Context: Exoplanet science has made staggering progress in the last two decades, due to the relentless exploration of new detection methods and refinement of existing ones. Yet astrometry offers a unique and untapped potential of discovery of habitable-zone low-mass planets around all the solar-like stars of the solar neighborhood. To fulfill this goal, astrometry must be paired with high precision calibration of the detector. Aims: We present a way to calibrate a detector for high accuracy astrometry. An experimental testbed combining an astrometric simulator and an interferometric calibration system is used to validate both the hardware needed for the calibration and the signal processing methods. The objective is an accuracy of 5e-6 pixel on the location of a Nyquist sampled polychromatic point spread function. Methods: The interferometric calibration system produced modulated Young fringes on the detector. The Young fringes were parametrized as products of time and space dependent functions, based on various pixel parameters. The minimization of func- tion parameters was done iteratively, until convergence was obtained, revealing the pixel information needed for the calibration of astrometric measurements. Results: The calibration system yielded the pixel positions to an accuracy estimated at 4e-4 pixel. After including the pixel position information, an astrometric accuracy of 6e-5 pixel was obtained, for a PSF motion over more than five pixels. In the static mode (small jitter motion of less than 1e-3 pixel), a photon noise limited precision of 3e-5 pixel was reached.

Theia is an astrometric mission proposed to ESA in 2014 for which one of the scientific objectives is detecting Earth-like exoplanets in the habitable zone of nearby solar-type stars. This objective requires the capability to measure stellar centroids at the precision of 1e-5 pixel. Current state-of-the-art methods for centroid estimation have reached a precision of about 3e-5 pixel at two times Nyquist sampling, this was shown at the JPL by the VESTA experiment. A metrology system was used to calibrate intra and inter pixel quantum efficiency variations in order to correct pixelation errors. The Theia consortium is operating a testbed in vacuum in order to achieve 1e-5 pixel precision for the centroid estimation. The goal is to provide a proof of concept for the precision requirement of the Theia spacecraft. The testbed consists of two main sub-systems. The first one produces pseudo stars: a blackbody source is fed into a large core fiber and lights-up a pinhole mask in the object plane, which is imaged by a mirror on the CCD. The second sub-system is the metrology, it projects young fringes on the CCD. The fringes are created by two single mode fibers facing the CCD and fixed on the mirror. In this paper we present the latest experiments conducted and the results obtained after a series of upgrades on the testbed was completed. The calibration system yielded the pixel positions to an accuracy estimated at 4e-4 pixel. After including the pixel position information, an astrometric accuracy of 6e-5 pixel was obtained, for a PSF motion over more than 5 pixels. In the static mode (small jitter motion of less than 1e-3 pixel), a photon noise limited precision of 3e-5 pixel was reached.

NEAT is an astrometric mission proposed to ESA with the objectives of detecting Earth-like exoplanets in the habitable zone of nearby solar-type stars. NEAT requires the capability to measure stellar centroids at the precision of 5e-6 pixel. Current state-of-the-art methods for centroid estimation have reached a precision of about 2e-5 pixel at two times Nyquist sampling, this was shown at the JPL by the VESTA experiment. A metrology system was used to calibrate intra and inter pixel quantum efficiency variations in order to correct pixelation errors. The European part of the NEAT consortium is building a testbed in vacuum in order to achieve 5e-6 pixel precision for the centroid estimation. The goal is to provide a proof of concept for the precision requirement of the NEAT spacecraft. The testbed consists of two main sub-systems. The first one produces pseudo stars: a blackbody source is fed into a large core fiber and lights-up a pinhole mask in the object plane, which is imaged by a mirror on the CCD. The second sub-system is the metrology, it projects young fringes on the CCD. The fringes are created by two single mode fibers facing the CCD and fixed on the mirror. In this paper we present the experiments conducted and the results obtained since July 2013 when we had the first light on both the metrology and pseudo stars. We explain the data reduction procedures we used.

NEAT is an astrometric mission proposed to ESA with the objectives of detecting Earth-like exoplanets in the habitable zone of nearby solar-type stars. NEAT requires the capability to measure stellar centroids at the precision of 5e-6 pixel. Current state-of-the-art methods for centroid estimation have reached a precision of about 2e-5 pixel at two times Nyquist sampling, this was shown at the JPL by the VESTA experiment. A metrology system was used to calibrate intra and inter pixel quantum efficiency variations in order to correct pixelation errors. The European part of the NEAT consortium is building a testbed in vacuum in order to achieve 5e-6 pixel precision for the centroid estimation. The goal is to provide a proof of concept for the precision requirement of the NEAT spacecraft. In this paper we present the metrology and the pseudo stellar sources sub-systems, we present a performance model and an error budget of the experiment and we report the present status of the demonstration. Finally we also present our first results: the experiment had its first light in July 2013 and a first set of data was taken in air. The analysis of this first set of data showed that we can already measure the pixel positions with an accuracy of about 1e-4 pixel.