[FFmpeg-devel] [PATCH 12/15] libavcodec: Implementation of AAC_fixed_decoder (PS-module) [3/4]

Nedeljko Babic nedeljko.babic at imgtec.com
Fri Nov 7 16:08:16 CET 2014


From: Djordje Pesut <djordje.pesut at imgtec.com>

Add fixed point implementation.

Signed-off-by: Nedeljko Babic <nedeljko.babic at imgtec.com>
---
 libavcodec/Makefile            |  14 +-
 libavcodec/aac_defines.h       |  36 +++++
 libavcodec/aacps.c             | 353 ++++++++++++++++++++++++++++++++++-------
 libavcodec/aacps.h             |  28 ++--
 libavcodec/aacps_fixed.c       |  24 +++
 libavcodec/aacps_float.c       | 157 +-----------------
 libavcodec/aacpsdata.c         |   6 +-
 libavcodec/aacpsdsp.c          | 216 -------------------------
 libavcodec/aacpsdsp.h          |  30 ++--
 libavcodec/aacpsdsp_fixed.c    |  23 +++
 libavcodec/aacpsdsp_float.c    |  23 +++
 libavcodec/aacpsdsp_template.c | 228 ++++++++++++++++++++++++++
 12 files changed, 671 insertions(+), 467 deletions(-)
 create mode 100644 libavcodec/aacps_fixed.c
 delete mode 100644 libavcodec/aacpsdsp.c
 create mode 100644 libavcodec/aacpsdsp_fixed.c
 create mode 100644 libavcodec/aacpsdsp_float.c
 create mode 100644 libavcodec/aacpsdsp_template.c

diff --git a/libavcodec/Makefile b/libavcodec/Makefile
index bfaeadf..5cbfa4b 100644
--- a/libavcodec/Makefile
+++ b/libavcodec/Makefile
@@ -102,12 +102,12 @@ OBJS-$(CONFIG_WMA_FREQS)               += wma_freqs.o
 OBJS-$(CONFIG_ZERO12V_DECODER)         += 012v.o
 OBJS-$(CONFIG_A64MULTI_ENCODER)        += a64multienc.o elbg.o
 OBJS-$(CONFIG_A64MULTI5_ENCODER)       += a64multienc.o elbg.o
-OBJS-$(CONFIG_AAC_DECODER)             += aacdec.o aactab.o aacsbr.o aacps.o \
+OBJS-$(CONFIG_AAC_DECODER)             += aacdec.o aactab.o aacsbr.o aacps_float.o \
                                           aacadtsdec.o mpeg4audio.o kbdwin.o \
-                                          sbrdsp.o aacpsdsp.o
-OBJS-$(CONFIG_AAC_FIXED_DECODER)       += aacdec_fixed.o aactab.o aacsbr_fixed.o \
+                                          sbrdsp.o aacpsdsp_float.o
+OBJS-$(CONFIG_AAC_FIXED_DECODER)       += aacdec_fixed.o aactab.o aacsbr_fixed.o aacps_fixed.o \
                                           aacadtsdec.o mpeg4audio.o kbdwin.o \
-                                          sbrdsp_fixed.o
+                                          sbrdsp_fixed.o aacpsdsp_fixed.o
 OBJS-$(CONFIG_AAC_ENCODER)             += aacenc.o aaccoder.o    \
                                           aacpsy.o aactab.o      \
                                           psymodel.o mpeg4audio.o kbdwin.o
@@ -880,6 +880,7 @@ TOOLS = fourcc2pixfmt
 
 HOSTPROGS = aac_tablegen                                                \
             aacps_tablegen                                              \
+            aacps_fixed_tablegen                                        \
             aacsbr_tablegen                                             \
             aacsbr_fixed_tablegen                                       \
             cabac_tablegen                                              \
@@ -912,7 +913,7 @@ else
 $(SUBDIR)%_tablegen$(HOSTEXESUF): HOSTCFLAGS += -DCONFIG_SMALL=0
 endif
 
-GEN_HEADERS = cabac_tables.h cbrt_tables.h cbrt_fixed_tables.h aacps_tables.h aacsbr_tables.h \
+GEN_HEADERS = cabac_tables.h cbrt_tables.h cbrt_fixed_tables.h aacps_tables.h aacps_fixed_tables.h aacsbr_tables.h \
               aacsbr_fixed_tables.h aac_tables.h dsd_tables.h dv_tables.h     \
               sinewin_tables.h sinewin_fixed_tables.h mpegaudio_tables.h motionpixels_tables.h \
               pcm_tables.h qdm2_tables.h
@@ -924,7 +925,8 @@ $(GEN_HEADERS): $(SUBDIR)%_tables.h: $(SUBDIR)%_tablegen$(HOSTEXESUF)
 ifdef CONFIG_HARDCODED_TABLES
 $(SUBDIR)aacdec.o: $(SUBDIR)cbrt_tables.h
 $(SUBDIR)aacdec_fixed.o: $(SUBDIR)cbrt_fixed_tables.h
-$(SUBDIR)aacps.o: $(SUBDIR)aacps_tables.h
+$(SUBDIR)aacps_float.o: $(SUBDIR)aacps_tables.h
+$(SUBDIR)aacps_fixed.o: $(SUBDIR)aacps_fixed_tables.h
 $(SUBDIR)aacsbr.o: $(SUBDIR)aacsbr_tables.h
 $(SUBDIR)aacsbr_fixed.o: $(SUBDIR)aacsbr_fixed_tables.h
 $(SUBDIR)aactab.o: $(SUBDIR)aac_tables.h
diff --git a/libavcodec/aac_defines.h b/libavcodec/aac_defines.h
index 3ba9957..5c2b181 100644
--- a/libavcodec/aac_defines.h
+++ b/libavcodec/aac_defines.h
@@ -33,6 +33,7 @@
 #define AAC_RENAME(x)       x ## _fixed
 #define AAC_RENAME_32(x)    x ## _fixed_32
 #define INTFLOAT int
+#define INT64FLOAT          int64_t
 #define SHORTFLOAT int16_t
 #define AAC_FLOAT SoftFloat
 #define AAC_SIGNE           int
@@ -43,9 +44,33 @@
 #define Q31(x)              (int)((x)*2147483648.0 + 0.5)
 #define RANGE15(x)          x
 #define GET_GAIN(x, y)      (-(y) << (x)) + 1024
+#define AAC_MUL16(x, y)     (int)(((int64_t)(x) * (y) + 0x8000) >> 16)
 #define AAC_MUL26(x, y)     (int)(((int64_t)(x) * (y) + 0x2000000) >> 26)
 #define AAC_MUL30(x, y)     (int)(((int64_t)(x) * (y) + 0x20000000) >> 30)
 #define AAC_MUL31(x, y)     (int)(((int64_t)(x) * (y) + 0x40000000) >> 31)
+#define AAC_MADD28(x, y, a, b) (int)((((int64_t)(x) * (y)) + \
+                                      ((int64_t)(a) * (b)) + \
+                                        0x8000000) >> 28)
+#define AAC_MADD30(x, y, a, b) (int)((((int64_t)(x) * (y)) + \
+                                      ((int64_t)(a) * (b)) + \
+                                        0x20000000) >> 30)
+#define AAC_MADD30_V8(x, y, a, b, c, d, e, f) (int)((((int64_t)(x) * (y)) + \
+                                                     ((int64_t)(a) * (b)) + \
+                                                     ((int64_t)(c) * (d)) + \
+                                                     ((int64_t)(e) * (f)) + \
+                                                       0x20000000) >> 30)
+#define AAC_MSUB30(x, y, a, b) (int)((((int64_t)(x) * (y)) - \
+                                      ((int64_t)(a) * (b)) + \
+                                        0x20000000) >> 30)
+#define AAC_MSUB30_V8(x, y, a, b, c, d, e, f) (int)((((int64_t)(x) * (y)) + \
+                                                     ((int64_t)(a) * (b)) - \
+                                                     ((int64_t)(c) * (d)) - \
+                                                     ((int64_t)(e) * (f)) + \
+                                                       0x20000000) >> 30)
+#define AAC_MSUB31_V3(x, y, z)    (int)((((int64_t)(x) * (z)) - \
+                                      ((int64_t)(y) * (z)) + \
+                                        0x40000000) >> 31)
+#define AAC_HALF_SUM(x, y)  (x) >> 1 + (y) >> 1
 #define AAC_SRA_R(x, y)     (int)(((x) + (1 << ((y) - 1))) >> (y))
 
