This includes some default values of #defined constants used in the
code; it won't compile without it.

This patch prepares support for floating point samples (and probably
other formats).  It changes the meaning of the "bits" attribute from a
bit count to a symbolic value.

Copied sources from
http://beesbuzz.biz/code/audiocompress/AudioCompress-2.0.tar.gz

[mk: created this patch under fluffy's name and fixed some gcc
signed/unsigned comparison warnings]

This fixes a regression due to a typo caused by "decoder: use
audio_format_init_checked()".

Support 32 bit samples with software mixer.

Assign default value only if none was given on the command line.

Conflicts:

	NEWS
	configure.ac

Removed the "vtrack" local variable (which triggered a gcc warning
because it was after the newly introduced NULL check), and run
strtol() on the original parameter.

The function flac_vtrack_tnum() was missing a strrchr()==NULL check.

Allow RIFF/AIFF ID3 tags up to 4 MB (old limit was 256 kB).  This
might still be too small for some users, and when somebody complains,
we might do something more clever (like streaming the data into
libid3tag?).

On some platforms, libavcodec wants the output buffer aligned to 16
bytes (because it uses SSE/Altivec internally).  It will segfault when
you don't obey this rule.

Pass the audiofile_setup_sample_format() result to
audio_format_init_checked().

Don't maintain the current time stamp in a floating point variable,
because this is subject to rounding errors.

More exact total time.

Negative return values are not documented here, but since the function
prototype is signed, let's be sure.

Don't maintain the current time stamp in a floating point variable,
because this is subject to rounding errors.

The new option "sample_rate" sets the sample rate for libmikmod.

These functions are trivial, we don't need them separate.

Don't allocate this object, put it on the stack.

Unified function for converting an audio_format object to a string,
for log messages and for the "status" command.