Smash Happy Audio — User Manual
A precision metering suite — VU, full BS.1770-5 loudness, true peak, and a Real World Matrix that lets you audition how a master will actually translate.
Meridian is a pass-through analyser — it reads your audio, measures it in several genuinely useful ways at once, and passes it through unmodified except when you deliberately engage a Real World Matrix monitoring filter to audition a translation scenario. Nothing it does changes your master unless you ask it to.
Four instruments live side by side: a calibrated VU meter, a full BS.1770-5 loudness suite (momentary, short-term, integrated, and loudness range), a 4x-oversampled true-peak meter, and the Real World Matrix — three toggleable filters that simulate how your master will sound on a phone speaker, after streaming-platform loudness normalization, or over Bluetooth.
Four columns run left to right: VU Meter, LUFS Display, True Peak, and Real World Matrix, all reading the same signal simultaneously — you never have to switch modes to see level, loudness, and inter-sample peak behaviour together.
A calibrated segmented VU meter with proper 300 ms ballistics. Calibration is adjustable from −18 to −6 dBFS (default −18) so 0 VU can be set to match whatever reference level you're working against.
Full BS.1770-5 loudness metering with proper K-weighting: Momentary (400 ms), Short-Term (3 s sliding), and Integrated (the two-stage absolute + relative gated measurement the spec actually requires — not a simplified approximation), plus Loudness Range. Use START / STOP / RESET to control an integrated measurement pass over a full track.
Meridian's LUFS engine has been verified against an independent BS.1770 reference implementation on real mastered material, matching within hundredths of a LU — this is a genuine spec-accurate measurement, not a loudness estimate.
4x-oversampled true-peak metering catches inter-sample peaks that a plain sample-peak meter misses entirely — the kind of over that only appears after your master is converted to a lossy format. CLIP lights and latches when 0 dBTP is exceeded; RST clears it. An overall peak-hold tracks the loudest true-peak moment across the session.
Three toggles let you monitor through a simulated real-world listening scenario without leaving your DAW:
Toggle any of them on to hear — and see, via the dedicated meters each one drives — exactly how your master degrades in that scenario, then adjust with that translation in mind rather than guessing.
| Formats | Audio Unit, VST3 |
| Platform | macOS 12+, universal binary (Apple Silicon + Intel) |
| Loudness standard | ITU-R BS.1770-5, momentary/short-term/integrated + LRA |
| True peak | 4x oversampled, dBTP with latched clip + peak hold |
| VU calibration | −18 to −6 dBFS for 0 VU |
| Real World Matrix | Smartphone Trap, Streaming Penalty, Bluetooth Transcode |
| Processing | Pass-through — signal is unmodified unless a Matrix filter is engaged |
| Channel configuration | Mono or stereo in / stereo out |
Small differences (a few hundredths of a LU) between spec-accurate BS.1770 implementations are normal and usually trace to filter normalization or gating edge cases — Meridian's engine has been independently verified against a reference implementation. Larger gaps usually mean the other meter isn't fully spec-compliant, or you're comparing Integrated against Short-Term/Momentary by mistake.
That's expected and the whole point of true-peak metering — inter-sample peaks exceed 0 dBTP without any single sample in the file itself clipping. Back off your ceiling if this happens regularly.
Confirm you're actually listening through Meridian's output (not monitoring pre-fader) and that the toggle is actually engaged — each one is a real, audible filter change, not a meter-only simulation.