Spatial correlations of spontaneously down-converted photon pairs detected with a single-photon-sensitive CCD camera. B. Jost, A. Sergienko, A. Abouraddy, B. Saleh and M. Teich in Opt. Express 3:81 (1998).  What the paper says!?

This reports the observation of spatial-pairing correlations from SPDC bi-photons (both degenerate- and nondegenerate-SPDC photons) imaged on a single-photon-sensitive ICCD camera. Spatial here means transverse, i.e. correlations in detector coordinates, as opposed to the timing coincidences everyone had been doing since Burnham–Weinberg.

There is a time-to-space transposition:

Our results can be viewed as the spatial analog of the timing coincidence measurements that are found in the SPDC literature.

Two configurations are used, that correspond to near-field (they call it one-to-one and is a 4f lens setup, capturing the crystal plane) and far-field (angle-to-position 2f) imaging. In principle that allows a near/far-field pair of setups to image the biphoton's EPR position and momentum correlations—the two conjugate bases being in Fourier optics a lens apart—but only the far-field (momentum, opposite-side) correlation actually came out. Such EPR measurements will be done later by Howell et al.. The "spatial imaging" is because such momenta correlations get imaged in space but the real-space imaging provided no correlations: the near-field plane maps signal and idler to correlated positions, but because SPDC pairs are emitted with near-identical transverse position at the crystal, the one-to-one image doesn't resolve the pairing the way the momentum (far-field) plane does: momentum anti-correlation is what produces the opposite-side annulus signal, so "spatial correlations" is somewhat of a misnomer, if taken literally (and in contrast to our spatial correlations).

The far-field produces an annulus, with correlations on opposite sides:

In this way, they reproduce with a camera the measurement of which was scanned by hand. Note that A. Sergienko is an author to both of these papers.

Much attention is given to factors that blur or spread the correlations. Even the successful (annulus) configuration, required such practical attention:

We initially attempted to use the idealized correlation analysis described above but failed to produce meaningful results primarily because of two practical difficulties:

Interestingly, the measurement to a quantity that refers to the full structure provides one of the difficulties:

the center of the down-converted photon annulus is difficult to precisely identify (and is not very useful for full-ring analysis if the rings are distorted) and errors are introduced by the registration (mapping) of the rings in translating from a rectilinear (pixel) to a polar coordinate system

so it is better to use absolute measurements. A technique is presented to extract correlations from such annulus shapes on arrays of pixels (basically considering patches of pixels as opposed to single pixels).

The correlations are observed as follows:

It would have been nice to see the counterpart of this image at other (in particular, perpendicular) angles.

The image above includes an estimated 10 to 100 bi-photon events.

The correlations are not striking, especially as they are not presented in a very compelling way, but this is a pioneering demonstration of spatial correlations, here, from the source itself.

They also point to Malygin et al. (their Ref. [10]) as the first observation of spatial correlations of SPDC.

A related experiment was later performed by Oemrawsingh et al.