[CQ-Contest] computer Morse code recognition

Rob Locher W7GH rob at idiompress.com
Fri Feb 2 15:28:46 EST 2007

Hi Jim,

I'm younger than the average ham, and I took up Morse code recently.  I  
can tell you from personal experience that the method that you describe  
doesn't work very well.  I tried using the "CwGet" program on some DX  
pileups because the operators were going too fast for me to copy by ear,  
and the results were very poor.  CwGet seems to have a good reputation  
within its category, so I doubt that the problem was my choice of software.

The software deals with QRM brilliantly; a few Hertz of separation is all  
it needs to ignore an interfering signal.  However, when looking at the  
amplitude vs. time graph shown by the program (similar to an oscilloscope  
trace), it can clearly be seen that a lot of very-short-duration QSB  
frequently stymies the computer, which the "computer between the ears"  
handles effortlessly.  For instance, a "dit" will have its amplitude  
reduced by 50%, which will be below the threshold that the software will  
copy, changing a "D" to an "N".  (The threshold is adjustable, and I had  
tweaked it very carefully.)  What was RST 579 to my ear was more like 279  
for the software; several repetitions were required to understand what was  
being said.  To be fair, I think I was on 40m, but the QRN was very mild  
at the time.

It didn't take me very long to realize that computers are much better at  
comparing the amplitude of two tones to find which is louder (RTTY) than  
determining if a single tone is present or not (Morse code).

There is another annoyance: the software introduces a time delay.  If I  
remember correctly, it is between a half second to a second.  That is  
acceptable for a ragchew of course, but not for contesting or DXing.

I tried using CwGet a couple times, and that was enough.  I stopped using  
it and concentrated all my efforts on learning the code the old-fashioned  
way, and have been much happier with that approach.  Of course with more  
computational horsepower available in the future, computers will get  
better at copying Morse code.  However in my opinion Morse code appeals  
because it is fun to copy it by ear.  (Well, fun once it has been fully  
learned at least.)  If I wanted just a digital mode then I would pick one  
that that computers copy better, such as PSK31.

By the way, I have run into several other relatively young hams who are  
also learning Morse code.  There aren't nearly as many learners as there  
were in past years I'm sure.  In my opinion, fifty years from now there  
won't be nearly as many hams as there are today, and a smaller fraction of  
those will be using Morse code, but Morse code will still be alive and  
well.  I think that the main thing that will keep it alive will be  

- Rob W7GH

On Fri, 02 Feb 2007 06:39:09 -0800, <Jimk8mr at aol.com> wrote:

> Be wary of getting what you ask for.
> If there are to be any new CW operators in the future - unfortunately  
> not a
> certainty - it will probably be from people combining a code reader and a
> speaker. Watching and listening to repetitive patterns - CQs,  5NN OH,  
> your own
> callsign - will be the start of learning code as one  recognizes by ear  
> what
> one sees on the screen.
> I don't think many people will want to go through the work of learning   
> the
> code before they operate, but some just might learn the code while they
> operate.
> 73  -  Jim  K8MR

More information about the CQ-Contest mailing list