 #else
@@ -56,6 +81,7 @@
 #define AAC_RENAME(x)       x
 #define AAC_RENAME_32(x)    x
 #define INTFLOAT float
+#define INT64FLOAT          float
 #define SHORTFLOAT float
 #define AAC_FLOAT float
 #define AAC_SIGNE           unsigned
@@ -66,9 +92,19 @@
 #define Q31(x)              x
 #define RANGE15(x)          (32768.0 * (x))
 #define GET_GAIN(x, y)      powf((x), -(y))
+#define AAC_MUL16(x, y)     ((x) * (y))
 #define AAC_MUL26(x, y)     ((x) * (y))
 #define AAC_MUL30(x, y)     ((x) * (y))
 #define AAC_MUL31(x, y)     ((x) * (y))
+#define AAC_MADD28(x, y, a, b) ((x) * (y) + (a) * (b))
+#define AAC_MADD30(x, y, a, b) ((x) * (y) + (a) * (b))
+#define AAC_MADD30_V8(x, y, a, b, c, d, e, f) ((x) * (y) + (a) * (b) + \
+                                               (c) * (d) + (e) * (f))
+#define AAC_MSUB30(x, y, a, b) ((x) * (y) - (a) * (b))
+#define AAC_MSUB30_V8(x, y, a, b, c, d, e, f) ((x) * (y) + (a) * (b) - \
+                                               (c) * (d) - (e) * (f))
+#define AAC_MSUB31_V3(x, y, z)    ((x) - (y)) * (z)
+#define AAC_HALF_SUM(x, y)  ((x) + (y)) * 0.5f
 #define AAC_SRA_R(x, y)     (x)
 
 #endif /* USE_FIXED */
diff --git a/libavcodec/aacps.c b/libavcodec/aacps.c
index fdaae3f..142df00 100644
--- a/libavcodec/aacps.c
+++ b/libavcodec/aacps.c
@@ -17,16 +17,23 @@
  * You should have received a copy of the GNU Lesser General Public
  * License along with FFmpeg; if not, write to the Free Software
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ *
+ * Note: Rounding-to-nearest used unless otherwise stated
+ *
  */
 
 #include <stdint.h>
 #include "libavutil/common.h"
-#include "libavutil/internal.h"
 #include "libavutil/mathematics.h"
 #include "avcodec.h"
 #include "get_bits.h"
 #include "aacps.h"
+#if USE_FIXED
+#include "aacps_fixed_tablegen.h"
+#else
+#include "libavutil/internal.h"
 #include "aacps_tablegen.h"
+#endif /* USE_FIXED */
 #include "aacpsdata.c"
 
 #define PS_BASELINE 0  ///< Operate in Baseline PS mode
@@ -148,7 +155,7 @@ static void ipdopd_reset(int8_t *ipd_hist, int8_t *opd_hist)
     }
 }
 
-int ff_ps_read_data(AVCodecContext *avctx, GetBitContext *gb_host, PSContext *ps, int bits_left)
+int AAC_RENAME(ff_ps_read_data)(AVCodecContext *avctx, GetBitContext *gb_host, PSContext *ps, int bits_left)
 {
     int e;
     int bit_count_start = get_bits_count(gb_host);
@@ -302,35 +309,42 @@ err:
 
 /** Split one subband into 2 subsubbands with a symmetric real filter.
  * The filter must have its non-center even coefficients equal to zero. */
-static void hybrid2_re(float (*in)[2], float (*out)[32][2], const float filter[8], int len, int reverse)
+static void hybrid2_re(INTFLOAT (*in)[2], INTFLOAT (*out)[32][2], const INTFLOAT filter[8], int len, int reverse)
 {
     int i, j;
+
     for (i = 0; i < len; i++, in++) {
-        float re_in = filter[6] * in[6][0];          //real inphase
-        float re_op = 0.0f;                          //real out of phase
-        float im_in = filter[6] * in[6][1];          //imag inphase
-        float im_op = 0.0f;                          //imag out of phase
+        INT64FLOAT re_in = AAC_MUL31(filter[6], in[6][0]); //real inphase
+        INT64FLOAT re_op = 0.0f;                          //real out of phase
+        INT64FLOAT im_in = AAC_MUL31(filter[6], in[6][1]); //imag inphase
+        INT64FLOAT im_op = 0.0f;                          //imag out of phase
         for (j = 0; j < 6; j += 2) {
-            re_op += filter[j+1] * (in[j+1][0] + in[12-j-1][0]);
-            im_op += filter[j+1] * (in[j+1][1] + in[12-j-1][1]);
+            re_op += (INT64FLOAT)filter[j+1] * (in[j+1][0] + in[12-j-1][0]);
+            im_op += (INT64FLOAT)filter[j+1] * (in[j+1][1] + in[12-j-1][1]);
         }
-        out[ reverse][i][0] = re_in + re_op;
-        out[ reverse][i][1] = im_in + im_op;
-        out[!reverse][i][0] = re_in - re_op;
-        out[!reverse][i][1] = im_in - im_op;
+
+#if USE_FIXED
+        re_op = (re_op + 0x40000000) >> 31;
+        im_op = (im_op + 0x40000000) >> 31;
+#endif /* USE_FIXED */
+
+        out[ reverse][i][0] = (INTFLOAT)(re_in + re_op);
+        out[ reverse][i][1] = (INTFLOAT)(im_in + im_op);
+        out[!reverse][i][0] = (INTFLOAT)(re_in - re_op);
+        out[!reverse][i][1] = (INTFLOAT)(im_in - im_op);
     }
 }
 
