Recent lyot coronagraphic performances achieved with adaptive optics (AO) instruments ADONIS/SHARPII, PUEO/KIR and NAOS/CONICA are presented and compared to detection limit predictions estimated with a simple analytic approach. The aim is to discuss and to determine the different observing limitations, to optimize the use of the instrument capabilities and to support the design of future instruments in the specific field of high contrast observations.

We present the first results obtained at CFHT with the TRIDENT infrared camera, dedicated to the detection of faint companions close to bright nearby stars. The camera's main feature is the acquisition of three simultaneous images in three wavelengths (simultaneous differential imaging) across the methane absorption bandhead at 1.6 μm, that enables a precise subtraction of the primary star PSF while keeping the companion signal. The main limitation is non-common path aberrations between the three optical paths that slightly decorrelate the PSFs. Two types of PSF calibrations are combined with the differential simultaneous imaging technique to further attenuate the PSF: reference star subtraction and instrument rotation to smooth aberrations. It is shown that a faint companion with a ΔH of 10 magnitudes would be detected at 0.5" from the primary.

The detection of a faint companion to a star with achromatic coronographyand adaptative optics is only limited by the residual speckles.We study a method to increase the detectability of the companion, basedon the difference of two images taken in two different wavelengths, aftersuitable rescaling.

Deconvolution is a necessary tool for the exploitation of adaptive opticscorrected images, because the correction is partial. The Maximum APosteriori (MAP) framework is used to derive a deconvolution method(MISTRAL) thatcombines the data with our knowledge of the noise statistics as well as ourprior information about the object and the variability of the Point SpreadFunction. The deconvolution of experimental and scientific data illustrates the capabilitiesof this method.

Highly performing adaptive optical systems are required for nextgeneration giant telescopes as well as next generationinstrumentation for 10 m-class telescopes. Different types of AOsystems are currently under study, including Multi-Conjugate AO(MCAO), high dynamic range AO, and low-order AO for distributedpartial correction AO. These systems require a large variety ofdeformable mirrors with very challenging parameters. Thedevelopment of new technologies based onmicro-opto-electro-mechanical systems (MOEMS) is promising forfuture deformable mirrors. We are currently developing a MDM basedon an array of electrostatic actuators with attachments to acontinuous mirror on top. The most challenging building block forthis device is the high optical quality mirror. Our first resultsshow a very efficient planarization of the surface with ourdesign. The integration of this mirror surface on top of anactuator array is under investigation.

We give in this short communication the principle of the densification of a large classical on-axis aperture,with aperture central obscuration and spider arms into a full circular unobstructed aperture, forapplication to high-contrast coronagraphy. An example is shown using a VLT like aperture. In space, such atechnique would allow to produce a perfect null of the star image using coronagraphic techniques such as the4QC or the PKC, at the expense of a small reduction of the transmitted light. The drawback of the techniqueis that a surprisingly very small ZOF is obtained, if the telescope full resolution is considered. However,if a loss of resolution of a factor 2 is accepted, then the system might behave as a perfect imaging system.The technique is found useless for ground based observations with the Strehl ratios reached at the presenttime.

A definition study for a Ground-based European NullingInterferometry Experiment (GENIE) has been initiated by theEuropean Space Agency (ESA) and the European SouthernObservatory (ESO). This instrument is planned to be installed in 2006 on the VLTI at Mount Paranal, Chile. The primaryobjective of the GENIE experiment is to gain experience in the design, manufacture and operation ofa nulling interferometer using a Darwin representative technology. Besides its technological goal,GENIE is designed to survey exozodiacal dust clouds around nearby stars and to study the brightestof them by means of low-resolution spectroscopy. GENIE also aims at low-mass companion detection,including hot Jupiters. In this paper, the expected performances of the GENIE interferometer areinvestigated as a function of the observation wavelength and of the performances of criticalsubsystems.

