Re: Wasn't this impossible ?

Not exactly. The signals are combined coherently and a huge effort is made to compensate the differential paths as the scopes track. It is quite an art. The phase centre is deliberately put just outside of the target field of view because all local interference is coherent there.

The signals from every antenna are phase compensated and digitised and combined in the correlators against every other antenna. On a good day VLA can do a quick snapshot in 5 minutes because it is a Y shape.

The original aperture synthesis algorithms were strictly limited to E-W baselines like the One Mile Telescope or nearly E-W like the Ryle 5km at MRAO Cambridge. It is pretty much like making Young's slits sensitivity patterns on the sky and measuring the sky brightness Fourier transform.

VLBI avoids the strict need to maintain perfect coherence by using invariant combinations of three antennas as closure phase and of four antennas for closure amplitudes. These remain good observables even when the absolute phase at each antenna is shot to hell. It needs a high precision synchronised maser clock at each dish and as many dishes in the network at possible to get enough good observables. It also require a lot of patience and some luck to find the interference fringes.

This technique is now also used in the optical at CHARA and before that on the experimental rig at COAST in MRAO. Essentially using radio astronomy methods to work with optical (near IR) light beams.

The only system I know of that used photon amplitude correlation only was Hanbury-Brown & Twiss's intensity interferometer from Jodrell Bank which surprised most physicists at the time by actually working! They measured the diameters of numerous bright stars for the first time since Michelson & Pease had done it with an iron girder on the Mt Wilson 100".

There is a book about it which I think is online somewhere.

Is is not coherence that allows the laser light to spread less rapidly? Do I have this wrong?

Coherence is the only light-emission plan that gives a lot of light into a single direcction (etendue problems). Thermal sources the size of stars can be seen at a distance; light bulbs, not so much.

Hmm OK, I thought coherence was only about the wavelength / frequency purity. As in coherence length. Send a thermal source through a narrow interference filter and you increase the coherence of the output light...

George H.

George H.

See there you have it. It might surprise people by working. I'm not asking if known tech can do it. I'm asking if it's feasible for an array to collect a lot of visible light and process multiple images into one with more clarity. Like a common deblurring algorithm except it would be given multiple images to start from.

Thinking about light propagation as photons bouncing around is a one-hundred-percent guaranteed way to get the wrong answer.

Cheers

Phil Hobbs

It is. ;)

Cheers

Phil Hobbs

That's temporal coherence. There's also spatial coherence to worry about.

Cheers

Phil Hobbs

But the *BIG* snag is that without phase information the result you get is very ambiguous. You don't get an image so much as a centrosymmetric autocorrelation of the sky brightness distribution. That isn't too bad on a stellar disk but it is pretty whacky on most complex objects.

Closure phase (and amplitude to a lesser extent) is the preferred technique these days to get good observables out of tricky data. A technique devised by Jennison at Jodrell Bank.

ISTR Baldwin did a phaseless image of Cygnus A at 151MHz with synthetic phases added to the measured amplitudes based on other observations at

It is actually rather old technology. Frieden developed a photographic method of speckle interferometry using narrow band starlight that could resolve some close doubles with enough aperture. It was exacting work and rather tedious. Never really caught on.

Lucky imaging which arose from faster more sensitive cameras and a rather simple piece of logic sometimes used in aperture synthesis of aligning the highest spike of each image has tended to take over these days. Astonishing resolution limited by available aperture images are possible today by amateurs using what are pretty much everything modified webcams to state of the art CCDs. Basically take lots of shots and throw away all the bad blurry ones and shift and add the remaining high contrast images using a smart algorithm.

A side effect is that Jupiter is now under observation somewhere in the world by at least two or three amateur scopes with enough resolution to show any impacts. Turns out there are more than we ever imagined.

Speckle interferometry was a good technique that you could do using photographic plates, but it's obsolete now, as you say. (*)

The lucky imaging trick works fine as long as the object is bright enough. Seeing fluctuates in a bandwidth of up to 20 Hz or thereabouts, so you need to be able to estimate image sharpness reasonably well in at most 100 ms.

Cheers

Phil Hobbs

(*) Even though silver halide film can now be made as sensitive as a good CCD, by adding formate ions to the emulsion.

Right, thanks. AFAIK, spatial coherence involves a pin hole that the laser/ light source is focused through. (memories of a holography lab.)

George H.

Yup, the fringe visibility as a function of separation of the two pinholes goes like the modulus of the complex degree of coherence. (Its phase sets the phase-shift of the fringe pattern.)

Cheers

Phil Hobbs

(Who had the privilege of taking statistical optics from Joe Goodman himself many moons ago, and remembers odd bits of it.)