TETRA mode

[joannesberndt]

Where did your fork version of DSD-FME disappear? Could you add TETRA mode decoding?

[romanremus]

if TETRA could also be implemented, it would be very good, I am attaching some links about some tetra codes
https://github.com/osmocom/osmo-tetra
https://github.com/MidnightBlueLabs/TETRA_crypto
Unfortunately, in my area I only have a few DMR signals, but I hope that one day someone will take care of TETRA and make us happy too.

[arancormonk]

Where did your fork version of DSD-FME disappear? Could you add TETRA mode decoding?

you can find it here: https://github.com/arancormonk/dsd-neo

i view it as a more experimental version of DSD as i re-architect the codebase... it is currently not maintenance friendly at all and adding new features/changing things is incredibly painful. this is going to be a considerable effort, so be aware that it will not be stable at all until i get through several refactors... though, the first commit is basically a restructured DSD-FME that includes my RTL-SDR refactoring. functionality-wise it's the same... for now.

regarding TETRA, can you start a "discussion" here: https://github.com/arancormonk/dsd-neo/discussions? i'll close this issue in favor of GitHub discussions, is that ok? nevermind, i just realized you can move issues across repos... so i'll keep this open as a dsd-neo issue.

[romanremus]

I tried compiling the code in cygwin, but I got a lot of errors, can you help me?

[arancormonk]

I tried compiling the code in cygwin, but I got a lot of errors, can you help me?

to keep things organized, please create a new issue so as to not pollute this topic... https://github.com/arancormonk/dsd-neo/issues/new/choose

but i just compiled it in win11 cygwin and it should work if you follow: https://github.com/arancormonk/dsd-neo?tab=readme-ov-file#build-from-source (assuming you had DSD-FME installed in the same environment)

git clone https://github.com/arancormonk/dsd-neo
cd dsd-neo
cmake --preset dev-release
cmake --build --preset dev-release -j

compiled binary is ./build/dev-release/apps/dsd-cli/dsd-neo

if it's still giving you a problem, create a new GitHub issue with the cygwin compiler log and i'll investigate.

[romanremus]

Hello arancormonk, and congratulations for the code, it is superb, I have the latest version but in the Ncurses terminal SNR does not look exactly like it does in the code, do you have any idea? I have attached a picture

[romanremus]

[romanremus]

I think this symbol would work best with SNR, displaying a level, what do you think?

[romanremus]

[romanremus]

Good morning, I think SNR still doesn't work, but now those dotted lines have been enlarged, or is there a bug on my part?

[romanremus]

[romanremus]

I honestly don't understand how it works or how it displays the SNR, shouldn't it increase from a small line to the largest line as a function of the signal strength when it receives a signal?

[arancormonk]

Good morning, I think SNR still doesn't work, but now those dotted lines have been enlarged, or is there a bug on my part?

Fyi, this would be a good candidate for a discussion topic in https://github.com/arancormonk/dsd-neo/discussions/categories/general, not this TETRA issue...

i guess i need a better idea of what you're thinking about the SNR indicator, it is using those symbols you mentioned but they change "height" and color dynamically depending on how much the SNR varies (from 0 to 30 dB... which may be a unrealistic range since this is being calculated at the output of the demodulator).

also, don't expect any of this stuff to be stable... i'm adding a bunch of these visual indicators/candy so i can tune my DSP pipeline. until i tag a "stable" release things may break at any time... 😄

/* Render a compact horizontal meter for current SNR using existing glyphs. */
static void
print_snr_meter(const dsd_opts* opts, double snr_db) {
    /* Preserve the current color pair so our temporary colors don't clear it */
#ifdef PRETTY_COLORS
    attr_t saved_attrs = 0;
    short saved_pair = 0;
    attr_get(&saved_attrs, &saved_pair, NULL);
#endif
    /* Color bands match sparkline thresholds */
    const short C_GOOD = 11, C_MOD = 12, C_POOR = 13;
    /* Map 0..30 dB onto 8 glyph levels (same set used elsewhere) */
    static const char* uni8[] = {"▁", "▂", "▃", "▄", "▅", "▆", "▇", "█"};
    static const char ascii8[] = ".:;-=+*#"; /* 8 levels */

    double v = snr_db;
    if (v < 0.0) {
        v = 0.0;
    }
    if (v > 30.0) {
        v = 30.0;
    }
    const int levels = 8;
    int li = (int)floor((v / 30.0) * (levels - 1) + 0.5);
    if (li < 0) {
        li = 0;
    }
    if (li >= levels) {
        li = levels - 1;
    }

    int use_unicode = (opts && opts->eye_unicode && MB_CUR_MAX > 1);
    short cp = (snr_db < 12.0) ? C_POOR : (snr_db < 18.0) ? C_MOD : C_GOOD;
#ifdef PRETTY_COLORS
    attron(COLOR_PAIR(cp));
#endif
    if (use_unicode) {
        addstr(uni8[li]);
    } else {
        addch(ascii8[li]);
    }
#ifdef PRETTY_COLORS
    attroff(COLOR_PAIR(cp));
#endif
#ifdef PRETTY_COLORS
    /* Restore previously active color pair (e.g., green call banner) */
    if (saved_pair >= 0) {
        attron(COLOR_PAIR(saved_pair));
    }
#endif