Hello friends,


my comments to the discussion WSPR vs. CW:

(1) Modes only can be compared at a specified data rate. 
Of course, CW is "better" than every thinkable digital 
mode if we allow it to be slowed down until something 
can be identified on the screen. What really counts is 
the duration of a random QSO.

(2) Modes only can be compared at a specified error rate. 
Try to communicate completely unknown messages! The 
symbol error rate of CW at it's ultimate snr is 
relatively high. Getting confidence in a callsign 
needs a low symbol error rate, i.e. a good snr.

(3) I am working on a digital mode for minimal QSOs. 
In simulations with Gaussian noise and Rayleigh fading 
it is 1 dB above the Shannon limit. On-off-keying as 
used by CW by principle cannot reach that.


Wishing the best for 2009,

73 de Klaus, DJ5HG