3 ms·

Interesting, whenever I failed it was between 320kbps and uncompressed wav. I know that I did a blind test about 10 years ago and could distinguish between ~192

Interesting, whenever I failed it was between 320kbps and uncompressed wav. I know that I did a blind test about 10 years ago and could distinguish between ~192kbps variable bitrate and lossless audio with my left ear, mostly influenced by percussion (left ear for some reason can hear 21kHz frequencies).

dehrmann 5y ago
Uncompressed CD audio is 1.4 Mbps. FLAC can do roughly 2:1 compression, so that's 700 kbps. It's not crazy that modern mp3s (encoders have gotten better over the years) sound good at half the size of something lossless. They're only removing subtle details at that point. You can see the same thing with high-quality JPEGs vs PNGs for photos. There's a decent size reduction, but to see the difference, I have to open the files in Photoshop, zoom in, and toggle which one is visible. If I compute the difference in Photoshop, it's usually only a few levels different, and never more than ~8. That's really hard to see. Sadly, they're not even using a modern codec like Vorbis or Opus. That said, while I doubt I can tell the difference between FLAC and high-quality Vorbis, once I learned to listen for distortions on the high end, it's a little hard to unhear. Same as looking for blocking artifacts in JPEGs or video.
- teawrecks 5y ago
  FLAC is lossless, and given a random piece of audio it would be impossible to always encode into 700kbps something that is originally 1.4Mbps without loss of information. So your 2:1 samples claim doesn't sound accurate. Maybe file size, but not sample count.
  - colanderman 5y ago
    Yes, if you are encoding level-maximized white noise, FLAC cannot compress at that ratio, or likely at all. Music is not white noise. It does not maximize the channel capacity of the audio spectrum. As a rule of thumb, FLAC compresses most anything a human might call "music" in about a 2:1 file size ratio.
    - teawrecks 5y ago
      Right, so file size, not sample count. It's lossless.
bydo 5y ago
It's the sibilance that always gives it away for me. Even the 128kbps mp3 sounded fine for most of these, but when it fails, it really fails.

dehrmann 5y ago

Uncompressed CD audio is 1.4 Mbps. FLAC can do roughly 2:1 compression, so that's 700 kbps. It's not crazy that modern mp3s (encoders have gotten better over the years) sound good at half the size of something lossless. They're only removing subtle details at that point. You can see the same thing with high-quality JPEGs vs PNGs for photos. There's a decent size reduction, but to see the difference, I have to open the files in Photoshop, zoom in, and toggle which one is visible. If I compute the difference in Photoshop, it's usually only a few levels different, and never more than ~8. That's really hard to see. Sadly, they're not even using a modern codec like Vorbis or Opus. That said, while I doubt I can tell the difference between FLAC and high-quality Vorbis, once I learned to listen for distortions on the high end, it's a little hard to unhear. Same as looking for blocking artifacts in JPEGs or video.

teawrecks 5y ago
FLAC is lossless, and given a random piece of audio it would be impossible to always encode into 700kbps something that is originally 1.4Mbps without loss of information. So your 2:1 samples claim doesn't sound accurate. Maybe file size, but not sample count.
- colanderman 5y ago
  Yes, if you are encoding level-maximized white noise, FLAC cannot compress at that ratio, or likely at all. Music is not white noise. It does not maximize the channel capacity of the audio spectrum. As a rule of thumb, FLAC compresses most anything a human might call "music" in about a 2:1 file size ratio.
  - teawrecks 5y ago
    Right, so file size, not sample count. It's lossless.

bydo 5y ago

It's the sibilance that always gives it away for me. Even the 128kbps mp3 sounded fine for most of these, but when it fails, it really fails.