At 02:53 PM 3/23/2008, Joe Subich, W4TV wrote:
>Nobody has even used CW Skimmer "in anger" yet. To be
>fretting about fully automated stations is an awfully big
>leap. I have been watching the stuff that comes out of
>Pete's Skimmer and there is a lot of trash ... in the
>last hour I see three spots for North Korean stations,
>Andaman and Nicobar, one Bhutan, one Macquarie Island,
>and probably two dozen other less "rare" multipliers I
>know to be badly busted calls.
Well, I think the Skimmer vs Human test on the 1999 KCDXC pileup recording
suggests that it has a lot of potential.
For the last couple of weeks I have been using CW Skimmer with a beta
version of N1MM Logger that validates Skimmer spots by comparing them to
the current contest log and to the master.dta file. This produces a very
high-quality bandmap, at the cost of perhaps missing a few "uniques" -
which is, I think, a pretty good tradeoff. Come to think of it, similar
validation applied to traditional packet spots might produce a comparable
improvement.
It is also worth mentioning that even if Skimmer copies the calls
correctly, there is still a lot of human skill required to execute the S&P
QSO without losing your running rhythm (or your frequency).
73, Pete
_______________________________________________
CQ-Contest mailing list
CQ-Contest@contesting.com
http://lists.contesting.com/mailman/listinfo/cq-contest
|