5 Conclusions

Due to the progress in the determination of precise radial velocities by CORAVEL type instruments, the lowering of the observable separation limit for visual binaries by interferometric techniques and the concomitant drastic improvement of positional precision, more and more spectroscopic binaries are going to be visually resolved in the coming years. In a close future, new observing techniques (such as speckle and visual interferometry, satellites) are going to add to the set of stars that are resolved both visually and spectroscopically. Efficient reduction techniques will then be even more needed.

To obtain stellar masses and distances is still the main goal of binary studies. To get more powerful instruments and to improve the observation techniques should not be the unique objective of astronomers working in this field. In many instances, powerful reduction methods such as the one we propose can derive significantly improved data from already existing observations.

The approach presented in this paper is not limited to binary stars and other areas of astronomy could take advantage of our method. Any problem that can be considered as a minimization problem may profit from a similar approach. We hope this paper is going to promote SA in the astronomical community.

Acknowledgements

I express my thanks to H. Eichhorn (who also suggested this investigation field) and R.E. Wilson for having critically read the manuscript and for the many fruitful discussions we had about this topic. I also thank P. Lampens and the anonymous referees for their suggestions.