On 3 Jul, 03:22, HW@....(Henri Wilson) wrote:
On Mon, 02 Jul 2007 06:48:25 -0700, George Dishman wrote:
On 2 Jul, 01:19, HW@....(Henri Wilson) wrote:
On Sun, 1 Jul 2007 12:10:59 +0100, "George Dishman" wrote:
"Henri Wilson" HW@.... wrote in message
.. .

There are NO side bands with white light George.
You are raving again.

They are not mathematically definable side bands.

Of course they are, the same equations apply.
Multiplication is associative so modulating white
light by a sine wave is exactly the same as
modulating a sine wave carrier with a white noise
signal - e.g. voice on AM.

voice is not white noise George.

Aspirates henry, but stop changing the subject, it
should be obvious that they are as easily definable
mathematically as conventional sidebands on an AM
radio.

Henry, we have been talking for a couple of months
using these definitions and there's no way to go back
and revise the meanings now, but you are not really
proposing a change anyway, what you have said just
repeats how we always assumed the effect worked.

Yes. That is correct.

Right so let's stick with that terminology and not
confuse the issue.

However the bunching patterns are more than just an
acceleration curve.

Yes, as I said above, both apply and ADoppler is
the extra factor beyond VDoppler that only appears
in ballisitic theory.

ADoppler is usually by far the dominant factor.

Forget the mantra Henry and think. We have proved
there is no ADoppler on pulsars and for contact
binaries where the speeds are the highest and
ADoppler should show up. If you look at the radius
curves, there is no ADoppler on Cepheids either,
so no, it is not "usually .. dominant", it is
non-existent in every case we have examined.

....
We have always understood the luminosity to be the product
of the two. The energy per photon also has an effect in
theory but we always agreed to treat that as negligible
for the time being and modern detectors (CCD) are sensitive
to the number of photons but not their energy anyway.

Good.

Since the point is that the luminosity matches
the first derivative of the radius, not the
second, it is fundamental to the conversation.
You seem to have lost the plot a bit here Henry.

No, I'm only trying to match curves at the moment.

That's OK, you need to match the radius curve for
this example. That will resolve whether there is
any ADoppler effect or if it is only VDoppler.

VDoppler can't produce large magnitude changes like these.

I know, ballistic theory cannot explain Cepheid
curves.

I would call the first derivative of radius the
radial velocity and the second derivative the
radial acceleration. The luminosity matches the
velocity, not the acceleration.

Sorry George, you have regressed.

You are merely considering the energy of each photon.

Nope, I have ignored that as we agreed some weeks
ago. I am simply differentiating the radius to get
the velocity and then the acceleration.

You aren't using it as an indicator of bunching (or photon density).

Yes I am, I am comparing it against the luminosity
curve which is a direct count of photons in a given
band.

which IS velocity
dependent....h.(c+v)/lambda.
You are completely ignoring the principle factor involved, which is 'number of
photons arriving per second'.

On the contrary, that is the only factor I am
considering at the moment.

You aren't considering the bunching effect due to velocity differences.

Yes I am, that is the ADoppler part which should
be proportional to the top of my three plots, but
the curve shape is completely wrong - that shows
there is no ADoppler.

Ballistic theory says light energy travels at a certain
speed. The Doppler shift of sidebands determines the
speed at which the modulation travels, hence my point
is that the two are inextricably linked. You cannot
define one without that also defining the other.

Are you suggesting that modulating a light beam changes its speed?

No, I am saying (a) that pulses can be produced by
modulating a beam and ballistic theory says the
pulses travel at c+v, and (b) the speed of the
modulation is related to the Doppler shift equation
hence (c) the Doppler shift equation can be derived
from the ballistic speed by purely mathematical means.

That's a good one George....

It's your own Henry, do the sums.

f'=f(c+v)/c

OK, now apply that to the carrier and sideband
frequencies independently. Then inverse transform
the three to get the received waveform. What speed
does the modulation travel at? Show your working ;-)

I don't see the point.
What are you getting at.

Yet again Henry, consideration of sidebands allows
you to calculate the Doppler shift directly from
the speed of modulating pulses hence a speed of
c+v determines the shift.

You have to consider monchromatic light for this
anyway.

The maths is easier that way (but it isn't essential
as any arbitrary waveform can be converted to frequency
space via a Fourier Transform).

There are two possibilities. Either the sphere remains virtually at rest
as the
star moves around a small orbit at its centre....or the sphere moves in
phase
with and by the same amount as the star.

Or some mix.

,,,, which would show up as an anomalous phase difference between brightness
and observed velocity.

Yes.

1. give 100% VDoppler.

That is true if the speed equalisation distance is
very small (i.e. less than the sphere size).

yes

The same is true without the sphere if the distance
is less than the depth of the stellar atmosphere,
but that is what you claimed above regarding
"molecular velocities" so the additional sphere is
redundant.
2. gives 100% ADoppler with a small amount of VDoppler.

Again, that is true only if the speed equalisation
distance is large so that it travels a long way
through the ISM at variable speed. The same is true
without the sphere if the light so again it still
appears redundant.

NO NO NO!!!!!. You are quite wrong there George.
The sphere effectively becomes the source. If it moves with the star, then the
original c+v relationship with Earth holds.

Exactly, so what is the difference from saying it
leaves the star at c+v?

But George, if TWO stars are in close orbit, the common sphere remains
vurtually at rest wrt both.

You spent a long time telling me there were two spheres and
each moved with its parent star. You lost the plot somewhere.

There is very little if any 'c+v'. The light from
both leaves at c/n...but is wavelength shifted dring the unification process.

Surely you can see this.

Yep, but the same is true if the light changes to speed
c at surface of the heliopause as it moves into the ISM
so what does the sphere do?

