Any tips for learning?

[IZ2UUF]

After lamenting my apartment situation for over a year, I finally decided to buy an apartment antenna and try again. So, I finally have a good excuse to learn CW.

Hello dear friends.

I would like to share my experience regarding CW because it has been quite "extreme".
When I started learning Morse code I could merely spell a few letters, nothing more.
After having read all I could find about mastering CW, and taking in account my personal character, I decided to go for the one that I believed would have given me the best results: the Koch method.
So I started excercising every single day: maybe thirthy minutes, maybe five, but regularly.
I kept 20 to 25 wpm and I stayed on a letter set until I could reach an accuracy of 90% at 20WPM with normal spacing; then I moved to the next letter.
I never tried to copy real CW from my radio, because with the Koch method, it is very frustrating: every time an unknown letter is copied, the "comprehension flow" will block immediately.
I never ever used the computer to decode CW at any time: not for of some rules of honor or moral virtuousness, but simply because I did not want to waste any time having my brain learning to "peek" into the computer monitor instead of decoding itself.

Day by day, after about ten months, all the 40 Koch letters were learned with 90% accuracy.

Reached this goal, I begun tuning on the CW bands to see what I was able to copy.
Just listening, I could barely tell that someone was calling CQ.
But with a sheet and a pencil, I found myself able to copy most of what I could tune at. I can't say I was actually "understanding" anything, but those sound made me kind of "feel" that I had to write those letters.
I can't tell you the emotion and the surprise when, after almost an year writing down random letters generated by a computer, on my paper appeared a message that made sense and that was sent by a human being!

At the beginning, in order to understand, I had to write down and read everything, including "UR RST IS" or "73".
But it didn't take long before I begun to recognize the standard QSO parts as standalone symbols. Still, In a casual QSO I have to write down what I ear and then read back what I have written, otherwise I don't understand anything at all.

Regarding transmission, with a perfect timing, a Begali paddle appeared on the second-hand desk of a nearby HAM shop.
As soon as I had this paddle at home, I plugged it in my radio and I prepared a setup with a computer software decoder that could read what I was typing. I never used a paddle before nor sent any Morse with any other mean, so I couldn't wait to try.
I was afraid it would have taken another long time to learn TX, but with my surprise, I discovered that the conditional reflex works both ways: not only when I ear a given sound I'm conditioned to write the related letter, but also when I saw a letter I recalled the sound that was related to it! It works in this way: when I see a letter, I immediately know what is the sound I want to ear, so I try to reproduce it with the paddle.
That first time, with my great surprise, I spent maybe three hours transmitting newspapers, books, numeric tables, making very few errors at 20 WPM.

PROS:
- very fast (I do not thing there are many other techniques that can bring anyone from zero to 20-25WPM in such a short time)
- being used to copy random letters, you can copy anything, not only "standard" QSO
- conditioned reflexes require very little "CPU-time" from our brain: I found myself writing down a QSO while speaking to my son;
- transmission comes almost for-free;
- even sparse 2' sessions are effective;
- you never experience 5 WPM wall (actually, you never experience 5 WPM at all);

CONS:
- it requires an exceptionally high grade of persistence and commitment: this is for sure the greatest obstacle for many people
- it is useless until you are finished. 50% done doesn't mean that you can copy 50% of QSOs: you can't copy any. So, if you give up at 80%, you wasted your time.
- it requires writing down everything: this is temporary, because with some real-radio experience, it becomes soon limited to the callsign, name and little more.

73 de IZ2UUF
Davide

[M0LEP]

Clearly the Koch random character progression method works for some. If it works for you, that's great. It does not work for everyone. Personally I think most of the time (well over a year) I spent trying to learn Morse using it was time I wasted. I should have looked at the statistics LCWO threw back at me a lot more critically, and gone elsewhere much much sooner.

[IZ2UUF]

Clearly the Koch random character progression method works for some. If it works for you, that's great. It does not work for everyone.

I would rather say that it works for a minority: at my local HAM club, I'm the only one that succeeded.
Most people simply aren't able to sustain this commitment for enough time to have a result: so I agree with you, it is time completely wasted.
I have not been able to "study" any case of anyone that put my same effort without obtaining my same results, so honestly I can't tell whether, besides the commitment, this method could work with anyone or not.
What I know that at my club there are a lot of people that know "M" and "K" very well. 

73 de IZ2UUF
Davide

[AK4YH]

Hello Philip,

At 5wpm you will be tempted and most likely use a look-up table in your head, which will slow you down a great deal later. I highly suggest learning at least at 12wpm with 20wpm characters. That will give you a much better base to build upon. If it's too fast for you, reduce the number of letters. Also see on lcwo.net: "Morse Machine." And try to go as fast as you can.

I made the same mistake you are making now, and after ten months, I am only at 15wpm. Only now am I able to transcribe directly from sound for most letters. 5wpm makes things easy now, but you will pay for it later...

Gil.

[M3KXZ]

I also have an interest in learning CW . . I found lcwo and some apps for the phone/iPad etc. I'll work on those. My question is "what is a good way to learn to send"? I know listening and transcribing is one thing, but sending might be a whole other thing to learn. And should we just start with paddles? I know someone will say straight key out of a mentality of "Thats what I learned on or that how we used to do it" . . But honestly, in 2013 what's the best way to learn to send? And learn with what device?

