Issue #24: Cool Devices to Learn CW

I can specifically vouch for the “Morse-It” iPhone & iPad app. I’ve been using it to try to get my own code speed up to around 15 _ 18 WPM by using it as a “crutch” while listening to W1AW’s propagation forecast & DX bulletins on 80 meters. It works well on a strong signal such as W1AW normally is here my QTH is in Southeastern Massachusetts). While listening to W1AW, I try to mentally predict what the next character is before it pops up on the iPad screen, & when the app stumbles at decoding if W1AW takes I fade, I can frequently copy it by ear. The propagation forecast bulletins are especially good due to having lots of numbers & punctuation characters in them. I find it’s definitely helping my ability to copy by ear. Just my 2 cents worth.

Tom WA1LBK

 

Code readers ........... Why bother - just learn morse code. There are absolutely NO short cuts other than increasing the amount of practice you do. 'Copying Behind", Instant word recognition, head copy, blah blah etc.,

Given there have been many, many thousands of merchant marine operators trained to read morse in many colleges around the world over the last 100 years - well up to 2000.!!. Many more military operators (airforce, army &navy) trained in military establishments around the world all using roughly the same method. Then there are the many thousands of instructors who've spent, in many cases years teaching morse code. Let alone the huge numbers of operators who have been trained. Not one had a eureka moment and invented a quicker method....

https://davidwperry.blogspot.com/2022/04/learning-morse-in-royal-navy.html

 

And all that happened before Moore's law allowed most people to carry tiny supercomputers in their pocket. Which is a better and quicker way to communicate via digital modes (including ITU timed Morse Code). Ask yourself why did the U.S. Navy and merchant marine, with their mission critical communications needs, obsolete all their CW stations after training those thousands and thousands of operators?

 

Ans: Financial reasons and efficiency.

The merchant marine, gradually made redundant by INMARSET and https://www.gov.uk/maritime-safety-...he-global-maritime-distress-and-safety-system (GMDSS)

Likewise, the RN and I assume other navies (like yours) simply gradually replaced CW and RATT/RTTY with more modern methods of comms. Satellite systems were already being used or trialled in the RN as early as 1969. The traditional 'RO' job, including morse, was largely phased out and operators were - if they wished - retrained as computer / communications specialists and so on. Our RN never made anyone redundant as far as I recall.

PS: I've just realised your may be commenting on the sentence about inventing a quicker method = I was commenting on inventing a quicker method of teaching morse - not quicker communications...

 

The issue with electronic decoders is that the software is not intiutive and must be tuned to both the speed and the audio frequency.
While this can be done be eye and ear its not perfect.
Electronic latency even if the speed is set correctly still will lag behind the mind and trained ear.
Apps abound and my favorite is morse mania because it simulates static noise along with the tone.
But i recommend earphones.
Unless you use it to drill with friends.
( dont know why but some people get irritated by morse code)

 

As someone who has written an automatic CW decoder, my experience is that noise or static is one of the easier signal corruptions to remove. If the signal frequency is constant, a simple FIR filter can reject most of the broadband static noise. Much harder is:

1. Fading: to know the difference in amplitude between the tone and no tone, you need a threshold level. If the channel fades, you don't know if the signal is actually no tone, or if fading has occurred.

2. Nonuniform dits, dahs, and element spacing. Especially with straight key, there can be huge variation in dit, dah, and element spacing as much as by a factor of two.

3. Frequency wander of the CW signal, as this can throw off filters tracking the tone, though for most rigs (especially using a TCXO) this isn't a problem.

More intelligent software typically either uses some sort of context clues (for example CQ is a common pair of letters that is run together) and perhaps something like a Markov chain to weight the various possible decoding probabilities. I think CW operators also underestimate how often they get decoding wrong or use retransmission (either requested or by repetition by the source) to get the message through. There is also other context, for example, an operator trying to land a POTA or field day station would be listening to other QSOs and therefore have an idea of the calling station's callsign and other information already before initiating a QSO. Also, a lot of QSOs have an extremely constrained vocabulary and well-known abbreviations which would be incorporated into the probability matrix of a Markov chain.

While generally such context clues are a good thing, they might be a problem if a rare or unusual communication occurs (for example an emergency break in) as it might take a while for an operator or software to determine that the communications are not to be interpreted as a contest QSO or other traffic.

73,
Dan
KW4TI

 

Your ORGANIC BRAIN.

DECODE ONLY.

NO BS SOFTWARE .


K8PG Paul Morse Music 73’s