(I suppose it could be (c+v)/n,
where n is quite small ).

Right - see "dispersion measure".

Let's not worry about n.

No problem but keep it in mind.

I imagine a neutron star to have a large sphere of matter and fields
around
itself. ...and that sphere would not keep up with the pulsar's movement.

It makes no difference I can see, if the speed
equalisation distance is small you get VDoppler while
if it is large you get ADoppler.

You see wrongly George. You don't have the picture.

You are right, I can't see why you think it makes any
difference.

see above. It makes a difference in cases like contact binaries.

You said the sphere moved with the star.

... what I said was that
we know there are no errors in the derivation
of the predicted angle from the respective
theories and they give different answers.

Ther is an optical lens effect anyway.

That's what we are talking about, effectively
Newton predicts twice the focal length for a
given mass.

I meant an atmospheric lens as well as the gravitational one.

That doesn't come into the maths of GR or Newton,
they give different predictions.

It isn't typical.

I suggest you go and look, see if you can find
one that looks any different.

None looks like it at all.

I repeat, I suggest you go and look ;-)

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

is typical

It doesn't show a radius curve.

Wake up, George.
The curves I produce show 'arrival photon density' vs time.

The curves _I_ am showing you are real observations,
and the luminosity match the velocity, not the
acceleration.

You are thinking solely of h.nu. ...individual photons. You are not considering
photon density at all.

Nope, the calculations are of photon rate only,
the energy per photon is not considered.

You are missing the whole point.
You are becoming quite clueless George.

Since you can't even work out what was being
calculated, it is you who needs the clue - see
above.

George, I'll let the computer calculatebthe bunching.

Let me give you the clue again - we were talking about
differentiating the radius to get the radial velocity.
You have wandered off on all sorts of tangents.

It even includes YOUR suggested method of doing it....which matches mine
exactly.

I never suggested it was, what I said is that the
luminosity matches the first derivative, not the
second. Try listening to the argument first.

They are both roughly in phase ....

What!!! One is a sawtooth and the other is rectangular.
You really are losing it these days.

Again, You are becoming quite clueless George.

Simple statement of fact Henry.

Even Max Keon thinks you are losing it.

Have a look at Max's first attempts at writing
equations, he though he needed two, one for negative
numbers and another for positive. Both took the
square root of a square. He didn't know the
associative, distributive and commutative laws until
I pointed him at K12 pages. His maths is way behind
yours!

There is no "orbit" Henry, we are talking about
the radius of a single star.

Well I am now reverting to my original view that cepheids are just orbiting
stars. I have found a perfectly sound explanation for the presence of presumed
'overtones'.

So find one that gets eclipsed at the fundamental.
Statistically there must be many.

I would like to find one.

Exactly, without them you have a problem to explain.

I start with the measured radius and differentiate.
That gives me the radial velocity and the shape
matches the measured luminosity. If I differentiate
twice I get the radial acceleration and that doesn't
match the luminosity curve, hence we know that the
luminosity is _not_ the result of ADoppler.

George, I can match any cepheid curve using ADoppler.

You will need to add another curve to your program
if you want to try. That is the distance of the source
from us. Match that to the radius curve and then see
whether it is the velocity or the acceleration that
matches the luminosity.

I don't need any other curves George.

Yes you do, you have nothing to compare against the
radius.

The radial velocity of the surface of a star that goes 'huff puff' is very
similar to that of one in elliptical orbit.

You think? So add the curve to your software
and let's see it.

I think the point you are missing is that an ADoppler curve can be almost
the
same as a VDoppler one...but for different reasons. At an eccentricity of
around 0.25, the typical cepheid velocity curve resembles either a
VDoppler one
for a yaw angle of 180 or an ADoppler one for a yaw angle of about -50.

Exactly the point, when you claim you can match a
curve using ADoppler, it could as easily be VDoppler

Except for one thing George.
VDoppler variations are minute. ADoppler can easily produce variations up to
mag 4.

However, differentiating the radius twice is nothing
like the luminosity curve, it only matches the velocity.

George, your own suggested method of calculating photon bunching matches just
about any brightness curve.

Don't try to change the subject Henry.

In the latter case, the brightness curve is in phase with the velocity
curve
and has the same basic shape....which is what is observed.

The point you are missing is radius curve in those two
conditions would be quite different and it is only the
case where the luminosity matches the VDoppler where
the radius matches what is observed. There is nothing
to corroborate your suggestion of ADoppler and solid
evidence to rule it out.

You are referring to the TRUE radius variation. It is the OBSERVED variation
that matters.

I am referring to the radius measured by means of the
angle subtended by the star so I don't see what
distinction you are drawing.

OK, some people have claimed to have seen cepheids actually pulsating.

For goodness sake Henry, what do you think we have been
talking about for the last several weeks ????? The
ESO page is exactly that measurement.

There might be stars that actually do that and it might indeed be possible to
see them....but I would be very suspicious..

Welcome to the conversation.

Go ahead then, add the radial distance curve, match
it to the radius curve for L Car and "let your program
provide the answers".

Again, all you are considering is the h.c/lambda energy effect.....not the
'photon density' one.

Nope, you are lost entirely. I am discussing the
measured luminosity and CCD detectors are photon
counters, not sensitive to the energy. None of
this discussion has been related to photon energy
at any point.

Just let the computer do the sums and produce the curves George.
My program is corrrect.

ROFL, I have pointed out the error in it many, many
times.

Incidentally, I have now included the effects of tidal bulges and have found
that their effects are very similar to those of a first overtone. I have
matched some brightness curves very closely.

Worthless until you match the radius at the same time,
or find an eclipser so you can match the phase, As you
admitted, you can just as easily match a curve with
VDoppler as with ADoppler

George