In the late 80's, the scientific support group of the 2 meter Télescope Bernard Lyot (TBL) at Pic du Midi, Southern France, decided to include a speckle camera in the new instrumentation program, to take advantage of the unique low-seeing quality of that site. The instrument was to be used with the new CP40 photon-counting detector duplicated by Observatoire de Paris and INSU (Institut National des Sciences de l'Univers) from the prototype developed at Observatoire de la Côte d'Azur (Blazit, 1976, 1987). Because of its involvement in image restoration techniques, the ``Aperture Synthesis'' team of Observatoire Midi-Pyrénées (OMP) was selected to design, construct and test this instrument.
The original goal of the scientific support group was to incorporate the speckle camera within the new bonnette at the Cassegrain focus, along with the acquisition and guiding detectors) of the TBL in order to be able to switch to speckle observations during a run in the case of good seeing (``flexible scheduling''). The concept looked attractive at first, but quite difficult to implement in practice. It would have been hard for an observer to interrupt his program and give up the best images to a speckle program. This procedure would also have been quite risky, since it involved switching instruments during the night, at a crucial moment when the images were good. Another problem was that a substantial amount of telescope time would have been used for technical purposes, i.e., assembling, tuning and testing the instrument. This would have temporarily reduced the scientific output of the TBL which was not a good thing, as the survival of the telescope itself had been put to question for several years because of budget restrictions.
All these reasons lead to a more conventional solution, which consisted in developing an independent instrument that could be fully mounted and tested in the optical laboratory. Another advantage was the possibility of using this speckle camera on larger telescopes such as the 3.6-meter ESO or CFH telescopes.
As most of us belonged to the ``Aperture Synthesis'' team of Midi-Pyrénées Observatory and were promoters of new numerical methods for high resolution imaging in optical and radio astronomy, we also wanted to test these methods with the new speckle camera. This required that the pupil plane would be accessible and pupil masks could be put into it. Such masks could simulate telescope arrays such as the ESO VLTI (European Southern Observatory Very Large Telescope in the Interferometric mode), or others. This speckle camera would thus provide appropriate experimental tools for the investigation of image restoration techniques with the optical telescope interferometric networks that are currently being built all around the world (VLTI, GI2T, SUSI, COAST, see f.i. Beckers and Merkle, 1991).
We first present a general overview of PISCO and its optical concept (§2). Then we introduce the various operative modes and illustrate them with observational results (§3). The detectors that have been used with PISCO are presented in §4. In §5, we discuss the current performances of the instrument in relation with the physical limitations of the speckle techniques and the image restoration methods. In §6 we briefly describe the on-going scientific programs which use PISCO, and conclude about the use of a speckle camera in the current context of new high angular resolution techniques.