shithub: sox

Info  •  Files  •  Log  •  Branches

Change MP3 input percision back to 16 bits to make default behavior
still be to write 16 bit output files.

--- a/ChangeLog

+++ b/ChangeLog

@@ -37,8 +37,6 @@

   o Added new variants of sox_open to allow read/write from/to memory

     buffers (in POSIX 2008 environment); see example5.c.  (robs)

-  o amrnb and amrwb formats can optionally use opencore-amr libraries.

-    (cbagwell)

 File formats:

@@ -54,6 +52,8 @@

     based formats).  (cbagwell)

   o Allow libsndfile to be dlopen()'ed at runtime if --enable-dl-sndfile

     is used. (Doug Cook)

+  o amrnb and amrwb formats can optionally use opencore-amr libraries.

+    (cbagwell)

 Audio device drivers:

--- a/soxformat.7

+++ b/soxformat.7

@@ -402,6 +402,12 @@

 from source code, or, in some cases, obtain pre-built dynamically loadable

 libraries.

 .SP

+When reading MP3 files, up to 28 bits of precision is stored although

+only 16 bits is reported to user.  This is to allow default behavior

+of writing 16 bit output files.  A user can specify a higher precision

+for the output file to prevent lossing this extra information.  MP3

+output files will use up to 24 bits of precision while encoding.

+.SP

 MP3 compression parameters can be selected using SoX's \fB\-C\fR option

 as follows

 (note that the current syntax is subject to change):

--- a/src/mp3.c

+++ b/src/mp3.c

@@ -66,8 +66,9 @@

 #define MP3_LAME_PRECISION 24

 /* MAD returns values with MAD_F_FRACBITS (28) bits of precision, though it's

-   not certain that all of them are meaningful. Let's just make it match LAME. */

-#define MP3_MAD_PRECISION  MP3_LAME_PRECISION

+   not certain that all of them are meaningful. Default to 16 bits to

+   align with most users expectation of output file should be 16 bits. */

+#define MP3_MAD_PRECISION  16

 static const char* const mad_library_names[] =

 {