Straight key all the way for sending. It's NOT what I first learned on. I started with paddles. BUT subsequently I found that physically keying each character with a straight key reinforced what I was wanting to send, which in turn has helped me ever more with receiving. Everything seemed to connect together more strongly and my receiving speed picked up really quickly as a result. It's all to do with rhythm and mimicking the sounds of the words. Just like listening to music and tapping out the rhythm, the music gets into your head. Tapping the rhythm on a table with a finger is easy. Imagine though trying to tap out a rhythm with a paddle where one side is a short note and one is a long note.... much harder.

[M0LEP]

Most people simply aren't able to sustain this commitment for enough time to have a result: so I agree with you, it is time completely wasted.

It's not just a matter of commitment; I put the time in (several hundred hours over a year and a half or so), but (with the odd exception) each new character took more time to learn than the one before. Sometimes it was only a bit more, but sometimes a lot more. My early estimates were that it'd take me three years to get the characters learned. By the time I'd got to lesson 12 I'd revised that estimate to nearer ten years. For me, it was far better to make sound files of "A B C", "1 2 3", "AA AB AC", "AA BA BC" and so on, and listen to them repeatedly. At least that way I got all (or most of) the characters at least part learned fairly quickly, and I could try to make sense of QSOs on air.

[K5UNX]

Clearly the Koch random character progression method works for some. If it works for you, that's great. It does not work for everyone. Personally I think most of the time (well over a year) I spent trying to learn Morse using it was time I wasted. I should have looked at the statistics LCWO threw back at me a lot more critically, and gone elsewhere much much sooner.

So did you learn using another method?

[M0LEP]

So did you learn using another method?

Pretty much the same way I learned to read and write. Progress is slow, but fairly steady.

[AK4YH]

For sending, just read a book and send what you read.. You can check with Fldigi if it decodes your code. Good way to get your timing right..

Gil.

[KD8POH]

Just Learn Morse Code has worked well for me. As others have mentioned, it is best to paste in an actual paragraph of text, so that you are not just listening to random letters. The great thing about using a computer program is the immediate feedback -- you know which ones you missed and need to work on next time.

[M0LEP]

The great thing about using a computer program is the immediate feedback -- you know which ones you missed and need to work on next time.

...but only if you can type it straight in. That was another mistake I made; trying to type in the Morse I heard. I'm not a good enough touch-typist. It was far better for me to write out what I heard long-hand, but I wasted about three months trying to type in Morse before I realised that. Of course, once I'm writing out what I hear in pencil on paper, the only thing the computer's giving me is the Morse to practice from. LCWO lets you type in what you've heard for checking, but JLMC doesn't (or didn't last time I tried it).

The second catch with pasting in text is that you already know what you're going to hear. G4FON's random QSO files get around that one, but there comes a point when you're better off listening on-air.

It'd would be really cool if the Koch-style programs could generate sensible text rather than random groups from the set of characters they're currently trying to teach. Wouldn't work much below about lesson 15, I expect, and would probably be an interesting programming challenge...

[AC2EU]

The great thing about using a computer program is the immediate feedback -- you know which ones you missed and need to work on next time.

...but only if you can type it straight in. That was another mistake I made; trying to type in the Morse I heard. I'm not a good enough touch-typist. It was far better for me to write out what I heard long-hand, but I wasted about three months trying to type in Morse before I realised that. Of course, once I'm writing out what I hear in pencil on paper, the only thing the computer's giving me is the Morse to practice from. LCWO lets you type in what you've heard for checking, but JLMC doesn't (or didn't last time I tried it).

The second catch with pasting in text is that you already know what you're going to hear. G4FON's random QSO files get around that one, but there comes a point when you're better off listening on-air.

You can do that with the "Just learn Morse code" program.
http://www.justlearnmorsecode.com/

It'd would be really cool if the Koch-style programs could generate sensible text rather than random groups from the set of characters they're currently trying to teach. Wouldn't work much below about lesson 15, I expect, and would probably be an interesting programming challenge...

[M0LEP]

You can do that with the "Just learn Morse code" program.

Fair enough, I'd forgotten JLMC did that too. I used G4FON because I liked the controls it has for messing with the sounds it sends, which makes the Morse it produces sound a lot more "real" than LCWO or JLMC manage.

[N5RDE]

For beginners, I highly recommend Chuck Adams' code course.  It has over 20 hours of practice materials and is free from several sites on the Internet.   It is in the form of mp3 files so you can use it on your computer, and s .pdf manual is included.  It is very well done.

I believe you can also obtain it from the FISTS site.

Just search for Chuck Adams, K7QO.

[MM0ZBH]

AA9PW iPhone app is good for a few minutes here and there during the day.  I am sending the letters on the app slighter faster than I can decode in my head and it allows me to push forward.  Currently sending at 15wpm spacing and I struggle but it pushed me past 12wpm spacing and that now seems easy.....just pushing my brain and it seems to work.