"du musstest jede der formate gleich aufnehmen... dann hast du gleiche voraussetzungen" - all three: gain forced to 1.0 (raw), played live 20s unrecorded to settle past any quiet intro, then 20s captured via MediaRecorder. Same duration, same method, no mixing offline-rendering (mod/rjp) with live-capture (midi) like the first attempt did.
| Format | Raw RMS | Peak | Play (raw, gain=1) |
|---|---|---|---|
| MOD (sound/Sensible_Soccer_Euro/mod.menu5) | 0.1194 | 0.5207 | |
| RJP (Sensible_World_Soccer_95/MENU) | 0.0487 | 0.2231 | |
| MIDI (sound/9495.mid, 20s-40s window) | 0.0528 | 0.3254 |
Calibrated gains derived from these (mod=0.321 fixed reference): rjp≈0.787, midi≈0.726. Play a "native" track below, then the matching WAV above, back to back - they should now sound the same loudness (the WAVs are raw/unscaled, so turn your volume up to compare against the calibrated native players fairly, or just compare native-vs-native by ear).
Green < -12dB, yellow -12..-6dB, red > -6dB (clipping risk, peak-based). White tick = peak-hold (slowly decays). Blue overlay bar = short-term RMS (perceived loudness - peak alone is misleading, e.g. mod has a lower peak but sounds louder because of its higher RMS). The "Bewertung" text rates THAT RMS value: <-26dB zu leise, -26..-20 etwas leise, -20..-14 optimal, -14..-8 etwas laut, >-8 zu laut. All four meters are always live - only the currently-playing track's bar will move, so start one, watch it settle, stop, start the next, and compare directly against each other.
Startet je einen Track pro Format gleichzeitig (jeweils der erste in der Liste unten) und schlägt, basierend auf den live gemessenen RMS-Werten, pro Format eine Lautstärke vor, die alle vier auf denselben Pegel bringt - "übernehmen" schreibt den Vorschlag direkt in den Regler oben.