 /** Split one subband into 6 subsubbands with a complex filter */
-static void hybrid6_cx(PSDSPContext *dsp, float (*in)[2], float (*out)[32][2],
-                       TABLE_CONST float (*filter)[8][2], int len)
+static void hybrid6_cx(PSDSPContext *dsp, INTFLOAT (*in)[2], INTFLOAT (*out)[32][2],
+                       TABLE_CONST INTFLOAT (*filter)[8][2], int len)
 {
     int i;
     int N = 8;
-    LOCAL_ALIGNED_16(float, temp, [8], [2]);
+    LOCAL_ALIGNED_16(INTFLOAT, temp, [8], [2]);
 
     for (i = 0; i < len; i++, in++) {
-        dsp->hybrid_analysis(temp, in, (const float (*)[8][2]) filter, 1, N);
+        dsp->hybrid_analysis(temp, in, (const INTFLOAT (*)[8][2]) filter, 1, N);
         out[0][i][0] = temp[6][0];
         out[0][i][1] = temp[6][1];
         out[1][i][0] = temp[7][0];
@@ -347,18 +361,18 @@ static void hybrid6_cx(PSDSPContext *dsp, float (*in)[2], float (*out)[32][2],
 }
 
 static void hybrid4_8_12_cx(PSDSPContext *dsp,
-                            float (*in)[2], float (*out)[32][2],
-                            TABLE_CONST float (*filter)[8][2], int N, int len)
+                            INTFLOAT (*in)[2], INTFLOAT (*out)[32][2],
+                            TABLE_CONST INTFLOAT (*filter)[8][2], int N, int len)
 {
     int i;
 
     for (i = 0; i < len; i++, in++) {
-        dsp->hybrid_analysis(out[0] + i, in, (const float (*)[8][2]) filter, 32, N);
+        dsp->hybrid_analysis(out[0] + i, in, (const INTFLOAT (*)[8][2]) filter, 32, N);
     }
 }
 
-static void hybrid_analysis(PSDSPContext *dsp, float out[91][32][2],
-                            float in[5][44][2], float L[2][38][64],
+static void hybrid_analysis(PSDSPContext *dsp, INTFLOAT out[91][32][2],
+                            INTFLOAT in[5][44][2], INTFLOAT L[2][38][64],
                             int is34, int len)
 {
     int i, j;
@@ -387,8 +401,8 @@ static void hybrid_analysis(PSDSPContext *dsp, float out[91][32][2],
     }
 }
 
-static void hybrid_synthesis(PSDSPContext *dsp, float out[2][38][64],
-                             float in[91][32][2], int is34, int len)
+static void hybrid_synthesis(PSDSPContext *dsp, INTFLOAT out[2][38][64],
+                             INTFLOAT in[91][32][2], int is34, int len)
 {
     int i, n;
     if (is34) {
@@ -429,7 +443,7 @@ static void hybrid_synthesis(PSDSPContext *dsp, float out[2][38][64],
 }
 
 /// All-pass filter decay slope
-#define DECAY_SLOPE      0.05f
+#define DECAY_SLOPE      Q30(0.05f)
 /// Number of frequency bands that can be addressed by the parameter index, b(k)
 static const int   NR_PAR_BANDS[]      = { 20, 34 };
 static const int   NR_IPDOPD_BANDS[]   = { 11, 17 };
@@ -483,28 +497,43 @@ static void map_idx_34_to_20(int8_t *par_mapped, const int8_t *par, int full)
     }
 }
 
-static void map_val_34_to_20(float par[PS_MAX_NR_IIDICC])
+static void map_val_34_to_20(INTFLOAT par[PS_MAX_NR_IIDICC])
 {
+#if USE_FIXED
+    par[ 0] = (int)(((int64_t)(par[ 0] + (par[ 1]>>1)) * 1431655765 + \
+                      0x40000000) >> 31);
+    par[ 1] = (int)(((int64_t)((par[ 1]>>1) + par[ 2]) * 1431655765 + \
+                      0x40000000) >> 31);
+    par[ 2] = (int)(((int64_t)(par[ 3] + (par[ 4]>>1)) * 1431655765 + \
+                      0x40000000) >> 31);
+    par[ 3] = (int)(((int64_t)((par[ 4]>>1) + par[ 5]) * 1431655765 + \
+                      0x40000000) >> 31);
+#else
     par[ 0] = (2*par[ 0] +   par[ 1]) * 0.33333333f;
     par[ 1] = (  par[ 1] + 2*par[ 2]) * 0.33333333f;
     par[ 2] = (2*par[ 3] +   par[ 4]) * 0.33333333f;
     par[ 3] = (  par[ 4] + 2*par[ 5]) * 0.33333333f;
-    par[ 4] = (  par[ 6] +   par[ 7]) * 0.5f;
-    par[ 5] = (  par[ 8] +   par[ 9]) * 0.5f;
+#endif /* USE_FIXED */
+    par[ 4] = AAC_HALF_SUM(par[ 6], par[ 7]);
+    par[ 5] = AAC_HALF_SUM(par[ 8], par[ 9]);
     par[ 6] =    par[10];
     par[ 7] =    par[11];
-    par[ 8] = (  par[12] +   par[13]) * 0.5f;
-    par[ 9] = (  par[14] +   par[15]) * 0.5f;
+    par[ 8] = AAC_HALF_SUM(par[12], par[13]);
+    par[ 9] = AAC_HALF_SUM(par[14], par[15]);
     par[10] =    par[16];
     par[11] =    par[17];
     par[12] =    par[18];
     par[13] =    par[19];
-    par[14] = (  par[20] +   par[21]) * 0.5f;
-    par[15] = (  par[22] +   par[23]) * 0.5f;
-    par[16] = (  par[24] +   par[25]) * 0.5f;
-    par[17] = (  par[26] +   par[27]) * 0.5f;
+    par[14] = AAC_HALF_SUM(par[20], par[21]);
+    par[15] = AAC_HALF_SUM(par[22], par[23]);
+    par[16] = AAC_HALF_SUM(par[24], par[25]);
+    par[17] = AAC_HALF_SUM(par[26], par[27]);
+#if USE_FIXED
+    par[18] = (((par[28]+2)>>2) + ((par[29]+2)>>2) + ((par[30]+2)>>2) + ((par[31]+2)>>2));
+#else
     par[18] = (  par[28] +   par[29] +   par[30] +   par[31]) * 0.25f;
-    par[19] = (  par[32] +   par[33]) * 0.5f;
+#endif /* USE_FIXED */
+    par[19] = AAC_HALF_SUM(par[32], par[33]);
 }
 
 static void map_idx_10_to_34(int8_t *par_mapped, const int8_t *par, int full)
@@ -589,7 +618,7 @@ static void map_idx_20_to_34(int8_t *par_mapped, const int8_t *par, int full)
     par_mapped[ 0] =  par[ 0];
 }
 
