There's a solution here and I don't know why nobody's talking about implementing it.<p>First, some context: the problem of dark scenes is the same problem as unintelligible dialog is the same problem as a lot of classical music being quieter than pop music.<p>All of these creators are striving for <i>greater range</i>. Dark scenes work fantastically if you watch them on a bright screen in a pitch-black room. Dialog is perfectly intelligible in surround sound at full volume, and marvelously expressive. And classical music is delicately soft and then powerfully loud in an otherwise perfectly quiet room, to incredible effect.<p>But as soon as you're watching on a screen in a normally lit room, or listening on your TV speakers while people are talking in the kitchen, or trying to listen to classical in your car with the sounds of traffic... it <i>all falls apart</i>.<p>The solution is what audio engineers have known about for decades, which is called <i>compression</i>. Compression is: make the quiet parts of the classical music almost as loud as the loud parts. Boost the dialog channel so the mumbling parts are almost as loud as regular conversation. And boost the brightness of dark scenes. Compress the range -- go back to <i>less range</i>.<p>But we don't want to mess with the source material when viewing/listening in ideal conditions, because we want to keep that awesomeness. So we need <i>dynamic compression</i>. Which isn't really that hard.<p>A television can have an ambient light sensor just like your MacBook does, to boost dark scenes when you watch during the day, but not at light with the lights off.<p>A television can have a cheap microphone to detect ambient noise, and also be aware of its own volume setting, and so boost dialog at low volume settings and when it's noisy around, using the surround sound signal. (And a MacBook or iPad or iPhone can do this too.)<p>And your car radio or phone music player can apply dynamic audio compression to your classical music as well, similarly using a microphone to detect the need to compensate for the rumble of traffic or conversation in the coffee shop.<p>What baffles me is that all of this is possible <i>right now</i>, and it's actually <i>trivial</i> to implement, relatively speaking. (There's a bunch of tuning involved to make it work well and feel perceptually natural, but it's not like we need to invent new types of signal processing or anything.)<p>But nobody's even <i>talking</i> about dynamic compression as the solution. And I just don't get it. Why not?