Applying software volume to S16 samples means several bits of
precision are lost; at 25% volume, two bits are lost.  Additionally,
dithering adds some noise.

The problem gets worse when you apply the software volume code twice:
for the software mixer volume, and again for the replay gain.  This
loses some more precision and adds even more dithering noise, which
can become audible (see
https://github.com/MusicPlayerDaemon/MPD/issues/542).

By converting everything to 24 bit, we need to shift only two bits to
the right instead of ten, losing nearly no precision, and dithering is
not needed.  Even if the output device is unable to play S24 directly,
we can convert back to S16 with only one stage of dithering.

Closes https://github.com/MusicPlayerDaemon/MPD/issues/542

Prepare for a new mode which may convert to a different sample format
when applying volume, to reduce dithering.

Combining these two in one single value saves some memory, but is
complicated and we may need the src_sample_format for new features
later.

Similar to commit 32380d1d, these are
the final parts for really fixing
https://github.com/MusicPlayerDaemon/MPD/issues/469

These options are exclusive.

First part of the "real" fix for
https://github.com/MusicPlayerDaemon/MPD/issues/469

Preparing to add more state.

The dop_buffer will be moved out soon.

Reduce header dependencies.

Let the constructor and destructor do this.  This means that all users
have to be converted to allocate PcmConvert dynamically.