-static void map_val_20_to_34(float par[PS_MAX_NR_IIDICC])
+static void map_val_20_to_34(INTFLOAT par[PS_MAX_NR_IIDICC])
 {
     par[33] =  par[19];
     par[32] =  par[19];
@@ -620,27 +649,29 @@ static void map_val_20_to_34(float par[PS_MAX_NR_IIDICC])
     par[ 7] =  par[ 4];
     par[ 6] =  par[ 4];
     par[ 5] =  par[ 3];
-    par[ 4] = (par[ 2] + par[ 3]) * 0.5f;
+    par[ 4] = AAC_HALF_SUM(par[ 2], par[ 3]);
     par[ 3] =  par[ 2];
     par[ 2] =  par[ 1];
-    par[ 1] = (par[ 0] + par[ 1]) * 0.5f;
+    par[ 1] = AAC_HALF_SUM(par[ 0], par[ 1]);
 }
 
-static void decorrelation(PSContext *ps, float (*out)[32][2], const float (*s)[32][2], int is34)
+static void decorrelation(PSContext *ps, INTFLOAT (*out)[32][2], const INTFLOAT (*s)[32][2], int is34)
 {
-    LOCAL_ALIGNED_16(float, power, [34], [PS_QMF_TIME_SLOTS]);
-    LOCAL_ALIGNED_16(float, transient_gain, [34], [PS_QMF_TIME_SLOTS]);
-    float *peak_decay_nrg = ps->peak_decay_nrg;
-    float *power_smooth = ps->power_smooth;
-    float *peak_decay_diff_smooth = ps->peak_decay_diff_smooth;
-    float (*delay)[PS_QMF_TIME_SLOTS + PS_MAX_DELAY][2] = ps->delay;
-    float (*ap_delay)[PS_AP_LINKS][PS_QMF_TIME_SLOTS + PS_MAX_AP_DELAY][2] = ps->ap_delay;
-    const int8_t *k_to_i = is34 ? k_to_i_34 : k_to_i_20;
-    const float peak_decay_factor = 0.76592833836465f;
+    LOCAL_ALIGNED_16(INTFLOAT, power, [34], [PS_QMF_TIME_SLOTS]);
+    LOCAL_ALIGNED_16(INTFLOAT, transient_gain, [34], [PS_QMF_TIME_SLOTS]);
+    INTFLOAT *peak_decay_nrg = ps->peak_decay_nrg;
+    INTFLOAT *power_smooth = ps->power_smooth;
+    INTFLOAT *peak_decay_diff_smooth = ps->peak_decay_diff_smooth;
+    INTFLOAT (*delay)[PS_QMF_TIME_SLOTS + PS_MAX_DELAY][2] = ps->delay;
+    INTFLOAT (*ap_delay)[PS_AP_LINKS][PS_QMF_TIME_SLOTS + PS_MAX_AP_DELAY][2] = ps->ap_delay;
+#if !USE_FIXED
     const float transient_impact  = 1.5f;
     const float a_smooth          = 0.25f; ///< Smoothing coefficient
+#endif /* USE_FIXED */
+    const int8_t *k_to_i = is34 ? k_to_i_34 : k_to_i_20;
     int i, k, m, n;
     int n0 = 0, nL = 32;
+    const INTFLOAT peak_decay_factor = Q31(0.76592833836465f);;
 
     memset(power, 0, 34 * sizeof(*power));
 
@@ -658,19 +689,49 @@ static void decorrelation(PSContext *ps, float (*out)[32][2], const float (*s)[3
     }
 
     //Transient detection
+#if USE_FIXED
+    for (i = 0; i < NR_PAR_BANDS[is34]; i++) {
+        for (n = n0; n < nL; n++) {
+            int decayed_peak;
+            int denom;
+
+            decayed_peak = (int)(((int64_t)peak_decay_factor * \
+                                           peak_decay_nrg[i] + 0x40000000) >> 31);
+            peak_decay_nrg[i] = FFMAX(decayed_peak, power[i][n]);
+            power_smooth[i] += (power[i][n] - power_smooth[i] + 2) >> 2;
+            peak_decay_diff_smooth[i] += (peak_decay_nrg[i] - power[i][n] - \
+                                          peak_decay_diff_smooth[i] + 2) >> 2;
+            denom = peak_decay_diff_smooth[i] + (peak_decay_diff_smooth[i] >> 1);
+            if (denom > power_smooth[i]) {
+              int p = power_smooth[i];
+              while (denom < 0x40000000) {
+                denom <<= 1;
+                p <<= 1;
+              }
+              transient_gain[i][n] = p / (denom >> 16);
+            }
+            else {
+              transient_gain[i][n] = 1 << 16;
+            }
+        }
+    }
+#else
     for (i = 0; i < NR_PAR_BANDS[is34]; i++) {
         for (n = n0; n < nL; n++) {
             float decayed_peak = peak_decay_factor * peak_decay_nrg[i];
             float denom;
             peak_decay_nrg[i] = FFMAX(decayed_peak, power[i][n]);
             power_smooth[i] += a_smooth * (power[i][n] - power_smooth[i]);
-            peak_decay_diff_smooth[i] += a_smooth * (peak_decay_nrg[i] - power[i][n] - peak_decay_diff_smooth[i]);
+            peak_decay_diff_smooth[i] += a_smooth * (peak_decay_nrg[i] - \
+                                                     power[i][n] - \
+                                                     peak_decay_diff_smooth[i]);
             denom = transient_impact * peak_decay_diff_smooth[i];
             transient_gain[i][n]   = (denom > power_smooth[i]) ?
                                          power_smooth[i] / denom : 1.0f;
         }
     }
 
+#endif /* USE_FIXED */
     //Decorrelation and transient reduction
     //                         PS_AP_LINKS - 1
     //                               -----
@@ -681,8 +742,22 @@ static void decorrelation(PSContext *ps, float (*out)[32][2], const float (*s)[3
     //d[k][z] (out) = transient_gain_mapped[k][z] * H[k][z] * s[k][z]
     for (k = 0; k < NR_ALLPASS_BANDS[is34]; k++) {
         int b = k_to_i[k];
-        float g_decay_slope = 1.f - DECAY_SLOPE * (k - DECAY_CUTOFF[is34]);
-        g_decay_slope = av_clipf(g_decay_slope, 0.f, 1.f);
+#if USE_FIXED
+        int g_decay_slope;
+
+        if (k - DECAY_CUTOFF[is34] <= 0) {
+          g_decay_slope = 1 << 30;
+        }
+        else if (k - DECAY_CUTOFF[is34] >= 20) {
+          g_decay_slope = 0;
+        }
+        else {
+          g_decay_slope = (1 << 30) - DECAY_SLOPE * (k - DECAY_CUTOFF[is34]);
+        }
+#else
+         float g_decay_slope = 1.f - DECAY_SLOPE * (k - DECAY_CUTOFF[is34]);
+         g_decay_slope = av_clipf(g_decay_slope, 0.f, 1.f);
+#endif /* USE_FIXED */
         memcpy(delay[k], delay[k]+nL, PS_MAX_DELAY*sizeof(delay[k][0]));
         memcpy(delay[k]+PS_MAX_DELAY, s[k], numQMFSlots*sizeof(delay[k][0]));
         for (m = 0; m < PS_AP_LINKS; m++) {
@@ -690,7 +765,7 @@ static void decorrelation(PSContext *ps, float (*out)[32][2], const float (*s)[3
         }
         ps->dsp.decorrelate(out[k], delay[k] + PS_MAX_DELAY - 2, ap_delay[k],
                             phi_fract[is34][k],
-                            (const float (*)[2]) Q_fract_allpass[is34][k],
+                            (const INTFLOAT (*)[2]) Q_fract_allpass[is34][k],
                             transient_gain[b], g_decay_slope, nL - n0);
     }
     for (; k < SHORT_DELAY_BAND[is34]; k++) {
@@ -749,10 +824,170 @@ static void remap20(int8_t (**p_par_mapped)[PS_MAX_NR_IIDICC],
     }
 }
 
