
inbetween:

http://www.briar.demon.co.uk/Henri/speed.gif

My whole point is that your grating should not work in the HST while
mine
does.

And my point is that your basic algebra is not even
up to schoolboy level whaich means you are not even
capable of working out the correct equation for BaTh
even though you got the diagram right, and even after
I did the algebra for you.

there is nothing wrong with my algebra. ..and you know it...

Other than the fact that you didn't finish it, right.

I initially wanted to make the assumption that u = 0.
That can be checked by comparing diffraction angles produced by identically
marked transmission and reflection gratings.



Sorry, they can also be exactly matched (better in fact)
by the huff-puff mechanism which you know exists. Show
me evidence from an eclipsing Cepheid that the phase is
that of ADoppler instead of VDoppler.

Name one eclipsing cepheid and I will.

http://ebola.eastern.edu/model_display.php?model_id=186

Orbital period 1.3668 days. Cepheid variable period 0.0583
days.

You have both luminosity and velocity on that page and you
can also note the Shapiro effect on the velocity curve which
clearly shows the near discontinuous change at the peak
that I drew and you criticised before.

I think all these velocity curves are nothing but guesses.
I don't believe anyone actually measured doppler shift at the same time as they
measured brightness.

The phase of the velocity curve is that of VDoppler while
that of the Cepheid is at an entirely frequency.

Who said it's even a cepheid?


I've been clear about what is needed, no amount of
matching ever constitutes any proof at all, you need
either to prove that huff-puff cannot match (which we
know it does) or you need to show that the velocity
curve is ADoppler related to the orbital phase using
asome secondary reference such as an eclipse or Shapiro
delay (though I can't see how you could measure that to
be honest, it's not like the pulsars).

George, in matching a curve, I have to select fairly precise values for
eccentricity and yaw angle.
For instance typical cepheids have eccentricity between about 0.15 and 4
and
yaw angle -50 to -70.

I don't need any of your vague - even imaginary - tools to tell me what's
happening.

You need another phase refernece to prove it matches
ADoppler instead of VDoppler. Without that you have
no way to disprove the alternative. So far you have
nothing.

No I don't. It tells me all I want to know.



...and the proportions of A and V doppler vary widely.

Nope, VDoppler ADoppler in every case you have
offered so far, including Cepheids.

George, the whole theory behind brightness curves is based on the
'bunching' of
light, which as you know as an 'acceleration' effect....ADoppler....

No Henry the theory is the equations "c+v" and dv/ds ...
Luminosity curves are an _appliction_ of those equations.

'Bunching' and hence 'observed luminosity' is a result of differences in light
speed due to the star's orbital movement.

Cepheid and
most variable stars have brightness and velocity curves that match
ADoppler
predictions.

Prove it, I say they are totally due to the huff-puff
mechanism.

Nobody has managed to link the two effects yet...

So what are you on about?

You have no phase reference to show that the luminosity
matches ADoppler. I know you can never fidn one because
Cepheid variation is intrinsic, there is _no_ second
body involved, but the way science works, I don't have
to prove that, the onus is on you to provide proof that
the phase is what you claim.

http://www.users.bigpond.com/hewn/stupidjerry.jpg


Not in my model George.
It is of no use at all.

Ah so you haven't used Fourier at all I see. Fourier
analysis has always been the backbone of EM work. If
you look at the screen of a spectrum analyser, that
is what you are seeing.

What's this got to do with this topic?
The individual photons are particles.

It is still "classical", and it doesn't match the velocity curves.

It matches everything.

Crap, it fails Sagnac, Shapiro and Ives and Stilwell.

SR fails Sagnac.

ROFL, Henry you are totally clueless.

SR reverts to LET to explain sagnac. In fact it does so whenever it is called
upon to find a physical connection to anything .

Then you can back-calulate the speed and find it isn't
c+v. If your maths had progressed beyond schoolboy
level, you wouldn't need me to tell you that.

...says George, who can't even understand my simple grating equation....


Ah yes, that'll be the one I had to write for you:


"George Dishman" wrote in message
...

"Henri Wilson" HW@.... wrote in message
...

see: www.users.bigpond.com/hewn/bathgrating.jpg

Well done Henry. So your equation is

lambda_i * (c+u)
sin(phi) = ---------------
D * (c+v)

where lambda_i is the wavelength of the _incident_
light.

The wavelength of the reflected light, lambda_r, is
given by

lambda_r c+u
-------- = ---
lambda_i c+v

So your equation can also be written

lambda_r
sin(phi) = --------
D

You have been claiming that the speed didn't appear in
the equation and that wavelength couldn't change. One
or the other is wrong. You also claimed the formula
used frequency instead of wavelength but that too isn't
true. Naturally you can replace the wavelength by speed
over frequency but that just reintroduces speed in the
equation.

