Public bug reported: 1) This applies to Ubuntu 24.04 (and even 26.04, for that matter). 2) This affects the included 3.100 lame for both Debian and Ubuntu. 3) I expect this package to update to the patched version, as is the case for programs such as EasyTag. 4) Instead, this package languishes at version 3.100.
The current LAME is out of date. Most importantly, it is missing a fix to the buggy (and slower for no gain, only a sound penalty) q setting for CBR and ABR modes. The LAME maintainers have disappeared for the past few years and are unlikely to release an official 3.101. As a result, this package should be 3.101. Beta 3. Much like EasyTag, the situation will not get better. We have to go with a source build with patches. No official release will happen. The 3.101 Beta 3 has the new features of 3.101 and most importantly a fix for the CBR and ABR modes of encoding. The default quality level of 3 and the higher values of 0, 1, and 2, cause severe distortion on stereo files. These modes employ a buggy noise-shaping amplification filter that causes issues. Already, LAMEXP and dBPowerAmp have adopted LAME 3.101 Beta 3 for this reason. Multiple threads on HydrogenAudio and on Audacity attest to the problem: - HydrogenAudio: <https://hydrogenaudio.org/index.php/topic,126120.0.html> - HydrogenAudio: <https://hydrogenaudio.org/index.php/topic,128311.0.html> - Audacity: <https://forum.audacityteam.org/t/critical-mp3-encoder-quality-bug-needs-to-be-fixed/149016> Thank you for your time. ** Affects: lame (Ubuntu) Importance: Undecided Status: New ** Tags: encode lame mp3 patch quality -- You received this bug notification because you are a member of Ubuntu Desktop Bugs, which is subscribed to lame in Ubuntu. https://bugs.launchpad.net/bugs/2139642 Title: LAME Out-of-Date, Buggy Q Setting for CBR/ABR To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/lame/+bug/2139642/+subscriptions -- desktop-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/desktop-bugs