-int ff_ps_apply(AVCodecContext *avctx, PSContext *ps, float L[2][38][64], float R[2][38][64], int top)
+static void stereo_processing(PSContext *ps, INTFLOAT (*l)[32][2], INTFLOAT (*r)[32][2], int is34)
+{
+    int e, b, k;
+
+    INTFLOAT (*H11)[PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC] = ps->H11;
+    INTFLOAT (*H12)[PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC] = ps->H12;
+    INTFLOAT (*H21)[PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC] = ps->H21;
+    INTFLOAT (*H22)[PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC] = ps->H22;
+    int8_t *opd_hist = ps->opd_hist;
+    int8_t *ipd_hist = ps->ipd_hist;
+    int8_t iid_mapped_buf[PS_MAX_NUM_ENV][PS_MAX_NR_IIDICC];
+    int8_t icc_mapped_buf[PS_MAX_NUM_ENV][PS_MAX_NR_IIDICC];
+    int8_t ipd_mapped_buf[PS_MAX_NUM_ENV][PS_MAX_NR_IIDICC];
+    int8_t opd_mapped_buf[PS_MAX_NUM_ENV][PS_MAX_NR_IIDICC];
+    int8_t (*iid_mapped)[PS_MAX_NR_IIDICC] = iid_mapped_buf;
+    int8_t (*icc_mapped)[PS_MAX_NR_IIDICC] = icc_mapped_buf;
+    int8_t (*ipd_mapped)[PS_MAX_NR_IIDICC] = ipd_mapped_buf;
+    int8_t (*opd_mapped)[PS_MAX_NR_IIDICC] = opd_mapped_buf;
+    const int8_t *k_to_i = is34 ? k_to_i_34 : k_to_i_20;
+    TABLE_CONST INTFLOAT (*H_LUT)[8][4] = (PS_BASELINE || ps->icc_mode < 3) ? HA : HB;
+
+    //Remapping
+    if (ps->num_env_old) {
+        memcpy(H11[0][0], H11[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H11[0][0][0]));
+        memcpy(H11[1][0], H11[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H11[1][0][0]));
+        memcpy(H12[0][0], H12[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H12[0][0][0]));
+        memcpy(H12[1][0], H12[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H12[1][0][0]));
+        memcpy(H21[0][0], H21[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H21[0][0][0]));
+        memcpy(H21[1][0], H21[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H21[1][0][0]));
+        memcpy(H22[0][0], H22[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H22[0][0][0]));
+        memcpy(H22[1][0], H22[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H22[1][0][0]));
+    }
+
+    if (is34) {
+        remap34(&iid_mapped, ps->iid_par, ps->nr_iid_par, ps->num_env, 1);
+        remap34(&icc_mapped, ps->icc_par, ps->nr_icc_par, ps->num_env, 1);
+        if (ps->enable_ipdopd) {
+            remap34(&ipd_mapped, ps->ipd_par, ps->nr_ipdopd_par, ps->num_env, 0);
+            remap34(&opd_mapped, ps->opd_par, ps->nr_ipdopd_par, ps->num_env, 0);
+        }
+        if (!ps->is34bands_old) {
+            map_val_20_to_34(H11[0][0]);
+            map_val_20_to_34(H11[1][0]);
+            map_val_20_to_34(H12[0][0]);
+            map_val_20_to_34(H12[1][0]);
+            map_val_20_to_34(H21[0][0]);
+            map_val_20_to_34(H21[1][0]);
+            map_val_20_to_34(H22[0][0]);
+            map_val_20_to_34(H22[1][0]);
+            ipdopd_reset(ipd_hist, opd_hist);
+        }
+    } else {
+        remap20(&iid_mapped, ps->iid_par, ps->nr_iid_par, ps->num_env, 1);
+        remap20(&icc_mapped, ps->icc_par, ps->nr_icc_par, ps->num_env, 1);
+        if (ps->enable_ipdopd) {
+            remap20(&ipd_mapped, ps->ipd_par, ps->nr_ipdopd_par, ps->num_env, 0);
+            remap20(&opd_mapped, ps->opd_par, ps->nr_ipdopd_par, ps->num_env, 0);
+        }
+        if (ps->is34bands_old) {
+            map_val_34_to_20(H11[0][0]);
+            map_val_34_to_20(H11[1][0]);
+            map_val_34_to_20(H12[0][0]);
+            map_val_34_to_20(H12[1][0]);
+            map_val_34_to_20(H21[0][0]);
+            map_val_34_to_20(H21[1][0]);
+            map_val_34_to_20(H22[0][0]);
+            map_val_34_to_20(H22[1][0]);
+            ipdopd_reset(ipd_hist, opd_hist);
+        }
+    }
+
+    //Mixing
+    for (e = 0; e < ps->num_env; e++) {
+        for (b = 0; b < NR_PAR_BANDS[is34]; b++) {
+            INTFLOAT h11, h12, h21, h22;
+            h11 = H_LUT[iid_mapped[e][b] + 7 + 23 * ps->iid_quant][icc_mapped[e][b]][0];
+            h12 = H_LUT[iid_mapped[e][b] + 7 + 23 * ps->iid_quant][icc_mapped[e][b]][1];
+            h21 = H_LUT[iid_mapped[e][b] + 7 + 23 * ps->iid_quant][icc_mapped[e][b]][2];
+            h22 = H_LUT[iid_mapped[e][b] + 7 + 23 * ps->iid_quant][icc_mapped[e][b]][3];
+
+            if (!PS_BASELINE && ps->enable_ipdopd && b < NR_IPDOPD_BANDS[is34]) {
+                //The spec say says to only run this smoother when enable_ipdopd
+                //is set but the reference decoder appears to run it constantly
+                INTFLOAT h11i, h12i, h21i, h22i;
+                INTFLOAT ipd_adj_re, ipd_adj_im;
+                int opd_idx = opd_hist[b] * 8 + opd_mapped[e][b];
+                int ipd_idx = ipd_hist[b] * 8 + ipd_mapped[e][b];
+                INTFLOAT opd_re = pd_re_smooth[opd_idx];
+                INTFLOAT opd_im = pd_im_smooth[opd_idx];
+                INTFLOAT ipd_re = pd_re_smooth[ipd_idx];
+                INTFLOAT ipd_im = pd_im_smooth[ipd_idx];
+                opd_hist[b] = opd_idx & 0x3F;
+                ipd_hist[b] = ipd_idx & 0x3F;
+
+                ipd_adj_re = AAC_MADD30(opd_re, ipd_re, opd_im, ipd_im);
+                ipd_adj_im = AAC_MSUB30(opd_im, ipd_re, opd_re, ipd_im);
+                h11i = AAC_MUL30(h11,  opd_im);
+                h11  = AAC_MUL30(h11,  opd_re);
+                h12i = AAC_MUL30(h12,  ipd_adj_im);
+                h12  = AAC_MUL30(h12,  ipd_adj_re);
+                h21i = AAC_MUL30(h21,  opd_im);
+                h21  = AAC_MUL30(h21,  opd_re);
+                h22i = AAC_MUL30(h22,  ipd_adj_im);
+                h22  = AAC_MUL30(h22,  ipd_adj_re);
+                H11[1][e+1][b] = h11i;
+                H12[1][e+1][b] = h12i;
+                H21[1][e+1][b] = h21i;
+                H22[1][e+1][b] = h22i;
+            }
+            H11[0][e+1][b] = h11;
+            H12[0][e+1][b] = h12;
+            H21[0][e+1][b] = h21;
+            H22[0][e+1][b] = h22;
+        }
+        for (k = 0; k < NR_BANDS[is34]; k++) {
+            INTFLOAT h[2][4];
+            INTFLOAT h_step[2][4];
+            int start = ps->border_position[e];
+            int stop  = ps->border_position[e+1];
+            INTFLOAT width = Q30(1.f) / (stop - start);
+#if USE_FIXED
+            width <<= 1;
+#endif
+            b = k_to_i[k];
+            h[0][0] = H11[0][e][b];
+            h[0][1] = H12[0][e][b];
+            h[0][2] = H21[0][e][b];
+            h[0][3] = H22[0][e][b];
+            if (!PS_BASELINE && ps->enable_ipdopd) {
+            //Is this necessary? ps_04_new seems unchanged
+            if ((is34 && k <= 13 && k >= 9) || (!is34 && k <= 1)) {
+                h[1][0] = -H11[1][e][b];
+                h[1][1] = -H12[1][e][b];
+                h[1][2] = -H21[1][e][b];
+                h[1][3] = -H22[1][e][b];
+            } else {
+                h[1][0] = H11[1][e][b];
+                h[1][1] = H12[1][e][b];
+                h[1][2] = H21[1][e][b];
+                h[1][3] = H22[1][e][b];
+            }
+            }
+            //Interpolation
+            h_step[0][0] = AAC_MSUB31_V3(H11[0][e+1][b], h[0][0], width);
+            h_step[0][1] = AAC_MSUB31_V3(H12[0][e+1][b], h[0][1], width);
+            h_step[0][2] = AAC_MSUB31_V3(H21[0][e+1][b], h[0][2], width);
+            h_step[0][3] = AAC_MSUB31_V3(H22[0][e+1][b], h[0][3], width);
+            if (!PS_BASELINE && ps->enable_ipdopd) {
+                h_step[1][0] = AAC_MSUB31_V3(H11[1][e+1][b], h[1][0], width);
+                h_step[1][1] = AAC_MSUB31_V3(H12[1][e+1][b], h[1][1], width);
+                h_step[1][2] = AAC_MSUB31_V3(H21[1][e+1][b], h[1][2], width);
+                h_step[1][3] = AAC_MSUB31_V3(H22[1][e+1][b], h[1][3], width);
+            }
+            ps->dsp.stereo_interpolate[!PS_BASELINE && ps->enable_ipdopd](
+                l[k] + start + 1, r[k] + start + 1,
+                h, h_step, stop - start);
+        }
+    }
+}
+
+int AAC_RENAME(ff_ps_apply)(AVCodecContext *avctx, PSContext *ps, INTFLOAT L[2][38][64], INTFLOAT R[2][38][64], int top)
 {
-    LOCAL_ALIGNED_16(float, Lbuf, [91], [32][2]);
-    LOCAL_ALIGNED_16(float, Rbuf, [91], [32][2]);
+    LOCAL_ALIGNED_16(INTFLOAT, Lbuf, [91], [32][2]);
+    LOCAL_ALIGNED_16(INTFLOAT, Rbuf, [91], [32][2]);
     const int len = 32;
     int is34 = ps->is34bands;
 
@@ -762,7 +997,7 @@ int ff_ps_apply(AVCodecContext *avctx, PSContext *ps, float L[2][38][64], float
         memset(ps->ap_delay + top, 0, (NR_ALLPASS_BANDS[is34] - top)*sizeof(ps->ap_delay[0]));
 
     hybrid_analysis(&ps->dsp, Lbuf, ps->in_buf, L, is34, len);
-    decorrelation(ps, Rbuf, (const float (*)[32][2]) Lbuf, is34);
+    decorrelation(ps, Rbuf, (const INTFLOAT (*)[32][2]) Lbuf, is34);
     stereo_processing(ps, Lbuf, Rbuf, is34);
     hybrid_synthesis(&ps->dsp, L, Lbuf, is34, len);
     hybrid_synthesis(&ps->dsp, R, Rbuf, is34, len);
@@ -779,7 +1014,7 @@ int ff_ps_apply(AVCodecContext *avctx, PSContext *ps, float L[2][38][64], float
 #define PS_VLC_ROW(name) \
     { name ## _codes, name ## _bits, sizeof(name ## _codes), sizeof(name ## _codes[0]) }
 
-av_cold void ff_ps_init(void) {
+av_cold void AAC_RENAME(ff_ps_init)(void) {
     // Syntax initialization
     static const struct {
         const void *ps_codes, *ps_bits;
@@ -811,7 +1046,7 @@ av_cold void ff_ps_init(void) {
     ps_tableinit();
 }
 
-av_cold void ff_ps_ctx_init(PSContext *ps)
+av_cold void AAC_RENAME(ff_ps_ctx_init)(PSContext *ps)
 {
-    ff_psdsp_init(&ps->dsp);
+    AAC_RENAME(ff_psdsp_init)(&ps->dsp);
 }
diff --git a/libavcodec/aacps.h b/libavcodec/aacps.h
index 29323ff..4703ce4 100644
--- a/libavcodec/aacps.h
+++ b/libavcodec/aacps.h
@@ -61,24 +61,24 @@ typedef struct PSContext {
     int    is34bands;
     int    is34bands_old;
 
-    DECLARE_ALIGNED(16, float, in_buf)[5][44][2];
-    DECLARE_ALIGNED(16, float, delay)[PS_MAX_SSB][PS_QMF_TIME_SLOTS + PS_MAX_DELAY][2];
-    DECLARE_ALIGNED(16, float, ap_delay)[PS_MAX_AP_BANDS][PS_AP_LINKS][PS_QMF_TIME_SLOTS + PS_MAX_AP_DELAY][2];
-    DECLARE_ALIGNED(16, float, peak_decay_nrg)[34];
-    DECLARE_ALIGNED(16, float, power_smooth)[34];
-    DECLARE_ALIGNED(16, float, peak_decay_diff_smooth)[34];
-    DECLARE_ALIGNED(16, float, H11)[2][PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC];
-    DECLARE_ALIGNED(16, float, H12)[2][PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC];
-    DECLARE_ALIGNED(16, float, H21)[2][PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC];
-    DECLARE_ALIGNED(16, float, H22)[2][PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC];
+    DECLARE_ALIGNED(16, INTFLOAT, in_buf)[5][44][2];
+    DECLARE_ALIGNED(16, INTFLOAT, delay)[PS_MAX_SSB][PS_QMF_TIME_SLOTS + PS_MAX_DELAY][2];
+    DECLARE_ALIGNED(16, INTFLOAT, ap_delay)[PS_MAX_AP_BANDS][PS_AP_LINKS][PS_QMF_TIME_SLOTS + PS_MAX_AP_DELAY][2];
+    DECLARE_ALIGNED(16, INTFLOAT, peak_decay_nrg)[34];
+    DECLARE_ALIGNED(16, INTFLOAT, power_smooth)[34];
+    DECLARE_ALIGNED(16, INTFLOAT, peak_decay_diff_smooth)[34];
+    DECLARE_ALIGNED(16, INTFLOAT, H11)[2][PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC];
+    DECLARE_ALIGNED(16, INTFLOAT, H12)[2][PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC];
+    DECLARE_ALIGNED(16, INTFLOAT, H21)[2][PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC];
+    DECLARE_ALIGNED(16, INTFLOAT, H22)[2][PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC];
     int8_t opd_hist[PS_MAX_NR_IIDICC];
     int8_t ipd_hist[PS_MAX_NR_IIDICC];
     PSDSPContext dsp;
 } PSContext;
 
-void ff_ps_init(void);
-void ff_ps_ctx_init(PSContext *ps);
-int ff_ps_read_data(AVCodecContext *avctx, GetBitContext *gb, PSContext *ps, int bits_left);
-int ff_ps_apply(AVCodecContext *avctx, PSContext *ps, float L[2][38][64], float R[2][38][64], int top);
+void AAC_RENAME(ff_ps_init)(void);
+void AAC_RENAME(ff_ps_ctx_init)(PSContext *ps);
+int AAC_RENAME(ff_ps_read_data)(AVCodecContext *avctx, GetBitContext *gb, PSContext *ps, int bits_left);
+int AAC_RENAME(ff_ps_apply)(AVCodecContext *avctx, PSContext *ps, INTFLOAT L[2][38][64], INTFLOAT R[2][38][64], int top);
 
 #endif /* AVCODEC_PS_H */
diff --git a/libavcodec/aacps_fixed.c b/libavcodec/aacps_fixed.c
new file mode 100644
index 0000000..46af213
--- /dev/null
+++ b/libavcodec/aacps_fixed.c
@@ -0,0 +1,24 @@
+/*
+ * MPEG-4 Parametric Stereo decoding functions
+ * Copyright (c) 2010 Alex Converse <alex.converse at gmail.com>
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#define USE_FIXED 1
+
+#include "aacps.c"
diff --git a/libavcodec/aacps_float.c b/libavcodec/aacps_float.c
index 93a498c..73259c1 100644
--- a/libavcodec/aacps_float.c
+++ b/libavcodec/aacps_float.c
@@ -19,159 +19,6 @@
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  */
 
-static void stereo_processing(PSContext *ps, float (*l)[32][2], float (*r)[32][2], int is34)
-{
-    int e, b, k;
+#define USE_FIXED 0
 
