deletions | additions       diff --git a/scat5.tex b/scat5.tex  index ae09ea4..5a8e8a8 100644  --- a/scat5.tex  +++ b/scat5.tex 
...
All of the above discussion on scattering timescales would seem to suggest that the best-case scattering time could be taken into account when addressing requirements SKA1-SYS\_REQ-2961 and SKA1-SYS\_REQ-2962. However, there are reasons why we may wish to resolve the scattering tail of a given pulsar and so somewhat arbitrarily imposing a time resolution limit based on scattering is viewed as undesirable. This being said, wewish to ensure that we  do not have want  requirements that imply extremely computationally expensive operational modes and as such a limit must be determined on the minimum time resolution as a function of DM. In the next section we will attempt to provide such a limit based on keeping the computational cost of coherent dedispersion to a reasonable level.