[Elecraft] "Mechanical COntesting"

[Brian Kassel]

Folks:

     There seems to be a big misunderstanding here from folks who 
perhaps don't do too much CW contest operation.  I know of no serious 
contest operations that rely on computer recognition of CW.   True, some 
use programs that will help to occasionally decode a signal that  for 
some reason the op missed due to  inattention at the moment that the  
other station replied to a CQ.  But as others have pointed out, the 
signal must be very close to ideal.  Also some use these programs or 
devices to decode some of the really fast DX stations calling CQ, but 
those examples are rare.  I have played with these devices,  and 
programs over the years many times. I have built micro controlled 
devices both from others and my own design.  I have written endless PC 
programs as well. It all boils down to the fact that in the real world, 
as some have pointed out, there are just way too many variables that 
effect the detection process to allow relatively error-free 
recognition.  Now if you hang up in the upper end of the bands, where 
the slower QSO's tend to migrate, you can probably do a reasonable job 
on signals when you S&P, as you can listen with the decoder several 
times to get the exchange of the other station calling CQ.  Your rates 
will never approach the level needed for serious competition.

   These devices are very useful to new contesters, who have little CW 
skill.  In many cases the devices help some to upgrade their 
proficiency.  In other cases, the device becomes  a crutch.

    My point though is don't think for a minute that the serious CW 
contesters use these devices to any great advantage, although  maybe 
someday that will happen.  Perhaps the military has developed the 
detection of CW down to the level of unattended operation, I really 
don't know. It is a real eye opener to attempt to copy stations in the 
fast paced, QRM ridden contest scene, using any of these devices. One 
thing for sure that shocks me is how bad many fists are in this day of 
computer generated CW, either by PC or by paddle and keyers. The human 
brain's amazing decoding capability really comes to light  when one 
attempts to copy in one's head compared with the various devices.  
Character spacing is by far the most common error.

  I don't wish to discourage anyone from using or developing these 
devices.   Again under perfect conditions, the development of any device 
to copy CW is easy, almost trivial. Just be in for a shock when you see 
just how ineffectual these devices are in the real world. ;)

  Yep those top contenders actually sit and copy those thousands of 
QSO's in a big contest.  I have watched a few of these ops, and indeed 
it is a humbling experience to see just how good these folks have 
become. I do not include myself into this group, BTW,  but I am always 
in there slugging.

    BTW, the best software that I have found for CW decoding is CWGET.

Brian K7RE
10,000 to 15,000 Contest CW QSO's per year.