The lesson Henry, is to work out the equation before you
start telling people what it contains.

I had already pointed this out:

lambda_r c+u
-------- = ---
lambda_i c+v

I told YOU.

George


Yes I know that...but it is still very strange that Huygens principle
infers
that 'something' is radiating out over the full solid 360 degrees.

Yes, that's what QED addresses. I have no idea how you
will approach doing a non-classical version of ballistic
theory.

George, when it comes to explaining how light travels across the universe from
one star to another I don't need QED or anything other than plain old
ballistics.
Emitted light has ONE reference, its source.



I suggest you try it, they don't. I do that sort
of stuff every day at work.

Well try this experiment.
Modulate a laser beam's intensity with a 1000 hz signal. Do the same with
another laser only with a 1005 hz signal.
Fire them both at a photocell and see what output you get.

Bad experiment, you should have known that the photocell
responds to the envelope, not the carrier.

You accept the output will be the beat.


There are no components at the amplitude variation
frequancy, you need to study the difference between
beats and heterodyning.

George, performing a smartarse act doesn't impress me.

Henry, yur "getting it wrong all the time" act impresses
me even less.

You are continually irelevancies. I think you're stalling for time.

Modulation is when the AMPLITUDE of a carrier if varied according to a
signal.
For pure sine modulation, the result is a product of two sine waves. ...or
a
cos(a) - cos(b) wave.

NO! If you modulate pure sines, it produces sin(a)*sin(b)
which can be written as 1/2[cos(a-b) - cos(a+b)]

That's what I said.

More generally for smaller modulation factors there
will be a DC bias on the modulating waveform (so you
still have a carrier when the modulation goes through
zero) so there is some energy at the carrier frequency
but there is _none_ at the modulating frequency.

There is energy in TWO waves. Both will be difffracted.

Beating is also the sum of two sinewaves which, as you should know, can
also be
expressed as a product of a sin and a cos wave.

Indeed, but the energy remains in the original
frequencies sa superposition applies. A photocell
measures the envelope, not the components.

So really there isn't much difference....

The fundamental difference which you keep getting wrong
is that the product produces sidebands while the sum
doesn't, and neither has any energy at the modulating
frequency.

The 'sum' doesn't have to. The two frequencies will both be diffracted anyway.
I think you're completely lost George.

You don't know that.

I do. A photon in a 1MHz signal is simply the smallest
amount of power that can be emitted at that frequency
and has a value given by Planck's constant and the
frequency of 1MHz.

Hahahahahaha!
What if it arrives at 2c?

So what? If a signal arrives at 1MHz moving at 2c, the
smallest amount would still be a signal of 1MHz.

Bear in mind Planck's equation was derived from the
emission of black body radiation so the equation only
holds at that instant in ballistic theory and you will
need an alternative equation for energy at a later
time.

I thought you meant the 1 MHz was in relation to the source.

Think of a photon as being like a long arrow.....maybe billions of
wavelengths
long....When fired from a laterally moving object, its 'axis' will
generally
not lie parallel to its velocity vector.

Wavefronts are always perpendicular to the motion,
remember Huygens.

In this case the photon axis is not perpendicular to the wavefront.

The axis is perpendicular to the surface by definition.
If not, rotation about the axis changes the orientation
of the surface.

George, coinsider again what I said.

If you fire an arrow towards a target from a moving car, its axis will not
be
in line with its direction of travel.

Therefore an arrow is a bad analogy for a wave which always
moves perpendicular to the wavefronts - Huygens again.

It's certainly a bad analogy from YOUR point of view because you didn't think
of it before and it collaposes your whole sagnac argument.



Yes this one, check the maths, it's just Pythagoras.

George, in light of my 'arrow revelation' if a photon is ten billion
wavelengths long, how many wavelengths will one end be LATERALLY displaced
from
the other as it hits the mirror?

I don't care Henry, lateral displacement has no effect
since the wavefront is perpendicular to the direction
of propagation. Consider two waves moving to the right:

|
|
---------+---
|
|

The second is displaced laterally:

|
---------+---
|
|
|

It changes nothing.

You haven't changed the angle of the arrow.

I have introduced a factor that has the potential to cause fringe shifts for
very small angular velocities. The back end of a photon will hit the mirror at
a different point from the front end. It can easily be displaced sideways by a
number of wavelengths even at very small rotational speeds.

George




www.users.bigpond.com/hewn/index.htm

Einstein's Relativity - the greatest HOAX since jesus christ's virgin mother.