LAME MP3 has a -Y switch which tells the encoder to ignore/cut the 16KHz+ band if causes bloat or wasted bits. With V2 + -Y tracks that are 240kbps for no reason instead be 175 ~ 190kbps.
There have been "home grown" projects that tried to address these fundamental issues, one of the earliest being MP+ (then MPC, now Musepak). It was much more transparent at ~200kbps because it could actually allocate those bits to short transients and high frequencies. But people's HDDs were still small, the algorithms were still patented, and most people outside the enthusiast circles wanted more songs per megabyte, not less (anyone remember mp3pro?)
You forgot Musepack --Standard(170kbps) which despite being a pure subband codec holds fine to AAC/Vorbis/Opus. I have no idea how FHG thought MP3 would be transparent at 128 ~ 175kbps with it issues, Then they added salt to the wound by capping decoders to 320kbps despite MP3 max bitrate was 640kbps?.
That what bugged me about VBR in MP3 what the point if It locks to 320kbps, If a sections or a whole sample could be transparent at 350 ~ 500kbps like in other lossy codecs.
Edit: Some are forgetting that AAC/Vorbis/Opus can reach 512kbps even if at 192kbps VBR, If it is something very complex. While Musepack can go much higher like 1.4mbit if needed, I've had few albums with sections to whole song reach 540 ~ 790kbps even with the --Standard setting.
Last edited: