deletions | additions       diff --git a/section_Advantages_and_disadvantages_It__.tex b/section_Advantages_and_disadvantages_It__.tex  index bc8f627..43ced4a 100644  --- a/section_Advantages_and_disadvantages_It__.tex  +++ b/section_Advantages_and_disadvantages_It__.tex 
...
An additional and perhaps more important advantage is that the dual-injection system relieves an important constraint on BEASST, namely the requirement to work with as faint a star as possible in order to allow adjustments for as many targets as possible during the night. With this constraint relieved, BEASST can be designed to work with only bright reference sources and such a design can be made much simpler, and more importantly, more robust. One possible design for a revised BEASST would be to use a moderate-order (perhaps 10$\times$10 lenslet) Shack-Hartmann wavefront sensor to measure both tilt and shear on the same camera. This optical design would be extremely simple (two elements, a lenslet array and a camera) and would allow the crosstalk between tilt and higher order aberrations such as coma to be rejected, as well as allowing any systematic differences between artificial and natural light sources (beam size, aberrations etc) to be easily dealt with. Such a design would not be acceptable if BEASST had to work with faint sources, as the light would be spread over too many pixels and BEASST therefore would not achieve the required sensitivity.  The main disadvantages of going to the dual-injection system would be: are:  \begin{itemize}  \item Additional work would be required  to design and test the UTM beam injector. Relatively simple designs consisting of an LED, pinhole and lens could likely meet the requirements, since the stability requirements are modest. This work could be offset against the simpler design possible for BEASST. \item BEASST must be ready when the AAS functionality is needed. Again the simplicity of the revised BEASST design means that this is not likely to causegreat  schedule difficulties. \end{itemize}