rspektrum/README.md

rspektrum

An interactive spectrogram viewer for inspecting mLink radio captures (and any other audio). It computes an STFT of a WAV file, draws it as a zoomable, pannable spectrogram, and can overlay mLnL annotations — labelled regions (TX frames, assertion outcomes, impairment fires, …) carried inside the WAV file itself. You can box a time/frequency region, hear it back through a bandpass filter, and export either the picture (PNG) or the isolated audio (WAV). It runs as a native desktop app (C + raylib), as a headless CLI renderer, and as a WebAssembly build in the browser.

What it's for

The primary use case is reviewing captures from the mLink stack: a WAV recording of an over-the-air signal, with an embedded mLnL chunk describing what the modem/daemon believed it was transmitting at each instant. rspektrum renders those annotations on top of the measured spectrogram so you can see, frame by frame, what was intended versus what actually hit the air. See mlnl_chunk_spec.md for the annotation format. It also works as a general-purpose spectrogram tool for plain WAVs with no annotations.

Features

• STFT spectrogram with selectable colormaps, adjustable dB floor / dynamic range, and absolute (dBFS) or relative amplitude scaling.
• mLnL annotation overlay — labelled boxes drawn from the WAV's embedded annotation chunk; hover a box (or its region on the waveform scope) for a tooltip with per-frame detail (sequence, channel, rate, scheduling offset…).
• Zoom & pan the time/frequency view (mouse wheel + Alt/middle-drag).
• Region selection: box a time and frequency range with the mouse.
• Filtered playback: play just the selected region, band-limited to the selected frequency box (FFT bandpass — "what you hear is what you'd export").
• Waveform scope (toggle) showing the time-domain signal under the spectrum.
• Marker / ruler tool and a spectrum slice (PSD) readout.
• Export: save the view as a PNG, or the selected region as a WAV.
• Headless render mode: produce an annotated PNG from the command line with no window (see below).
• Broad input support: WAV directly (8/16-bit PCM, 32-bit float; stereo downmixed to mono); other formats transparently transcoded via ffmpeg if it is on PATH. Drag-and-drop loading.
• Cross-platform: Linux/desktop, Windows, and a WebAssembly build.

Building

The build is driven by a checked-in Makefile (premake5 is not required to build — only to regenerate the makefiles):

make -f rspektrum.make config=debug_x64      # -> bin/Debug/rspektrum
make -f rspektrum.make config=release_x64    # -> bin/Release/rspektrum

The web build:

./build_web.sh      # emscripten; emits the WebAssembly bundle

Note: the release build enables -O2, which turns on extra warnings (-Wformat-truncation) that the debug build doesn't — build release before declaring a change clean.

Running (GUI)

./bin/Debug/rspektrum [input.wav]

Load a file by passing it on the command line, dragging a .wav onto the window, or pressing O for the file browser.

Controls

Input	Action
O	Open file browser
Mouse wheel	Zoom time/frequency
Alt+drag / middle-drag	Pan the view
LMB drag	Select a time + frequency region
Space	Play / stop the selected region
Hover an annotation	Tooltip with that frame's mLnL detail
P	Show / hide the waveform scope
M	Marker / ruler tool
S	Spectrum slice (PSD)
E	Export PNG
W	Export selection as WAV
Home	Reset view (fit all)
End	Zoom to start
F11	Toggle fullscreen
F1	About / help
Esc	Clear selection / close dialog

Most controls are also available as buttons in the left sidebar (colormap, floor, dynamic range, annotation opacity, grid, etc.).

Headless rendering (CLI)

rspektrum can render an annotated spectrogram to a PNG without opening a window — useful for batch capture and for agents reviewing test output:

./bin/Debug/rspektrum --render OUT.png INPUT.wav [options]

Options:

Flag	Effect
-r, --render OUT.png	Render to OUT.png and exit (no window)
-a, --annotations	Force the annotation overlay on (solid, for review)
--no-annotations	Force the overlay off
--annotation-opacity=V	Resting overlay alpha 0..1 (default 0.06, faint)
--pane	Capture only the spectrogram pane (no sidebar/scope)
--width N / --height N	Output size (default 1280×800)
-h, --help	Usage

Default annotation opacity is intentionally faint (the overlay is meant to stay out of the way until hovered), so for a screenshot you actually want to review, pass --annotations. Example against the bundled sample:

./bin/Debug/rspektrum --render /tmp/review.png mlnl_samples.wav --annotations

mlnl_samples.wav is an in-repo WAV that carries an embedded mLnL chunk.

The hover tooltip (sched offset, per-frame detail) only appears with a live mouse over a box, so it cannot show up in a static --render. To verify tooltip behaviour you need a real (or virtual) display driving the GUI — see below.

Driving the GUI headlessly (for agents / CI)

The app can be run, screenshotted, and clicked on a virtual X display with no monitor or GPU (Mesa software GL under Xvfb). The accumulated playbook lives in AGENTS.md; the working reference implementation is shot_input.sh.

The loop in one breath:

Xvfb :99 -screen 0 1280x800x24 >/tmp/xvfb.log 2>&1 &     # 1. fake screen
DISPLAY=:99 ./bin/Debug/rspektrum mlnl_samples.wav \
        >/tmp/app.log 2>&1 &                              # 2. run on it
sleep 2                                                    # 3. reach a steady frame
DISPLAY=:99 import -window root /tmp/shot.png             # 4. grab the frame

Prerequisites (Debian/Ubuntu): sudo apt-get install xvfb imagemagick xdotool (plus libgl1-mesa-dri and LIBGL_ALWAYS_SOFTWARE=1 if GL fails / frames are black). To exercise UI paths, synthesize input with xdotool against DISPLAY=:99 — e.g. move the mouse over an annotation box and re-grab to capture the hover tooltip:

DISPLAY=:99 xdotool mousemove 640 400      # hover a box (coords from the spectrogram)
DISPLAY=:99 import -window root /tmp/hover.png
DISPLAY=:99 xdotool key space              # play the selection, etc.

There is no window manager, so the window sits at 0,0 and fills the Xvfb screen — match the screen size to the window and a root grab equals the app's frame. Always capture stdout/stderr to a log; it's your only view of TraceLog output and crashes. shot_input.sh wraps all of this (start → settle → screenshot → optional input → diff); read AGENTS.md for the traps (input focus, ImageMagick v6 vs v7 command names, software-GL timing).

Technical notes

• STFT: Hann-windowed, 2048-point FFT with 50% overlap by default; frequency resolution sampleRate / fftSize Hz per bin. Amplitude shown in dB.
• Axes: X = time (s), Y = frequency (Hz, scaled to the file's Nyquist), colour = amplitude.
• Playback / WAV export share the same processing path: the selected time span, FFT-bandpassed to the selected frequency box, peak-normalised.
• mLnL parsing: walks the WAV's RIFF chunks for the four-CC mLnL chunk (UTF-8 JSON Lines); unknown chunks are skipped, so annotated files remain standards-compliant audio everywhere else.

Source layout

src/
  spectrogram.c        # entry point, main loop, CLI args, headless render
  stft.c / fft.c       # STFT + FFT
  render.c             # spectrogram, annotations, tooltips, scope
  ui.c                 # sidebar, file browser, buttons
  audio.c              # WAV load (ffmpeg fallback), bandpass, playback, WAV export
  mlnl.c / mlnl.h      # mLnL annotation chunk parser
  platform_*.c         # per-OS shims (linux / win32 / web)

See raylib_for_desktop_applications.md for the performance/idle-CPU lessons behind the desktop build.