"I was in a conference call last Friday where I’d adjusted the volume up to compensate for the client’s quiet microphone, only to be audibly shot in the ears by

"I was in a conference call last Friday where I’d adjusted the volume up to compensate for the client’s quiet microphone, only to be audibly shot in the ears by an auto-playing video on a website. There is a lot of problematic stuff to unpack there, much of which is not the fault of the audio hardware or OS." If you're suddenly reaching for the mute button because you're being blasted then there has already been a significant failure. It just happens to be something everybody is failing at and nobody can be bothered to fix. We suck at quantifying (simply) the loudness of audio content. If you set volume according to average SPL, high transients (e.g. Cannon fire in a too-faithful recording of Tchaikovsky's 1812 overture) can still be shattering. If you set it according to the maximum SPL, an action movie with lots of explosions might have a lot of dialogue that's inaudible. Audio that has been highly compressed to the point of distortion may seem uncomfortably loud even at moderate SPL's. The thing is, we probably shouldn't even try to use simple rules for setting volume. We need to build systems that observe when people adjust the volume and learn from that. No OS should be dumb enough to suddenly let rip with a new audio stream whose first second is, on average, 30 dB higher than what was previously playing. Yet, they practically all do. A hardware volume knob is an important thing for any audio device to have but, if the device has the capacity to run programs and it's users are still reaching for that knob in startlement, there has been a failure.