The gravitational microlensing experiment currently underway on the Mt. Stromlo 50-inch telescope is another excellent example of how an old telescope can be outfitted with a modern detector system to produce superb results. This experiment makes use of a novel an sophisticated CCD imaging design whereby by 4 2048x2048 detectors are butted together to form a single 4096x4096 pixel detector system. In fact, a dichroic beam splitter placed in the optical path allows the incoming beam to be split into red and blue channels with each channel have a 4096x4096 detector. Since a single 2048x2048 detector is 16 bits deep in its imaging plane, then the readout from 4 million pixels is 8 Mbytes. The Mt. Stromlo system reads out 8 of these devices every few minutes and thus generates a data rate on the order of 10 Gigabytes per night of observing. Still, workstation and computing power are able to handle this load. Figure 4 (1.1 Megs) shows an example image from one of the LMC fields that is searched for gravitational microlensing amplification of stellar brightness. Overall, the MACHO project is a brilliant confirmation that small telescopes are still quite useful and can perform novel imaging studies.

The Short Focal Length Approach

In the late 80's, prior to the proliferation of million-pixel CCDs, the only practical way to achieve a wider imaging field was through the reduction of the telescope image scale. This reduction requires a modification of the telescope secondary mirror or some form of additional reducing optics. At the Michigan-Dartmouth-Mit Observatory, located on the Southwest Ridge of Kitt Peak, a focal reducing camera constructed by Greg Aldering, can be used on the 1.3-m telescope to reduce the image scale by a factor of 4.67. The cost of this camera, which contains 5 elements (lenses) was approximately $12k and it is more fully described in the December 1991 issue of the Publications of the Astronomical Society of the Pacific. Figure 5 (124 Kbytes) shows an image of NGC 253, taken with this camera and a 320x512 RCA CCD which has 30 micron size pixels. The image scale is 3.05 arcseconds per pixel. The rectangle in the lower right shows the field size that would be obtained without reimaging optics.

This figure also shows the downside of focal reducing cameras. The many new optical elements in the light path, required to reduce the image scale, also produce ghost images when there are very bright sources in the field. These spurious images are so marked in Figure 5. Figure 6 (151 Kbytes) shows another image, of the high surface brightness galaxy M101, obtained with this reimaging camera. In this case, the detector was a TI4849 CCD with 22 micron pixels resulting in an image scale of 2.32 arcseconds per pixel. The exposure time for this image was 30 seconds (through a Blue filter) which was sufficient to capture much of detailed structure in M101 (the black dots in the centers of some bright stars indicate saturation of the CCD).

As a final example, Figure 7 (202 Kbytes) shows a 30-second image of M33 taken with this camera where again much of the structure of M33 can be seen and the dust lanes are evident. Overall then, a modest focal reducer can produce a large enough field-of-view on 1-meter class telescopes with Cassegrain foci to effectively CCD image those nearby galaxies which have been adopted as morphological standards, otherwise known as the Hubble Sequence. In this way, images taken with this camera through different filters can show how the morphology of a galaxy is highly wavelength dependent

Page 5

The Electronic Universe Project