User name:
Password:

Forgot password? - Sign up

2GA49 BG3GTW BI1PVU biggus dd1lc df9tm EA5CPR F4FOA HETER ik3txt IU0RBF IU2PES IU5RCB iz8bir JF6CLO Liy mapleice Medved MorsetrainingF mrodrigues myles Nathan83 R4NCU RL7LBU sidu skmttty sm6oqx so7na SP9ALZ test ut4ubz

Just managed to make some time and get back to practicing. My 20/15; 1 min groups of 5; 2 min groups of 5; 1 min random groups; 2 mins random groups plan is still being kept to and I've completed the first 4 letters.

I'm having trouble writing fast enough, although I'd set it to 15 wpm effective speed. After a lesson completes, you can see the 'real' effective wpm which depend completely on which letters are now involved: in lesson 3 (K,M,U,R) the effective speed is around 16.8 wpm because the letters are, presumably, a bit shorter than the average across 46/47 characters. So I'm actually attempting 20/16.8 !

I've finally stopped practicing those and gone on to lesson 4 (adding E). Yep, the new effective wpm is roughly 19.5!

This is probably a good thing, and will help me do reflex responses rather than thinking, but I thought it might be worth pointing out if some people are getting stuck at lesson 4 and thinking they'll never succeed in learning CW - it could just be that the new letter is tiny and your effective wpm is way higher than you planned.

Anyone out there on a MacBook, remember to turn off your spelling auto-correct too: once you get enough letters going, the spell check starts 'fixing' things as you type them in, which won't help your scores :-) Click in the text box, then menu item Edit/Spelling and Grammar/Correct Spelling Automatically.

HTH

Back to 20/19.5 !!

The so called real speed, must be equal to the required writing speed, I understand.

When you complete the course the required writing speed is about 75% of the Morse PARIS speed, due to the code groups being random.

Yes, the effective speed is calculated from the number of characters sent and the time it takes to send them, and when a short letter is the new one (or the mix is weighted towards shorter letters) then the effective speed can come out a fair bit higher than the speed you've asked for. The reverse also applies, of course, once there are longer-than-average characters in the mix.



Once, way back, a colleague inadvertently allowed a piece of software to "correct" the spelling in a technical document that had taken a month to write. Of course, the software didn't know most of the technical terms, so it just replaced them all with words it thought were close enough. The result was hilarious to read, but not technically much use...

OK, given up on a fixed 20/15 as the letter E killed me: I had to write so fast, and if I thought about the character at all I missed the next few words groups.

Reducing the effective words per minute to whatever it takes to get a 'real' 15 wpm is my next plan. I've found 20/10 is needed for lesson 4 and 5. When I get onto lesson 6, it may have to come up to 20/11 as N is a bit longer.

I'm also binning the random length words for now - apparently Koch stuck to 5-letter groups the whole way.

I'm also trying to go for longer (in minute terms) before proceeding tomthe next lesson: if you only do 1 or 2 mins, you might never hear the new character at all, and that's the one you need to practice most.

I've also found that my writing has cost me a few errors so far :-) when I try and copy and then type the resulting scratches back into the input window.

Good luck, all! I'm going to kill lesson 5 off today!

The current random approach makes the lessons less and less effective for learning characters as you get further up the trail. A fixed biased distribution (so you always get N of the new character) would work better, but is probably a bit trickier to program.

With the current scheme, the later lessons are really only useful for testing whether you've learned a character, not for teaching it.

I agree. It wouldn't have to be that hard to make a difference: for each character choice, a random 1 out of 10 gets you the new character; if not, another random 1 out of 20 gets you the last new character; if not, choose randomly from ALL active characters.

There is some weighting, apparently. Look for Fabian's posts in http://lcwo.net/forum/574/Decreasing-the-frequency-of-old-letters