3 ms·

It might be a bit of a hot take but from my point of view audio is mostly a solved problem. It's not like video where we still have trade-offs in terms of quali

by simias 6y ago

It might be a bit of a hot take but from my point of view audio is mostly a solved problem. It's not like video where we still have trade-offs in terms of quality due to the huge bandwidth and processing involved for very high resolutions as well as the difficulty of reproducing very high dynamic ranges and framerates. None of this is relevant for audio. Despite what some audiocultists would have you believe CD quality is basically as good as it gets, you cover effectively 100% of the (normal) human hearing range, both in frequency range and dynamics. Ok, ok, technically not 100% of the dynamics, but few people care to listen to "distant ventilator on a quiet night" followed by "Airbus A380 taking off from the point of view of the jet turbine". Of course the quality of reproduction matters a lot, playing a CD through cheap $5 earphones is not going to sound great. But even there the cost of decent quality electronics keeps dropping and dropping. You can get a good quality DAC for a trivial amount of money (and stand alone DACs are still usually overpriced because they're for the niche/enthusiast/pro market, everybody else uses whatever their smartphones/laptop/desktop has integrated). Buying ultra high-end, ultra-expensive "audiophile" equipment these days is effectively like buying mechanical watches. It's because you like the luxury and craftsmanship, or you're making a fashion statement. Apple knows that, hence these AirPods. If you're about function over form you'd sooner get a cheapo Casio quartz watch. And there's nothing wrong with that, it just annoys me when audiophiles try to justify their fashion statements with pseudo-science.

TimTheTinker 6y ago
“spatial audio” is the killer feature here. It’s one of the last “unsolved” audio problems (at least in available headphone/earphone product lines), along with the upper frequency limit on noise canceling, which this product doesn’t solve.
- shortandsweet 6y ago
  Killer feature huh? Why have I never heard of it before this post.
  - TimTheTinker 6y ago
    Spatial audio (and by extension, binaural audio) has been an ongoing research project in a lot of different places over the last 10 or so years. The problems involved have been mostly theoretically solved, but Apple is the first to bring to market a consumer product that takes various Dolby multi-channel surround sound input types and outputs a true binaural stereo audio stream that varies with head (and screen) movement in 3d space. Here's one place to start if you're interested in the ongoing theory/research: https://www.bbc.co.uk/rd/projects/binaural-broadcasting https://www.bbc.co.uk/rd/projects/binaural-broadcasting
theincredulousk 6y ago
ehhh I'd say mostly solved. The problem of turning a recording into a perfect sound-wave replication is actually not easy, because the physical world is annoying with heat and friction and and physics. I'd say 24/96 is comfortably the "end game solved problem" quality for source recordings though, not CD. Even if it is just for the baseline safety margin that it is easier to implement perfect filters with no transition bands anywhere near the audible frequencies, and that additional bit precision has no downside unless you get into accuracy problems with noise on the recording side. From the speaker point of view, I think IEMs are the closest to being "solved", since the tiny drivers minimize the physics issues with larger drivers, they are probably closest to pure reproduction. For full-size speakers etc., pure reproduction can be accomplished for certain frequency ranges by individual drivers, but over the full range it is not so "solved", e.g. with bass there is often a physics issue with the kinetics of moving a large driver cone in perfect phase-sync with the relatively tiny and light tweeter. I've actually been wondering if speaker cones are a possible application for graphene.
ksk 6y ago
Audio (gear) is nowhere near to being a solved problem. One open problem is to match your own personal HRTF to your gear. Audio (production) is also not a solved problem. Your experience is decided by the recording/mixing engineer, and there may be multiple renderings possible of the same content. In much the same way, different conductors of the same orchestra can produce different flavors of the same composition. Then there is the spatial localization for games and movies, which is still not quite solved. >And there's nothing wrong with that, it just annoys me when audiophiles try to justify their fashion statements with pseudo-science. Yes, those people can get nutty, ignore them! Audio is much bigger than Audiophile :) Having said that, with no pressure from consumers, why would anyone invest money in new headphone designs? I recently purchased a pair of planar magnetic headphones which used to be crummy when I tried them a few years back. The newer design is very efficient, the manufacturing process is also pretty solid, and the sound is fantastic.

by simias 6y ago

TimTheTinker 6y ago

“spatial audio” is the killer feature here. It’s one of the last “unsolved” audio problems (at least in available headphone/earphone product lines), along with the upper frequency limit on noise canceling, which this product doesn’t solve.

shortandsweet 6y ago
Killer feature huh? Why have I never heard of it before this post.
- TimTheTinker 6y ago
  Spatial audio (and by extension, binaural audio) has been an ongoing research project in a lot of different places over the last 10 or so years. The problems involved have been mostly theoretically solved, but Apple is the first to bring to market a consumer product that takes various Dolby multi-channel surround sound input types and outputs a true binaural stereo audio stream that varies with head (and screen) movement in 3d space. Here's one place to start if you're interested in the ongoing theory/research: https://www.bbc.co.uk/rd/projects/binaural-broadcasting https://www.bbc.co.uk/rd/projects/binaural-broadcasting

theincredulousk 6y ago

ehhh I'd say mostly solved. The problem of turning a recording into a perfect sound-wave replication is actually not easy, because the physical world is annoying with heat and friction and and physics. I'd say 24/96 is comfortably the "end game solved problem" quality for source recordings though, not CD. Even if it is just for the baseline safety margin that it is easier to implement perfect filters with no transition bands anywhere near the audible frequencies, and that additional bit precision has no downside unless you get into accuracy problems with noise on the recording side. From the speaker point of view, I think IEMs are the closest to being "solved", since the tiny drivers minimize the physics issues with larger drivers, they are probably closest to pure reproduction. For full-size speakers etc., pure reproduction can be accomplished for certain frequency ranges by individual drivers, but over the full range it is not so "solved", e.g. with bass there is often a physics issue with the kinetics of moving a large driver cone in perfect phase-sync with the relatively tiny and light tweeter. I've actually been wondering if speaker cones are a possible application for graphene.

ksk 6y ago

Audio (gear) is nowhere near to being a solved problem. One open problem is to match your own personal HRTF to your gear. Audio (production) is also not a solved problem. Your experience is decided by the recording/mixing engineer, and there may be multiple renderings possible of the same content. In much the same way, different conductors of the same orchestra can produce different flavors of the same composition. Then there is the spatial localization for games and movies, which is still not quite solved. >And there's nothing wrong with that, it just annoys me when audiophiles try to justify their fashion statements with pseudo-science. Yes, those people can get nutty, ignore them! Audio is much bigger than Audiophile :) Having said that, with no pressure from consumers, why would anyone invest money in new headphone designs? I recently purchased a pair of planar magnetic headphones which used to be crummy when I tried them a few years back. The newer design is very efficient, the manufacturing process is also pretty solid, and the sound is fantastic.