ssbdaemon is a tiny wrapper script that takes cwdaemon output and turns this
into play commands allowing you to use a cwdaemon capable logger to generate
voice-keyer audio.
It does this by launching cwdaemon in debug mode, without connecting to a radio or sound card. The debug information from cwdaemon is captured by the wrapper and converted into play commands.
The ESC interrupt transmit is also captured and used to kill the currently running play command.
This wrapper is a simple bash script. Launch it on a terminal and it will stay in the foreground until you terminate it with Ctrl-C.
It requires access to cwdaemon, which on a Debian system is in /sbin/
It requires access to play, to actually play an audio file, but it's simple to
use any audio play command.
The script will fail if cwdaemon is already running.
It uses the default cwdaemon port. The default arguments sent to cwdaemon are:
--cwdevice=null --nofork -i --system=n
If you supply any parameters, you must also supply --nofork and -i.
In my shack the audio follows the same path as is used for WSJT-X. The radio is set to Digital VOX and sending audio will automatically set the radio to transmit. You should be able to use an alternative TX mode if needed, either by setting up your logger, or by enabling radio control in cwdaemon.
This voice-keyer requires a directory full of audio files. The files need to be named in the same way that your logging tool sends commands to cwdaemon. Using TLF, output of callsigns is in UPPER-case, but this might not be universal and both UPPER-case and lower-case versions of the individual word files are usable.
The script doesn't care which audio format you use, as long as you can play it from a command-line. The default extension is "aiff", but it's trivial to change.
For every message that your logger sends to cwdaemon, the script checks to see if there is a file with that name. If your logger sends out "VK6FLAB", the script will attempt to play "VK6FLAB.aiff" - note that the file name must match the case for a message file to play.
If that file doesn't exist, the script splits the message into individual words, if VK6FLAB.aiff doesn't exist, the script looks for: V.aiff, K.aiff, 6.aiff, F.aiff, L.aiff, A.aiff and B.aiff and plays those in order.
The file "A.aiff" needs to contain the audio for the phonetic word "Alpha", "B.aiff" needs to contain "Bravo", etc.
The script checks for a file name match in the following order:
- "VK6FLAB 5929.aiff"
- "VK6FLAB.aiff"
- "5928.aiff"
- "V.aiff"
- "v.aiff"
- "K.aiff"
- "k.aiff"
- ...
The following characters are translated:
.= decimal/= stroke?= query
You can record the audio yourself and process it using a tool like audacity. You
can use the MacOS say command and output the text to aiff like this:
`say "Hello World" -o "Hello World.aiff"`
To create a file capable of tuning your radio, name it "tune.aiff". You can use the following command to generate a 10 second file with 1 kHz tone:
`sox -V -r 22050 -n -b 16 -c 2 tune-test.aiff synth 10 sin 1000 vol -1dB`
Tip: Name the files for their phonetic content and make a symbolic link to play them. For example:
`ln -s "Alpha.aiff" A.aiff`
`ln -s "CQ Contest, Victor Kilo 6 Foxtrot Lima Alpha Bravo.aiff" "CQ.aiff"`
The following known issues exist:
- Only three of the features of cwdaemon are handled: request, tune and abort.
- WPM has no effect on the audio output.
- There is no volume control.
ssbdaemon was written by Onno (VK6FLAB) Benschop [email protected] after discovering a distinct lack of Linux based voice-keyers that integrate with logging software.
This script is a companion to Episode 333 of the "Foundations of Amateur Radio" podcast titled: "How to run an SSB contest without using your voice...".
See https://podcasts.vk6flab.com/ for details.