Second build has shipped. Message will appear here when store is ready for third build ordering.

WSPR: Remote WSJT-x input from Kiwi audio output

On busy bands the Kiwi's WSPR decoder can't fully process the 2 minute capture, so I have been opening a browser window on a PC/Mac and feeding the audio output through a virtual audio device into the WSJT-x. This is a complex and fragile SW architecture and I wonder if it would be possible to simplify it by running 'ncat' or other utility on the PC and feeding it to a virtual audio device or a capture file for processing by WSJT-x.
Suggestions would be welcome.


  • jksjks
    edited May 2018
    The "kiwiclient" Python-based software exists ( It is currently used to connect to a Kiwi and store IQ/mono samples into a .wav file (with/without GPS timestamps), or process a FAX and store in a .png file or process waterfall data. Perhaps a mode can be added where it acts as a pipeline filter for use with ncat. Can WSJT-x take IQ input from standard input and not just a sound card device?
  • I have cloned from to Win7 and Linux. I'm not a Python expert but assume I can modify the makefile for the specific function I want to run (record, fax etc) and the kiwiserver address and port, frequencies etc in line with the examples and --help info. I have tried this: python makefile [edited] but get a syntax error. I was trying to record an iq file from two KiwiSDRs, including one of my own. I suspect I am making a fundamental error! Help would be appreciated (Terry)
  • edited June 2018
    Thanks, records a wav file.
    However I want to feed audio to WSJT-x. So to do that it appears that I will need to modify
    Am I correct?
    Ideally I would have kiwirecorder sync up to the beginning of a two minute window, capture that to a wav file, then execute the K9AN wspr decode lilbrary. Perhaps I could borrow that from jks's auto-wspr in the Kiwi ;=)
Sign In or Register to comment.