Lots of stuff here, but there is a lot of overlap and it gets finicky in a hurry. Audio is difficult.
Audio input stuff:
- Mic auto level. Horribly difficult to do across all hardware.
- Noise removal
- Everything that is not the users voice should be removed
- Probably requires some neural network shenanigans
- Should be a toggle, in case someone wants to play the guitar
- All audio that comes out of the users speakers should be filtered from the mic input
- For those who don’t want to put on their headset
- or listens to music
- or gets annoying notification noises all the time
- should be a toggle for those who wants to share their hard bass
- Hardmode extra credits: broadcast-like audio processing
- EQ - either automatic or with voice friendly presets. Focus on clarity
- Compressor/limiter to keep voice level consistent over time
Audio out stuff:
- Mechanism that catches unwanted/unexpected loud noises
- Bit broader control over other users audio levels - current slider seems too mild for some
- EQ or perhaps a bit more advanced audio out controls, to facilitate “placement” in the audio, to make voice chat stand out or sit better in the overall audio image.
- imagine a “night mode” that turns your overly excited friends victory squeals to be a bit more tolerable
- Mostly to fix those with mic/voice combos that seem to drown in gunfire and explosions in game
Some of the above is a bit on the extreme side, and is mostly meant as systems that can do wonders for the overall experience by doing small gentle nudges, and should be mostly automatic, if an overall “best practice” can be achieved across the board.
Think of it implemented like an “advanced input” and “advanced output” option, that tries to make audio pleasant for the majority of users.