Got my radio back on the air after six months off. I’d decommissioned the ancient ThinkPad that spoke to the radio, and hadn’t worked out what I’d needed to get it running with the MacBook. All it took was a FTDI-based USB to serial converter for the RigBlaster Pro’s PTT line control, and all works perfectly!
Fldigi used to work fine, but recent updates may have caused me to drop off the face of the (radio) earth. What it seems to be doing — and I don’t find this at all plausible — is causing interference with its own audio stream when its window has focus, but receiving perfectly when the program window is hidden. As Fldigi is a highly interactive program, this is not much use. Here’s an audio sample showing what I mean: fldigi-psk14070-VA3PID-201206092107z. It’s about 45 seconds long, a sample of the 20m PSK31 band this afternoon, and comprises:
- 0-15 seconds: fldigi’s window is in focus. None of the traces in the waterfall resolve to meaningful text.
- 15-30 seconds: I changed focus to another program. The waterfall traces snap into focus; QSOs become readable. The conversation at 2383 Hz goes from line noise to a very clear “… 73 73 Jim and thanks for ans[w]ering the cq …”
- 30-45 seconds: fldigi’s window is back in focus, and all decoding is cut off.
I’m running fldigi 3.21.43-1~kamal~precise from the Ubuntu Amateur Radio Software Updates repo. Hardware is a Thinkpad R51 (a bit old), latest Ubuntu 12.04 LTS, a FT-857D into a Signalink USB, and the audio’s being handled by PortAudio. I’m stumped!
Update: It was a volume thing. Linux had decided that I didn’t need my main system volume above 10%, so fldigi was picking up noise only.
Last night I picked up a weird signal on 14.074 MHz that wasn’t quite Olivia, and wasn’t quite Contestia. After a while I worked out it was MFSK16; but only after I started calling CQ nearby in Olivia and the original caller, Don (W5VRX) in Arlington, TX, told me what it was.
Don told me that MFSK could send pictures, and that Fldigi supported it well. It’s a bit slow, can be subject to noise, but you can get quite clean results. Here’s what Don sent me: