http://www.spacetelescope.org/news/html/heic0803.html

Years ago, when I first heard about the possibility of gravitational
lensing, where a star bends the light around it, I saw a really sloppily
done graphic.  The graphic implied you  would only get two points, left and
right to the star.  I immediately thought there ought to be a ring. The
article I read seemed to confirm that you'd get one image superimposed to
the left and right.   I then assumed there was some property of light or
gravity or magnetism I didn't know about, or the graphic  designer and/or
the article writer didn't know what they were talking about.

They didn't!

They're called Einstein RINGS!

Post a reply to this message

Greg M. Johnson wrote:
> They're called Einstein RINGS!

I think if it's not lined up precisely, the left and right side get 
shorter and shorter until it's a dot. If it's not behind the item at 
all, you'll see the direct light and then the "slingshotted" light, 
causing basically two dots.

But yah, that's pretty cool.

-- 
   Darren New / San Diego, CA, USA (PST)
     It's not feature creep if you put it
     at the end and adjust the release date.

Post a reply to this message

Greg M. Johnson <pte### [at] thecommononethatstartswithycom> wrote:
> Years ago, when I first heard about the possibility of gravitational
> lensing, where a star bends the light around it, I saw a really sloppily
> done graphic.  The graphic implied you  would only get two points, left and
> right to the star.  I immediately thought there ought to be a ring. The
> article I read seemed to confirm that you'd get one image superimposed to
> the left and right.   I then assumed there was some property of light or
> gravity or magnetism I didn't know about, or the graphic  designer and/or
> the article writer didn't know what they were talking about.

http://en.wikipedia.org/wiki/Einstein_cross

-- 
                                                          - Warp

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
>
> http://en.wikipedia.org/wiki/Einstein_cross
>

At least that's SYMMETRIC.

http://en.wikipedia.org/wiki/Einstein_ring

I think that gravitational lensing makes more intuitive sense as a ring, or a
symmetric arrangement of points.  Two points I think is easier to represent in
vector art, and I still think a temptation if you don't really understand the
phenomenon.

Post a reply to this message

gregjohn <pte### [at] yahoocom> wrote:
> I think that gravitational lensing makes more intuitive sense as a ring

  I wonder if the shape of the lensing object makes a difference. After all,
galaxies are seldom spherical.

-- 
                                                          - Warp

Post a reply to this message

"Greg M. Johnson" <pte### [at] thecommononethatstartswithYcom> wrote

> http://www.spacetelescope.org/news/html/heic0803.html
>
> Years ago, when I first heard about the possibility of gravitational
> lensing, where a star bends the light around it, I saw a really sloppily
> done graphic.  The graphic implied you  would only get two points, left
and
> right to the star.  I immediately thought there ought to be a ring. The
> article I read seemed to confirm that you'd get one image superimposed to
> the left and right.   I then assumed there was some property of light or
> gravity or magnetism I didn't know about, or the graphic  designer and/or
> the article writer didn't know what they were talking about.
>
> They didn't!

Sure, they did. You can get other effects than a ring, depends on the
configuration. The very first gravitational lensing observed was the so
called twin quasar (ie, two images), so that's probably what you remember.

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote:
> gregjohn <pte### [at] yahoocom> wrote:
> > I think that gravitational lensing makes more intuitive sense as a ring
>
>   I wonder if the shape of the lensing object makes a difference. After all,
> galaxies are seldom spherical.
>
> --
>                                                           - Warp

Yes, the shape of the projected mass distribution of the galaxy alters the
number, location and distortion (both angular size/magnification, and shape) of
the images.

A galaxy whose projected mass distribution has pure circular symmetry can
produce three distinct images (per source) when the source is away from the
centre of mass (but is still within a critical distance of the centre), and
will produce an Einstein ring when some part of the source falls directly
behind it.

A projected mass distribution that has a degree of ellipticity can produce five
distinct images (per source) when the source is close enough to the centre of
mass, or three images as the source is moved far enough away from the centre
(but is still kept within a critical distance).

The third and fifth images form close to the centre of mass (and light), and
unlike the other images are strongly demagnified instead of boosted by the lens
effect. This means that they are usually not observable, and it's common to talk
of two- or four- image systems as a result.

Tom

Post a reply to this message

Okay, I'm back to square one.  I'm bothered by getting any points at all. I want
rings.

Say your interposed object is a super-dense ball (in case I don't understand
"circular symmetry")  in front of a star, then you would get a ring.  If it's
off to the side, don't you get a distorted ring instead of points?

Unless the magnetism of the interposed ball has an effect on the light, or the
polarized photons from the star are affected by gravity in some way, you should
get a band of light and not discrete points.  Or maybe magnetism of the
intervening ball affects how its material is "available" for lensing, making an
assymetrical lens.  There's gotta be one more phenomenon at work besides
gravity.


"Tom York" <alp### [at] zubenelgenubi34spcom> wrote:

> The third and fifth images form close to the centre of mass (and light), and
> unlike the other images are strongly demagnified instead of boosted by the lens
> effect. This means that they are usually not observable, and it's common to talk
> of two- or four- image systems as a result.
>
> Tom

Post a reply to this message

gregjohn nous apporta ses lumieres en ce 2008/01/16 07:10:
> Okay, I'm back to square one.  I'm bothered by getting any points at all. I want
> rings.
> 
> Say your interposed object is a super-dense ball (in case I don't understand
> "circular symmetry")  in front of a star, then you would get a ring.  If it's
> off to the side, don't you get a distorted ring instead of points?
> 
> Unless the magnetism of the interposed ball has an effect on the light, or the
> polarized photons from the star are affected by gravity in some way, you should
> get a band of light and not discrete points.  Or maybe magnetism of the
> intervening ball affects how its material is "available" for lensing, making an
> assymetrical lens.  There's gotta be one more phenomenon at work besides
> gravity.
> 
>  
If the lencing object is precisely on the line from the observer and the lensed 
object, you get a ring.
If the lencing object is slightly shifted, you get one or more cressent(s) and 
possibly a specular dot.
If the lecing object is even more shifted sideway, you see a slightly shifted 
primary image and a lenced punctual or arced image.

If the lencing object is homogeneous, the effect is regular.
If it is diffuse and somewhat irregular, like a galaxy, the effect become 
distorded. The more irregular, the more distortion you get.

-- 
Alain
-------------------------------------------------
Never delay the ending of a meeting or the beginning of a cocktail hour.

Post a reply to this message

"gregjohn" <pte### [at] yahoocom> wrote:
> Say your interposed object is a super-dense ball (in case I don't understand
> "circular symmetry")  in front of a star, then you would get a ring.  If it's
> off to the side, don't you get a distorted ring instead of points?

You'll find that the ring splits into two arcs for a real source, yes, because
the lensed source will have a finite size. But if you take any single point on
the source, you'll find that there are no more than two points that correspond
to it on the lensed image - except for the source point that's directly behind
the super-dense ball (you understood correctly), which maps to a complete ring
in the lensed image due to the symmetry.

You can model the lens effect with an appropriately shaped glass lens. The
bottom of a wineglass can be quite close to the required shape, you need
something roughly conical (or more steeply sloped if possible). Or you could
use POV, just create a conical lens with its apex facing towards you and move a
bright sphere behind it. Annoyingly none of the interactive simulations (java
and such) I used to be aware of seem to be around any more.

Tom

Post a reply to this message