Images would be equal to or superior to those produced by multiple-grid-pair telescopes.
The figure is a simplified depiction of a proposed Fourier telescope for imaging in hard-x rays and γ rays. This instrument would contain only one pair of grids made of an appropriate radiation- absorpting/ scattering material, in contradistinction to multiple pairs of such as grids in prior Fourier x- and γ-ray telescopes. This instrument would also include a relatively coarse gridlike image detector appropriate to the radiant flux to be imaged. Notwithstanding the smaller number of grids and the relative coarseness of the imaging detector, the images produced by the proposed instrument would be of higher quality.
Both grids would have the same overall width. If n were the number of slits or slats in one of the grids, then the other grid must contain n + 1 slits, respectively. Because both grids would have the same overall width, the width of an individual slit or slat would be slightly greater in the n grid than in the n + 1 grid. It would not matter which grid was characterized by the greater number; for the initial design, n and n + 1 would be chosen for the outer and inner grid, respectively. The image detector could be composed of as few as two elements; however, prior research has shown that seven elements would represent a better compromise between the quality of image data and the complexity of the hardware.
Although practically any alignment could be used as long as it were known a priori, it would be convenient to align the middle element of the detector with the central slits of the inner and outer grids. With this alignment, a point source on the axis of symmetry of the telescope would produce a fringe pattern having peak intensity on the middle detector element. As the point source moved off the axis, the fringe pattern would shift accordingly, enabling acquisition of data on the amplitude and phase of the spatial-frequency component corresponding to the slit width, distance between grids, and grid angle. The processor would sum the photon counts on the detector elements to produce a four-parameter output data stream indicative of the intensity and location of the peak amplitude on the detector (equivalently, of magnitude and phase) as functions of the angle of rotation and the distance between the grids.
This work was done by Jonathan Campbell of Marshall Space Flight Center.