The strobed word benchmarking script could show interval timing in a more
explicit way: it should plot each "dOut" interval vs. each "getInterval"
interval to show whether jitter in the dOut measurements is alway, sometimes,
or never reflected in the externally measured interval. As of r459 it's not
clear which dOut measurement corresponds to which getInterval, so there's only
an overall sense of how well kne accounts for jitter in the other.
Original issue reported on code.google.com by Benjamin.Heasly on 18 Oct 2010 at 7:09
Original issue reported on code.google.com by
Benjamin.Heasly
on 18 Oct 2010 at 7:09