News for hackers. Stuff that matters.
Re:Receiving images of Earth from satellites with software d... by looklookatme on Tuesday June 09, 2020 @05:52PM (#23471346)
> Once you start playing around in GNU Radio and decoding more exotic stuff

Do you have any tips or resources on how to get to a competency level where you get to decode arbitrary digital signals you find from a scan?

I've read through all the GNU Radio documentation, along with the examples on their wiki and have even written my own blocks but keep finding that opening a grc file that someone else authored, there will be a critical block I've never seen before (and likely, never see again) which hampers experimentation. As soon as I attempt to decode a real-life signal, it feels like I dove in too fast -- there's too many unknowns, since the end result is [presumably] not going to be a nice, human-readable string. However, occasionally, I find challenges like [1] which are great, since you have the expectation that the signal _can_ be decoded; it's just that I've found examples like these to be quite rare.

[1] https://www.gnuradio.org/blog/2018-02-21-gnu-radio-challenge...

Re:Receiving images of Earth from satellites with software d... by jcims on Tuesday June 09, 2020 @06:48PM (#23471811)
Not the person you asked but once gnuradio flowgraphs started to function materially outside of basic flowgraph blocks stitched together, or worse, wrapped by some python script, I would have to tap out.

It's one thing to demodulate some manchester encoded OOK/FSK signal from some ISM-band hardware monitor vs. 64-QAM or GMSK. At some point it, to me, becomes indistinguishable from magic. Fluency in DSP is, IMHO, mandatory to be broadly effective with GNU Radio.

Re:Receiving images of Earth from satellites with software d... by timeinput on Tuesday June 09, 2020 @09:18PM (#23472841)
I'm fairly fluent in both DSP and software, and find GNU radio indecipherable.

A plain C/C++ implementation is often so much clearer than a flow graph built in python.

Re:Receiving images of Earth from satellites with software d... by jcims on Tuesday June 09, 2020 @09:56PM (#23473073)
Copy that. By far the most informative bit of dsp code for me was the core libraries in SDRSharp.