NOTICE: I have released a completely new version 2 of the IZ2UUF Morse Koch CW application. The new application is currently at the stage of open beta testing. If you wish to participate as a beta tester, go to https://www.iz2uuf.net/cw and join the Beta Testers program.
I am a professional software developer and I spent a huge amount of my time writing and debugging code.
About twenty years ago, I was called by a company to teach C++ to a group of their C programmers. During that course, it happened something that I never noticed before: when they were writing their simple coding exercises, with a simple glance I could list all syntax, semantic and logic errors contained in the source file they were writing.
I was very much surprised by that, because I realized that I was not reading their source files: I kind of “felt” the presence of those errors before even starting to read the screen.
My brain was so trained in associating those coding patterns to the related compiler errors that it was doing it without even me paying attention to it.
That event, being so unusual, attracted my attention: however, we all use this mechanism every day. For example, when we ear people speaking in the background, we understand what they are talking about even if we are doing something else and no paying attention to them.
What we develop to obtain such results are called conditioned reflexes.
I believe that the need for strong conditional reflexes is unquestioned for effective CW operations. So next step is: which is the most effective way to develop such reflexes?
First step, we have to determine what we want those reflexes to be able to do. Reasoning on my experience, I found that I was able to recognize the single letters of the source code, words and syntactical constructs made of such words (for example, a function, a code block); actually, my brain was able to recognize even higher (i.e. more abstract) combinations, such as logical errors.
I am sure that a good CW operator reflexes are able to decode letter by letter, but also whole words and even whole concepts. After all, all of us are able to read single letters but also whole words at once, if not even whole concepts if they are stereotyped.
We certainly have developed such an ability by reading and reading and reading for all of our life. Hovever I am well above my 40’s and I have no time to do this all over for Morse code too.
So I decided to settle for a less ambitious goal, i.e. being able to recognize and type letter by letter at a reasonable speed.
The Android application
Methods exploiting this ideas exists since decades, like the one proposed by German psychologist Ludwig Koch. However, to be real effective, these methods need a computer software able to produce really random sequences at calibrated speed.
On Windows I use G4FON’s Morse Trainer, which is very good.
However, I really felt that there were many times where I had five minutes that I could invest in learning CW, but no computer handy. This is why I wrote the IZ2UUF Morse Koch CW Android application.
This application can be downloaded for free from the Android Market (now Google Play) at this link: IZ2UUF Morse Koch CW.
Using IZ2UUF Morse Koch CW
The Android app is very simple. By pressing the blue arrow “play” button, the application starts playing a random CW sequence on the louspeaker or headphones; in the meantime, the characters played appear on the main screen.
Playback stops when the timer is expired (default is 5 minutes) or when the red square “stop” button is pressed.
The method is simple: start with two characters, for example “K” and “M”. The program plays these two characters only and you just have to recognize and write them down.
Initially it will be hard, but in a few session you will start to understand better and better; do not slow the playback speed, or it will not work! Just enlarge the inter-character spacing or, better, the inter-word spacing.
When you are able to write down at least 90% of a 5 minutes sequence, add another letter and go on.
Your proficiency will drop, but it will be back again very soon.
The application has several configuration options.
This sets the words-per-minute, using the standard morse speed rules.
This should be set at least at 20WPM, better if more.
This value, starting from version 1.0.012, can be changed also with a slidebar on the main screen.
This settings allows characters to be spaced more than the standard 3-dots time. This allows an extra thinking without reducing the speed of the single character.
In this way our brain can still learn how the characters sound like at high speed and not slowed down. This implements the same concept of Farnsworth timing used by ARRL.
This setting allows higher separation between words. In this way we can train on small sequence of letters at high speed and with normal spacing, but still having time to write them down at the end of the word. I found myself enlarging word spacing more effective than enlarging character spacing.
Sets the pitch of the CW note.
Sets an higher pitch for dots. This is another learning technique that uses this artifact to easen learning sequences of dots-dashes. The pitch difference should be reduced as soon as the learner’s proficency gets better.
This allows setting a pause before the first character; in this way, we have time to settle down and become ready!
New feature! When this option is selected, after each word is sent in CW, a synthesized voice will read each character of the same word
using the ICAO standard spelling (alpha, bravo, etc.).
After having listened the sequence of CW characters, the student repeats them mentally using the ICAO alphabet; after a programmable delay, the voice will repeat the same characters for a check.
This feature has two goals:
- training hands-free: this exercise can be done while walking, driving or anything else that reduces hands availability;
- training to quickly remember a few random characters; this is extremely useful to rapidly catch callsigns and other pseudo-random words (locators, names, etc.).
Time before speaking
Time allowed between the end of CW and beginning of spoken words.
Time after speaking
Time allowed between the end of spoken words and the beginning of the next CW word.
Length of the exercise
Set the length of each session. Koch method sets it to five minutes, but I find better 2 minutes on very fast sequences, because after that I start really messing up.
This allows you to choose among the following options:
- G4FON sequence – play a random sequence of characters taken from the current Koch level (see below) using the G4FON program sequence
- LCWO sequence – play a random sequence of characters taken from the current Koch level (see below) using the LCWO site sequence
- Chars from custom string – play a random sequence of characters taken from the string set in the “Custom string” (see below) option;
- Custom string – play the string set in the “Custom string” (see below) option; this allows playing any text.
- Words from custom string – randomly play words taken from the “Custom string” (see below); use “_” (underscore) to join sentences in a single word (e.g. “MY_NAME_IS” will be played altogether as a word replacing underscores with spaces)
- Words from custom string and G4FON – same as above, but some words will be randomly generated using the G4FON squence.
- Words from custom string and LCWO – same as above, but some words will be randomly generated using the LCWO squence.
- Koch training on words – New feature! – this feature takes the words listed in the custom string (or loaded from a file) and plays them randomly. This exercise is very useful to memorize entire words at once at high speed.
Koch defined in which order characters should be learned. With this setting, you can choose how many Koch characters you want to use.
The “hard letters” feature allows to mark the letters that are harder for us to understand. The program will insist on them, proposing them with a much higher frequency.
Auto hard letters
If set, the last two letters of our current sequence are treated as “hard letters”. In this way, we get an higher frequency for the letters that have been just added to the sequence.
This is the custom string which use is explained under the “Character sequence” option.
Note that this string can contain prosigns like <BT> and accented characters.
The custom string can be set by loading a file on the SD card. This is done by pressing the file-open icon in the main screen.
Files must have extension “.TXT” and they must be saved in the SD card, under a directory named “cw” (/sdcard/cw).
My personal tips
Reflex jam – When you are working on a set of characters you know very well, you will see that you will able to write them in a continuous flow; but as soon as you encounter a new character, this mechanism will jam and you will not be able to recognize the following characters, even if well known, until you get to focus on it again. My tip is: move to next Koch character only when you get very little jamming on the ones you know: in this way, when you “jam” you know that it must be the newly added character and you do not have to think on it.
Spacing – initially character spacing (see Farnsworth timing) is highly needed, otherwise it would be very frustrating; however, I found very useful to progressively reduce that spacing down to 1x and using the “word spacing” setting. This gives me a little time to think over what I heard but in the meantime I am training actual full speed sequences, including their normal spacing. This is also an exercise to temporarely memorize small sequence of random letters, useful for callsigns and locators.
Garbled audio (solved)
UPDATE (25-Jan-2013): I have been able to setup a device showing this problem. Audio doesn’t work on Android 4.2, while it works ok on 4.0.3. After some tests, I found that the ENCODING_PCM_8BIT setting doesn’t work anymore. Instead, I had to switch to ENCODING_PCM_16BIT, that seems working.
06-Feb-2013 – I would like to share with you this very kind piece of “ham spirit” that really touched me. Yesterday in my mail box I found a small carton box that contained a nice hand-written letter from DO2CW Thomas. Thomas thanked me for publishing this app, for promoting the diffusion of CW and gifted me with this:
It is a Russian “spy-key” combined with a single-lever mini-paddle homebrew by Thomas.
I tried both the vertical and the paddle and both have perfect and confortable movements. The little paddle is a tiny jewel of fine mechanics, really impressive! The entire assembly is very light but though at the same time, sure able to survive my backpack during mountainering with my FT817.
By the way, I was looking for a light key for portable use because every key I have is heavier than the 817 itself, never mind having along both the vertical and the paddle!
Thanks again Thomas!
- first release
- on back button execution terminates
- added accented letters (as defined in Wikipedia morse page)
- added custom strings playback
- added playback from custom charset
- added programmable word spacing
- fixed error when activating empty custom string
- replaced zero with slashed zero
- added 2.5x, 3x and 3.5x spacings
- added programmable pause before playback
- changed info page
- fixed a bug when using spacing 1x
- fixed screen rotation bug
- added different pitch settings for dot and dash
- added programmable word length
- (never released)
- hopefully fixed the garbled audio problem on newer devices
- Full support for LCWO sequence
- Added pause button
- Added fast button to immediately load custom files
- WPM speed can now be changed with a spinner at 1 WPM steps
- Koch level reports LCWO/G4FON series correctly
- Added morse dash/dot reminder beside “Koch level”
- Added file load option
- Custom strings now support prosigns
- G4FON sequence extended with prosigns
- Added support for user defined words
- Added scrollbar in the main screen
- Added speech support. When active, each word is read after a pause by a synthetic voice using the ICAO alphabet. This allows exercising without writing: first repeat the characters yourself, then wait for the voice for a check.
- New “koch word” exercise. Words from a file are presented randomly and have to be mentally recognized.
- Added user defined hard letters. Some letters can be marked as “hard to understant”: they will be presented with higher frequency during exercises.
- Added auto-hard letters. This feature automatically define as “hard letters” the last two letters of the current Koch sequence.
- Made moveable to SD-card
- When a file is loaded, its name is printed on the main screen
- Fixed bug: the number of letters used to compose a word were one more than the value selected in “Word length”.
- Speed now goes up to 100 wpm
- Added a WPM quick bar on the main screen.
- Fixed minor bug: welcome screen was reprinted when device was rotated.
Comments and suggestions
I hope you will find this application useful and please, do not hesitate to let me know your opinion or if you have any suggesion.
73 de IZ2UUF!