The measurement. R 128 measures with ITU-R BS.1770: K-weighting, 400 ms blocks with 75 % overlap, an absolute gate at -70 LUFS and a relative gate at -10 LU. WavSpec implements BS.1770-5 natively, with short-term and momentary loudness per EBU Tech 3341 and loudness range per Tech 3342, and its test suite checks the meter against the EBU Tech 3341 and 3342 test signals.
True peak, not sample peak. A sample-peak meter misses overs that appear between samples once audio is converted or encoded. WavSpec measures true peak with the 4× oversampled filter from BS.1770-5 Annex 2, and its limiter uses the same 4× detection with a 2 ms look-ahead, so the ceiling holds on the file that is delivered.
Target and tolerance. The EBU R 128 broadcast preset aims at -23.0 LUFS with ±1.0 LU and a ceiling of -1 dBTP. WavSpec applies gain to land on the target itself, not just inside the window, and limits only when the predicted peak would cross the ceiling. A file already within the spec passes unchanged.
Format rules. The preset accepts 48 kHz, 16- or 24-bit PCM, in mono, stereo or 5.1. A file at another rate is resampled; a float or 32-bit file is converted with TPDF dither; 5.1 files are measured and corrected with the BS.1770 channel weights. AIFF, FLAC, MP3 and AAC sources are decoded and delivered as PCM WAV.
Loudness metadata. R 128 says loudness metadata should state the actual programme loudness. After verification, WavSpec writes the measured values into the bext chunk of delivered Broadcast Wave files using EBU Tech 3285 version 2 fields, refreshing any fields the source carried, and keeps iXML, LIST and cue chunks in their original order.
The rest of the R 128 family. The R 128 s1 short-form preset adds a -18 LUFS short-term maximum. The UK DPP AS-11 preset targets -23 LUFS ±0.5 LU at 48 kHz, 24-bit, with a -3 dBTP ceiling and 18 LU of loudness range as guidance. The ARD/ZDF/ORF TV (2025) preset uses ±0.5 LU and a 20 LU guideline. Duplicate any of them to match a station’s own sheet.
Proof that travels. Every delivered file is re-read from disk and measured again. The report gives before and after values per criterion, a short-term loudness timeline, the spec’s reference document, and a SHA-256 digest for each delivered file, as PDF, HTML, CSV and JSON.