vlc_aout: remove comments copied from libmad

include/vlc_aout.h

@@ -76,31 +76,9 @@
        || ((p_format)->i_format == VLC_CODEC_A52)       \
        || ((p_format)->i_format == VLC_CODEC_DTS) )
 
-/* This is heavily borrowed from libmad, by Robert Leslie <rob@mars.org> */
-/*
- * Fixed-point format: 0xABBBBBBB
- * A == whole part      (sign + 3 bits)
- * B == fractional part (28 bits)
- *
- * Values are signed two's complement, so the effective range is:
- * 0x80000000 to 0x7fffffff
- *       -8.0 to +7.9999999962747097015380859375
- *
- * The smallest representable value is:
- * 0x00000001 == 0.0000000037252902984619140625 (i.e. about 3.725e-9)
- *
- * 28 bits of fractional accuracy represent about
- * 8.6 digits of decimal accuracy.
- *
- * Fixed-point numbers can be added or subtracted as normal
- * integers, but multiplication requires shifting the 64-bit result
- * from 56 fractional bits back to 28 (and rounding.)
- */
 typedef int32_t vlc_fixed_t;
 #define FIXED32_FRACBITS 28
-#define FIXED32_MIN ((vlc_fixed_t) -0x80000000L)
-#define FIXED32_MAX ((vlc_fixed_t) +0x7fffffffL)
-#define FIXED32_ONE ((vlc_fixed_t) 0x10000000)
+#define FIXED32_ONE      (1 << FIXED32_FRACBITS)
 
 /* Values used for the audio-device and audio-channels object variables */
 #define AOUT_VAR_MONO               1