* I'll anser your second question first because it's the easiest. Look into the audio descriptions from Amazon Prime where they use their new system with automatic ducking and text-to-speech instead of real live humans to read (High Life, Under the Silver Lake, Sleeping Dogs Lie, Dazed and Confused, House of Demons, Coppola's adaptation of the Outsiders, Vivarium, Skin, It's a Wonderful Life). The uncredited description scripts are surprisingly competent;they don't give you the comprehensive detail or the graceful prose of a DVS/Deluxe/IDC script, but they get the job done (if the exact same words were being read out by Hannah Brownlie or Georgina Rose instead of RoboNarrator [F US] I would have few complaints). And you can get used to the TTS, after a while it just fades into the background. But the mixing can be fucking ghastly (depending on the volume of the movie itself), they almost mute the audio every time the describer "speaks" so that you strain to hear the actual film sounds. Since it is automatic and indescriminate (amazon has insured that these descriptions can be churned out as fast n' cheap as possible by removing virtually all human involvement), it doesn't matter if you're watching a quiet musicless conversation at a restaurant table, or a Christopher Nolan shootout on a hill with a $1 it's all mercilessly cranked down to the same low volume.

* Though the ducking in your linked episode is undoubtedly too much for this narrator, it's not what I would call "extreme", yes the sound is lowered, but it is still clearly audible. A year ago I saw an eightyeight-minute film with the same describer and didn't notice any audio problems (he also voices Green Room). Keep in mind, however, that I watch all movies with headphones so everything is crystal clear no matter what, speaker/television/home theatre users may have a different experience.

The audio quality is perfect. Continue