Subj : Yet another morse proggie
To   : Deuce
From : Angus McLeod
Date : Tue Aug 23 2005 09:14 am

  Re: Yet another morse proggie
  By: Deuce to Angus McLeod on Mon Aug 22 2005 23:22:00

 > ii) I'm currently using a computer-adapted koch-farnsworth method... the cod
 > is played full speed, then the program blocks until you press a key
 > (preferrably, the key that corresponds to the code you just heard.  :-)
 > If you get more than 90% over a 5 minute drill, it will add another letter t
 > the mix.  The Farnsworthy bit comes from the blocking between sequences.  I
 > does use the reccomended Koch learning sequence though, specifically:
 > KMRSUAPTLOWI.NJEF0Y,VG5/Q9ZH38B?427C1D6X<BT><SK><AR>
 > 
 > Which I swiped from a few different sites, including:
 > http://hfradio.org/koch_2.html

I'd recommend dumping Koch and sticking with Farnsworth.  When learning, 
you want to hear the characters at full speed with long think-time spaces 
between characters.  The character elements and the intra-character gaps 
are as per selected speed.  As the student improves, you close up the 
inter-character gaps.  A la Farnsworth.

But this random mish-mash of characters suggested by Koch is rubbish.  You 
start with groups of letters, with similar character elements.  These are 
the characters that people find hard to distinguish, so these are the ones 
they practice hearing and distinguishing.  EISH5.  TMO0.  AUV4.  NDB6.  
KX=.  And so on; I can't remember them all off the top of my head.

 > I'm thinking that after you complete the modified farnsowrth, it would go to
 > modified koch which sends a random-length sequence then waits for the whole
 > sequence to be entered.

My instructor (Sam) would teach us a group of letters, then drill us on 
them by sending individual letters and wait for a response.  Then he would 
start using letters from letter groups we'd learned earlier.  Then he 
would move on to short groups using all the letters but emphasizing the 
new set.  Then we would do words made up from the letters learned so far.

 > Following that, I was thinking of an ELIZA type program for the more 
 > "standard" CW conversations.

That's an interesting idea.  I can just picture it sending signal reports 
and talking about the WX :-)

I don't know how you're doing the tone generation, but here's a thought:  
Add some QRM!  Generate several other "on-air" signals tuned to a slightly 
different frequency and running quietly at slightly different speeds...

 > Instead of learning the differences between codes that sound similar, 
 > you're learning the sound of codes in general using this method I made 
 > more sense to me anyways. *shrug*

Well, it's a school of thought, I suppose.  But not a good one!  ;-)

---
 þ Synchronet þ CQ DX!  The ANJO BBS calling on 56K dial-up...

.