Einstein was right (+/- 0.05%)

Joseph Taylor and Russell Hulse discovered the binary pulsar PSR1913+16 in 1974, and found that their orbits do shrink just as predicted, with a smaller than 0.5% margin of error.

http://nobelprize.org/nobel_prizes/physics/laureates/1993/illpres/discovery.html

Thanks for pointing it out!
( I've got online subscriptions for Science and Scientific American, so let me know anytime in the future .... )
It does look good indeed. The data ( in black ) is accumulated from many orbits ( 100+ ), and the exact shape of the predicted behaviour ( in red ) is sensitive to whichever theory of gravity you want to test. It's from the stats analysis on the curve/data fit which yield a confidence of 0.99987 +/- 0.00050 to Einstein's GR. Whew!
If you can imagine the plot being repeated many many times along the horizontal axis, because it represents a cyclic event, you can envision the delay stretching in and out as the pulsars orbit. It seems to be significant when the pulsars are within about 30-45 degrees either side of aligned ( from our viewpoint ). Stretch the space, let it relax, stretch the space, let it relax ..... you get the idea :-)
Of course the whole space between the pulsars is pretty warped, but this study has isolated the 'traverse' ( as in surveying ) along our line of sight that cuts the system. Thus it's trimmed a few dimensions out, but heck I can't visualise beyond three anyhows ... :-)
Cheers, Mike.

I'm not sure 'lensing' is the right term, as no image is being formed, but you've grasped the idea that gravity is altering the spacetime so that the radiation emitted from the rear pulsar is being distorted by it's passage past the companion star when the alignment ( to hit us ) is right. Hence the delay.

Imagine a really long piece of string, as long as one Imperial foot per nanosecond of transit time, and held tight between the rear pulsar one end with one of the radio dishes at Parkes [AUS], Jodrell Bank[UK] or Green Bank[USA] at the other. So when the second pulsar ( which is a neutron star of course ) goes in front then you have to pay out the string only to wind it in later - all because of the spacetime distortion. That would be on top of a lot of other motions of course. It's a sloppy analogy, but one way to visualise it.

I'll mention that there were three other parameters measured and compared to GR expected values, which were accurate to within about one percent, so not as good as the Shapiro number. These were

(i) the derivative of the orbital period [GR accurate to 0.5%] - a negative number, so the period is getting shorter, meaning they orbit each other more rapidly as time goes on;

(ii) the gravitational redshift parameter [GR accurate to 1.2%], which indicates the decrease in frequency/energy of the radiation as it 'climbs out' of the system;

(iii) a Shapiro 'range' parameter [GR accurate to 1.5%] which when combined with the 'shape' parameter [in a fairly compicated way] models the delay due to the other pulsar's gravity well. It was the shape parameter which was quoted as having that high concordance with GR.

Cheers, Mike.

Naw, 'lensing' is fine, I like 'diffracted' too, and 'blurred' is tres cool .... I suppose if we had radio dishes for eyes we'd be saying stuff like that .... :-)

Mind you, then we'd be wondering what Neutrino Man was looking at when he came around for pizza & a beer ..... :-)

You're understanding is right on, but it's hard to get the right language isn't it? Maybe 'warping' and 'curvature' is best .. dunno ?

Being such a foreign circumstance ( it doesn't happen hereabouts ) the visualising is difficult too. Rubber sheets give the flavour for sure though.

Cheers, Mike.