šŸ“‹ What was this drill?

Drill: Numeric display: a dashboard with 6 metrics. Mono numerals, tabular-nums, right-aligned, consistent decimal places. Target: numbers scannable like a spreadsheet.

Interpretation: Broadcast loudness compliance desk (EBU R128) for a fictional regional public-radio station: 6-KPI strip over a 15-programme as-run ledger whose durations sum to exactly the 19:00ā€“00:00 window. Numeric stresses: all-negative signed columns in U+2212, tolerance-banded values with derived signed deviations, hh:mm:ss hero time objects, and paired aggregation candor ā€” the duration-weighted energy mean (āˆ’22.7 LUFS, teal) shown against the falsely-reassuring arithmetic mean (āˆ’23.1). Inter + JetBrains Mono tnum/zero, describe-first dataset with 13 on-load audit checks against python ground truth.

Fellside Radio Ā· Transmission Loudness Desk
AUDIT ā€”

Evening Session Loudness Ledger

Window 19:00:00 ā€“ 00:00:00 Ā· Mon 26 Oct 2026 Ā· sheet closed 00:01:04 Tue 27 OctĀ·Norm EBU R128Ā·Target āˆ’23.0 LUFSĀ·peak limit āˆ’1.0 dBTPĀ·tolerance Ā±1.0 LU
Session loudness
āˆ’22.7LUFS
target āˆ’23.0dev +0.3
Peak max
āˆ’0.6dBTP
limit āˆ’1.02 overs
Widest LRA
11.2LU
Night Signalsavg 7.2
In tolerance
12/ 15
2 marg (quiet)1 over
Aired
05:00:00
fill 100.0%15 items
Energy vs arith
āˆ’22.7energy
arith āˆ’23.1Ī” 0.4 LU
As-run ledger Ā· integrated loudness per programme
swipe sideways ā€” full ledger ā†’
StartDurProgramme Int LUFSDev LULRA LU Peak dBTPOversVerdict
Session 05:00:00 15 programmes āˆ’22.7 +0.3 7.2 āˆ’0.6 2 12/15 ok

Night ledger

Twelve of fifteen programmes aired inside Ā±1.0 LU of target. Exceptions: Night Signals 03 ran +2.2 LU hot for its full 62 minutes (21:19ā€“22:21, archive-vinyl transfer) and exceeded the āˆ’1.0 dBTP limit at 21:34:07 and 22:02:33; Weather & Shipping measured āˆ’1.1 LU and The Long Wave āˆ’1.4 LU, both quiet-side marginal. No listener complaints logged; no TX fault raised.

Aggregation note ā€” the arithmetic mean of the Int column reads āˆ’23.1, inside tolerance. Loudness aggregates by acoustic energy, so the session figure is āˆ’22.7 LUFS (+0.3 LU vs target, within session tolerance). Both figures are recorded so the 0.4 LU divergence stays auditable.

Action ā€” Night Signals archive transfer recalled for re-transfer (ref TX-2610-E); desk gain automation re-checked 22:24. Prepared R. Aldous, continuity. Reviewed ā€” duty engineer, closed 00:01. Deviation accepted.