This paper describes our strategy and presents the specific additional calibration tool added to the near-infrared focal instrument AMBER of the Very Large Telescope Interferometer, to achieve the direct detection of extrasolar planets with color-differential interferometry. The measurement of the angular orbits, masses and spectra of hot giant extra solar planets can be achieved for Jupiter size planets up to 0.2 a.u. away from a 10 pc distant star, if it is limited only by photon, detector and background noises. However, this requires an extremely good control and calibration of the atmospheric and instrumental stability.

Adaptive Optics is now making it possible to efficiently use large telescopes in ground-based interferometers. Spatial filtering, especially with optical fibers, allows accurate interferometric measures to be done on these new interferometers (VLTI, Keck, OHANA). However, as telescope sizes increase and sensitivity improves, the problem of field of view of interferometers become more acute. In this paper, I focus on the observation of extended sources with fiber interferometers, and show that accurate visibility measurements are made difficult by the extent of the observed source. Out of the several solutions explored, one seems especially attractive: multi-fiber coupling. In this optical scheme, the pupil is subdivided in sub-pupils, each being coupled into a separate optical fiber. I show how this technique extends the field of view of fiber interferometers, reduces the errors in the interferometric measurements of extended sources, and provides flexibility which improves the sensitivity of fiber interferometers.

In the frame of ESA's DARWIN mission technological studies, Alcatel Space, supported by ESA, is designing and manufacturing a nulling interferometer whose aim is to demonstrate in laboratory the extinction of a point source with a stable nulling ratio of 10-6. This breadboard is presented with emphasis on the integrated optics beam combiner.

"Densified-pupil multi-aperture imaging interferometers" also called "hypertelescopes" can provide direct images at their focal plane. Coronagraphic masking is applicable for nulling a star's image and searching associated planets. The scheme is applicable at the infra-red wavelengths where a planet has its best contrast, but also potentially at visible wavelengths where chlorophyll-like spectral features are to be searched. In the infra-red, where zodiacal and exo-zodiacal clouds tend to contaminate the image, the scheme provides a better rejection of this contamination than schemes using a beam-splitter and flat interference. For a typical Earth-like exo-planet the detection time is reduced 10 to 100 times. In the visible, planets are often resolved from their star by an interferometer's sub-aperture. Visible coronagraphy is then achievable in each sub-aperture before combining the beams and producing the high-resolution image. This relaxes the piston tolerances. Whether for sub-apertures or a densified pupil, or even a conventional telescope, multiple stages ofcoronagraphy are of interest, if equipped with as many deformable mirrors or transmissive plates. Located in the successive image planes, such adaptive correctors can phase the residual stellar speckles without affecting the planet peak. This generates a central peak of starlight, which can be masked without much obstructing the planet's wavefront. A design concept studied for an "Exo-Earth Discoverer", a hypertelescope variant proposed to NASA for its Terrestrial Planet Finder, involves 37 mirror elements of 1 m, driven by small solar sails and arranged to form a 40-300 m diluted spherical mirror. One or several focal combiners move along the focal surface. Optical solutions have been explored in preliminary detail.

In this work, I show that a wide field nulling imager is a very attractive solution for discovering and characterizing Earth-sized planets around other stars. The concept presented in this paper takes advantage of recent developments in nulling coronagraphy and a number of other techniques (aperture synthesis, pupil densification, pupil dilution) that make it possible to obtain high quality coronagraphic images with an interferometer.

This paper presents a theoretical limit to the field/resolution ratio for theimaging mode of aperture synthesis interferometers. This limit is a function ofboth the number of apertures in the interferometric array and the dynamic range ofthe visibility and phase data. It does not depend on the optical setup ofthe instrument. This work is based on an application of the theory of information.

Using simplificative assumptions, we show that a redundant configuration of diluted apertures should permitan easier exoplanet detection than a non-redundant one. This result assumes that after coronagraphy theplanet appears over an uniform background whose statistical fluctuation is the unique source of noise. Thedetection is made using a matched filter, cross-correlating the observed image with the telescope array PSF.In that case, the relevant parameter appears to be the integral of the modulus squared of the aperture MTF.