In the context of shortwave listening I am also developing a custom signal analysis software which I also plan to use with my software defined radio later. This is actually the third attempt to develop such a software (not counting the early approach which was not intended for shortwave signal analysis but just a test of an audio library).
The first was just a quick hack to get a waterfall of an audio signal and got very clumsy with time. The second one was more sophisticated in structure and features, but got performance problems as visualization and analysis ran in the same thread. Thus I went back to the scratchpad and redesigned the concept to be even more modular, and this will be the outcome.
- using libfftw for fft gives a certain speed
- separation of analysis and visualization to use multicore systems more efficiently
- detection and tracing of signals
- analysis core can be fed via and save to stdin/out, unix sockets, tcp (even with ssl) or files (maybe more to come).
- adjustable dynamic range of waterfall and spectra, even after waterfall has been “recorded”
- ringbuffer and linear buffer to allow on-demand scrollback of waterfall
- more to come (filters?)
If you want to follow the development, check out the github.com repositories:
No screenshots of the actual tool yet, as the visualization frontend is not ported to the new structure yet. However, let this image give you an idea about how it might look later: