Skip to main content

Topic: Why can mp3 go above 1.0 peak when other formats cant? (Read 2741 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.
  • CoyoteSmith
  • [*][*][*]
Why can mp3 go above 1.0 peak when other formats cant?
i have noticed many of the mp3s i download from archive.org (Creative Commons) go above the 1.0 Track Peak according Replay Gain, however i have not been able to notice this in other formats. when rendering audio losslessly it clips, why/is mp3 an exception?

  • lvqcl
  • [*][*][*][*][*]
  • Developer
Why can mp3 go above 1.0 peak when other formats cant?
Reply #1
AAC, Vorbis, Musepack, AC3 can also have replaygain_track_peak > 1.0

  • saratoga
  • [*][*][*][*][*]
Why can mp3 go above 1.0 peak when other formats cant?
Reply #2
i have noticed many of the mp3s i download from archive.org (Creative Commons) go above the 1.0 Track Peak according Replay Gain, however i have not been able to notice this in other formats. when rendering audio losslessly it clips, why/is mp3 an exception?


All lossy formats do this if they're peak normalized.

  • 2Bdecided
  • [*][*][*][*][*]
  • Developer
Why can mp3 go above 1.0 peak when other formats cant?
Reply #3
All lossy formats do this if they're peak normalized.
i.e. if the original was peak normalised, then when you encode it lossy-ly, some of the peaks get pushed higher.

Most lossy formats are floating point internally, so they can cause values to be reconstructed that are far higher than the defined 0dB FS.

Cheers,
David.


  • AndyH-ha
  • [*][*][*][*][*]
Why can mp3 go above 1.0 peak when other formats cant?
Reply #4
Uncompressed wav can do the same thing -- much higher peak that 0dB coming out the DAC if the DAC has adequate headroom.