4 Laboratory and sky results

We first tested the imaging capabilities of the hypertelescope with laboratory-simulated single or multiple point sources. Figure 4 shows the laboratory images obtained. The classical limitation of interferometers regarding field crowding also arises in hypertelescopes. With a periodic array of Nelements, field crowding occurs when the ZOF contains more than 4N stars. The image is aliased by dispersed higher-order peaks from those stars located in the HOF but outside of the ZOF, thus affecting the information retrieval. No such degradation occurs here with the numerous apertures and the few stars.

In a second step, we verified the hypertelescope's imaging properties on bright stars. Figure 3b shows an image of Pollux ( $\beta \ {\rm Gem}$, $m_{\rm v}=1.14$) exposed 10 min. The four residual peaks of order 1 surrounding the central zero-order peak result from the incomplete densification utilized. Their measured intensity relative to the central peak is 1/16, close to the theoretical value 1/20. In October 2001, we have obtained images of binary star Castor A-B ( $\alpha \ {\rm Gem}$, $m_{\rm v}=1.98$), for which the magnitude difference is $\Delta m=0.9$. Exposures were taken with and without Wratten filters W25(R), W58(G) and W80A(B) during 10 to 30 min. In Fig. 3c the companion $\alpha \ {\rm Gem}$ B appears clearly, in spite of seeing affecting the image's sharpness. We measured for JD = 2 452 193.67 a separation $\rho =3.8\pm 0.3$ arcsec and a position angle $\alpha=68.2\pm 5^{{\circ}}$, close to the expected values (Heintz 1988), $\rho =4.0$ arcsec and $\alpha=63.81^{{\circ}}$. The theoretical number of resolution elements inside the ZOF is 312 (4N). We measured in Figs. 3b and 3c, $307 \pm 26$ resolutions elements (see Eq. (3)), due to uncertainly of ZOF diameter ( $28 \pm 1.2$ arcsec) mesurement.

Figure 4: Laboratory simulated triple (left) and quintuple (right) stars observed with the hypertelescope. The image is a color composite of three images taken through R, G and B filters.

The intensity profile of Castor A contaminates the flux of Castor B by about 20%. After appropriate subtraction, the magnitude difference between Castor A and Castor B is $\Delta m=0.8 \pm 0.15$. The significant magnitude inaccuracy is due to a nearly 2.7  $\hbox{$^{\prime\prime}$ }$ FWHM seeing.