• Thorry84@feddit.nl
    link
    fedilink
    English
    arrow-up
    17
    arrow-down
    2
    ·
    7 months ago

    Even a human with very good hearing and knowledge of how a song is supposed to sound cannot tell the difference between CD quality audio and 256k AAC like iTunes uses.

    Don’t believe all the nonsense audiophiles keep spewing out. Human ears suck. If we hadn’t had our giant brains to compensate, we’d be practically deaf.

    • aleph@lemm.ee
      link
      fedilink
      English
      arrow-up
      11
      arrow-down
      2
      ·
      7 months ago

      This. People assume that because it’s “compressed” it must sound flatter, less dynamic, or just vaguely worse than uncompressed audio, despite the fact that audio compression specifically uses psychoacoustic models to remove the bits of data that our human ears and brains cannot hear to begin with.

      Expectation bias is a helluva drug.

      • kingthrillgore@lemmy.ml
        link
        fedilink
        English
        arrow-up
        4
        arrow-down
        1
        ·
        7 months ago

        Even FLAC is compressed. Which is how I procure my music because I have the storage space.

        • bamboo@lemm.ee
          link
          fedilink
          English
          arrow-up
          5
          ·
          7 months ago

          FLAC is compressed, but unlike lossy codecs like AAC and MP3, FLAC is fully lossless. Lossy codecs delete information the authors believe you won’t notice, lossless compression keeps all the data and just tries to fit it in a smaller space. The original recording can be perfectly reproduced (taking into account sample rate and depth).

        • aleph@lemm.ee
          link
          fedilink
          English
          arrow-up
          5
          ·
          7 months ago

          Yup, although that doesn’t stop some weirdos out there claiming that CDs sound better than FLAC.

      • barsoap@lemm.ee
        link
        fedilink
        English
        arrow-up
        2
        ·
        edit-2
        7 months ago

        psychoacoustic models

        Sometimes they mess up. Actually only ever noticed it once and that was years ago CD vs. ogg vorbis at full quality level, this track. Youtube version is even worse, it seems (from memory): The guitars kicking in around 30 seconds should be harsh and noisy as fuck like nothing you’ve ever heard, they’re merely distorted on youtube.

        Then lossy codecs are a bad idea for archival reasons as you can’t recode them without incurring additive losses – each codec has a different psychoacoustic model, each deletes different stuff. Thus, FLAC definitely has a place.

        • aleph@lemm.ee
          link
          fedilink
          English
          arrow-up
          1
          ·
          edit-2
          7 months ago

          Killer samples do happen, sure but vorbis at Q9? I’m highly dubious. That track in particular just sounds badly recorded to begin with. If you have that same version in FLAC i would be interested to see some ABX test results or test it myself.

          For archival purposes, though, I agree FLAC is the way to go.

          • barsoap@lemm.ee
            link
            fedilink
            English
            arrow-up
            1
            ·
            7 months ago

            Killer samples do happen, sure but vorbis at Q9? I’m highly dubious.

            Back in 2004, when the album released, the encoder was barely past version 1.0. Though after 20 years I could misremember “full quality” as “whatever people said wouldn’t degrade quality”.

            That track in particular just sounds badly recorded to begin with.

            Heresy. Next thing you’re going to tell me is that Sunn O))) should move the mics away from the amps so the sound is cleaner. Granted, though, Sunn O))) does that live, blackmail live is quite different because they can’t layer a gazillion tracks for the mix. But yes the deliberateness of just how much noise is in those guitars doesn’t get conveyed after getting mangled by ten year old youtube compression.

            • aleph@lemm.ee
              link
              fedilink
              English
              arrow-up
              2
              ·
              7 months ago

              Lol, I’m not saying that brickwalling the mix to achieve a certain effect isn’t a thing, but at the extreme levels of compression and clipping apparent on that track, it’s unlikely that a FLAC would sound even remotely different. Apparently the band agreed - in 2020 they issued a remaster which seems noticeably less crushed:

              Dynamic range comparison screenshot

              Incidentally, I saw Sunn O))) live once. I can still feel my bowels shake.

    • tal@lemmy.today
      link
      fedilink
      English
      arrow-up
      5
      ·
      7 months ago

      I would guess that the fact that people aren’t all using some kind of standard-response reference headphones is probably going to have a considerably-larger impact on the human-perceivable fidelity of the audio reproduction than any other factor.

    • prole@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      2
      ·
      7 months ago

      This is true. That said, I’ve seen people claim that nobody can tell the difference between lossless and 128kbps mp3, but that’s complete bullshit.

      Though once you get above 192, it’s pretty indistinguishable.

      • Thorry84@feddit.nl
        link
        fedilink
        English
        arrow-up
        2
        ·
        7 months ago

        Would really depend on the version of MP3. The first versions had some major issues with artifacts being introduced. People probably listened to that and concluded all compressed music must be shit. Later versions were much better, even though I would think 128k is probably too low and would be noticeable with some effort. I agree, starting at 192k and people can’t tell anymore.

        Does anybody use MP3 anymore? I don’t really know to be honest.