Audio rescue for missing voices

From version 1.0.2. If the voices in a film are missing because the provider’s file was mixed wrongly, Neverlag brings them back by itself.

What it is

Some films and series are mixed wrongly at the provider: the speech is not where it belongs (in the centre channel) but in the bass channel (LFE, the channel for the subwoofer). When surround sound (5.1) is converted for the two speakers of a Mac, iPhone, iPad or Apple TV, every common player leaves the bass channel out – which is why the picture plays but you hear no voices. (The same happens in Apple’s Preview.)

Neverlag recognises exactly this case automatically and mixes the bass channel in for that one file, so the dialogue becomes audible. For all correctly mixed films, series and live channels, nothing changes.

How to use it

You do not need to do anything – audio rescue is always on and needs no setting. When you start a film or an episode, Neverlag listens briefly during the first few seconds: if the speech channel is empty and there is continuous sound in the bass channel instead, the app switches the bass channel into the mix. All you notice is that the sound is there after a few seconds.

In that case the journal (Settings → Advanced) contains a line “audio-rescue: Sprache im LFE, LFE in Stereo gemischt” (speech in LFE, LFE mixed into stereo) – the proof that the rescue has kicked in.

Limits

  • The rescue only applies to films and series, not to live channels. (Live usually plays correctly, and there is no clear start of a file to measure from.)
  • It only kicks in for surround tracks (5.1/7.1) that are mixed down to stereo. Multi-channel systems and direct audio passthrough are not affected.
  • The decision is made per file, once at the start. If you switch the audio track, Neverlag measures again.
  • If a file is so damaged that there is no usable sound at all, the rescue cannot add anything either.

For the technically curious

While loading, Neverlag continuously measures the volume of each channel (RMS level via the astats audio analysis filter). The rule: if the centre channel stays below −50 dB for at least 10 seconds within the first minute and the bass channel is above −45 dB during the same time, the app sets the mix level of the bass channel to normal level for that file (−3 dB to left and right, as for the centre). Otherwise the standard mix without the bass channel stays – exactly the same output as before, sound for sound.

Further reading