Release Notes: The main point of this release is the fixing of incorrect logic in libmpg123 so that the second (and third, and fourth) decoding of one file after seeking back to the beginning is as gapless as the first. There are some portability improvements (Solaris and Mac OS X 10.6) and a fix for stream input via standard input for the mpg123 binary on MS Windows.
Release Notes: A set of bugs have been fixed, along with some long- and short-standing regressions, namely NtoM CPU creep and parsing of Foobar2000-style RVA info in ID3v2 tags.
Release Notes: Free format support and ARM assembly optimizations were added. Dither noise is now runtime-generated, so the binary is less bloated with dithered decoders, and those are therefore included by default now. Several bugs have been fixed relating to UTF16 in ID3v2 tags and bad frame fillup for 8-bit samples. Generally the ID3v2 parser is more flexible with encodings. libmpg123 also offers a switch to defer the transcoding to UTF8 (so that a client application can handle strange encodings in disguise).
Release Notes: This update mainly adds the proper integer/fixed point decoding that accidentally had been promised for the 1.8.0 release. Minor bugs and build system problems were fixed.
Release Notes: This is a massively important release. There are major performance improvements (including new SSE code) for x86-64 systems (and others), as well as fixes to the correctness of decoding. For the first time, the fixed point decoding (compile-time option) works properly for layer 1 and 2.
Release Notes: A bad-sounding regression with seeking in layer 3 files has been fixed. This regression had been introduced in 1.7.0 with the --doublespeed revival.
Release Notes: This release fixes a vulnerability in the ID3v2 code with invalid text encoding values.
Release Notes: This release fixes a regression that prevented the RVA changes in the generic control interface or from the terminal from resulting in something other than silence.
Release Notes: This release features significant overhaul of the decoder engine to be more flexible at runtime and include more output formats. Float and 32-bit integer output are now standard features of a normal build. The --doublespeed functionality is back again after being silently absent in the 1.x series. Also, there is now support for configuring minimal libmpg123 builds by stripping out certain features.
Release Notes: This release fixes an issue with flexible resampling and the bad frame fill up introduced in the 1.6 series. Also, resampled output should now be consistent across seeks.