-    float (*H11)[PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC] = ps->H11;
-    float (*H12)[PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC] = ps->H12;
-    float (*H21)[PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC] = ps->H21;
-    float (*H22)[PS_MAX_NUM_ENV+1][PS_MAX_NR_IIDICC] = ps->H22;
-    int8_t *opd_hist = ps->opd_hist;
-    int8_t *ipd_hist = ps->ipd_hist;
-    int8_t iid_mapped_buf[PS_MAX_NUM_ENV][PS_MAX_NR_IIDICC];
-    int8_t icc_mapped_buf[PS_MAX_NUM_ENV][PS_MAX_NR_IIDICC];
-    int8_t ipd_mapped_buf[PS_MAX_NUM_ENV][PS_MAX_NR_IIDICC];
-    int8_t opd_mapped_buf[PS_MAX_NUM_ENV][PS_MAX_NR_IIDICC];
-    int8_t (*iid_mapped)[PS_MAX_NR_IIDICC] = iid_mapped_buf;
-    int8_t (*icc_mapped)[PS_MAX_NR_IIDICC] = icc_mapped_buf;
-    int8_t (*ipd_mapped)[PS_MAX_NR_IIDICC] = ipd_mapped_buf;
-    int8_t (*opd_mapped)[PS_MAX_NR_IIDICC] = opd_mapped_buf;
-    const int8_t *k_to_i = is34 ? k_to_i_34 : k_to_i_20;
-    TABLE_CONST float (*H_LUT)[8][4] = (PS_BASELINE || ps->icc_mode < 3) ? HA : HB;
-
-    //Remapping
-    if (ps->num_env_old) {
-        memcpy(H11[0][0], H11[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H11[0][0][0]));
-        memcpy(H11[1][0], H11[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H11[1][0][0]));
-        memcpy(H12[0][0], H12[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H12[0][0][0]));
-        memcpy(H12[1][0], H12[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H12[1][0][0]));
-        memcpy(H21[0][0], H21[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H21[0][0][0]));
-        memcpy(H21[1][0], H21[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H21[1][0][0]));
-        memcpy(H22[0][0], H22[0][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H22[0][0][0]));
-        memcpy(H22[1][0], H22[1][ps->num_env_old], PS_MAX_NR_IIDICC*sizeof(H22[1][0][0]));
-    }
-
-    if (is34) {
-        remap34(&iid_mapped, ps->iid_par, ps->nr_iid_par, ps->num_env, 1);
-        remap34(&icc_mapped, ps->icc_par, ps->nr_icc_par, ps->num_env, 1);
-        if (ps->enable_ipdopd) {
-            remap34(&ipd_mapped, ps->ipd_par, ps->nr_ipdopd_par, ps->num_env, 0);
-            remap34(&opd_mapped, ps->opd_par, ps->nr_ipdopd_par, ps->num_env, 0);
-        }
-        if (!ps->is34bands_old) {
-            map_val_20_to_34(H11[0][0]);
-            map_val_20_to_34(H11[1][0]);
-            map_val_20_to_34(H12[0][0]);
-            map_val_20_to_34(H12[1][0]);
-            map_val_20_to_34(H21[0][0]);
-            map_val_20_to_34(H21[1][0]);
-            map_val_20_to_34(H22[0][0]);
-            map_val_20_to_34(H22[1][0]);
-            ipdopd_reset(ipd_hist, opd_hist);
-        }
-    } else {
-        remap20(&iid_mapped, ps->iid_par, ps->nr_iid_par, ps->num_env, 1);
-        remap20(&icc_mapped, ps->icc_par, ps->nr_icc_par, ps->num_env, 1);
-        if (ps->enable_ipdopd) {
-            remap20(&ipd_mapped, ps->ipd_par, ps->nr_ipdopd_par, ps->num_env, 0);
-            remap20(&opd_mapped, ps->opd_par, ps->nr_ipdopd_par, ps->num_env, 0);
-        }
-        if (ps->is34bands_old) {
-            map_val_34_to_20(H11[0][0]);
-            map_val_34_to_20(H11[1][0]);
-            map_val_34_to_20(H12[0][0]);
-            map_val_34_to_20(H12[1][0]);
-            map_val_34_to_20(H21[0][0]);
-            map_val_34_to_20(H21[1][0]);
-            map_val_34_to_20(H22[0][0]);
-            map_val_34_to_20(H22[1][0]);
-            ipdopd_reset(ipd_hist, opd_hist);
-        }
-    }
-
-    //Mixing
-    for (e = 0; e < ps->num_env; e++) {
-        for (b = 0; b < NR_PAR_BANDS[is34]; b++) {
-            float h11, h12, h21, h22;
-            h11 = H_LUT[iid_mapped[e][b] + 7 + 23 * ps->iid_quant][icc_mapped[e][b]][0];
-            h12 = H_LUT[iid_mapped[e][b] + 7 + 23 * ps->iid_quant][icc_mapped[e][b]][1];
-            h21 = H_LUT[iid_mapped[e][b] + 7 + 23 * ps->iid_quant][icc_mapped[e][b]][2];
-            h22 = H_LUT[iid_mapped[e][b] + 7 + 23 * ps->iid_quant][icc_mapped[e][b]][3];
-
-            if (!PS_BASELINE && ps->enable_ipdopd && b < NR_IPDOPD_BANDS[is34]) {
-                //The spec say says to only run this smoother when enable_ipdopd
-                //is set but the reference decoder appears to run it constantly
-                float h11i, h12i, h21i, h22i;
-                float ipd_adj_re, ipd_adj_im;
-                int opd_idx = opd_hist[b] * 8 + opd_mapped[e][b];
-                int ipd_idx = ipd_hist[b] * 8 + ipd_mapped[e][b];
-                float opd_re = pd_re_smooth[opd_idx];
-                float opd_im = pd_im_smooth[opd_idx];
-                float ipd_re = pd_re_smooth[ipd_idx];
-                float ipd_im = pd_im_smooth[ipd_idx];
-                opd_hist[b] = opd_idx & 0x3F;
-                ipd_hist[b] = ipd_idx & 0x3F;
-
-                ipd_adj_re = opd_re*ipd_re + opd_im*ipd_im;
-                ipd_adj_im = opd_im*ipd_re - opd_re*ipd_im;
-                h11i = h11 * opd_im;
-                h11  = h11 * opd_re;
-                h12i = h12 * ipd_adj_im;
-                h12  = h12 * ipd_adj_re;
-                h21i = h21 * opd_im;
-                h21  = h21 * opd_re;
-                h22i = h22 * ipd_adj_im;
-                h22  = h22 * ipd_adj_re;
-                H11[1][e+1][b] = h11i;
-                H12[1][e+1][b] = h12i;
-                H21[1][e+1][b] = h21i;
-                H22[1][e+1][b] = h22i;
-            }
-            H11[0][e+1][b] = h11;
-            H12[0][e+1][b] = h12;
-            H21[0][e+1][b] = h21;
-            H22[0][e+1][b] = h22;
-        }
-        for (k = 0; k < NR_BANDS[is34]; k++) {
-            float h[2][4];
-            float h_step[2][4];
-            int start = ps->border_position[e];
-            int stop  = ps->border_position[e+1];
-            float width = 1.f / (stop - start);
-            b = k_to_i[k];
-            h[0][0] = H11[0][e][b];
-            h[0][1] = H12[0][e][b];
-            h[0][2] = H21[0][e][b];
-            h[0][3] = H22[0][e][b];
-            if (!PS_BASELINE && ps->enable_ipdopd) {
-            //Is this necessary? ps_04_new seems unchanged
-            if ((is34 && k <= 13 && k >= 9) || (!is34 && k <= 1)) {
-                h[1][0] = -H11[1][e][b];
-                h[1][1] = -H12[1][e][b];
-                h[1][2] = -H21[1][e][b];
-                h[1][3] = -H22[1][e][b];
-            } else {
-                h[1][0] = H11[1][e][b];
-                h[1][1] = H12[1][e][b];
-                h[1][2] = H21[1][e][b];
-                h[1][3] = H22[1][e][b];
-            }
-            }
-            //Interpolation
-            h_step[0][0] = (H11[0][e+1][b] - h[0][0]) * width;
-            h_step[0][1] = (H12[0][e+1][b] - h[0][1]) * width;
-            h_step[0][2] = (H21[0][e+1][b] - h[0][2]) * width;
-            h_step[0][3] = (H22[0][e+1][b] - h[0][3]) * width;
-            if (!PS_BASELINE && ps->enable_ipdopd) {
-                h_step[1][0] = (H11[1][e+1][b] - h[1][0]) * width;
-                h_step[1][1] = (H12[1][e+1][b] - h[1][1]) * width;
-                h_step[1][2] = (H21[1][e+1][b] - h[1][2]) * width;
-                h_step[1][3] = (H22[1][e+1][b] - h[1][3]) * width;
-            }
-            ps->dsp.stereo_interpolate[!PS_BASELINE && ps->enable_ipdopd](
-                l[k] + start + 1, r[k] + start + 1,
-                h, h_step, stop - start);
-        }
-    }
-}
+#include "aacps.c"
diff --git a/libavcodec/aacpsdata.c b/libavcodec/aacpsdata.c
index 7431cae..5c1a1b0 100644
--- a/libavcodec/aacpsdata.c
+++ b/libavcodec/aacpsdata.c
@@ -157,7 +157,7 @@ static const int8_t k_to_i_34[] = {
     33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33, 33
 };
 
-static const float g1_Q2[] = {
-    0.0f,  0.01899487526049f, 0.0f, -0.07293139167538f,
-    0.0f,  0.30596630545168f, 0.5f
+static const INTFLOAT g1_Q2[] = {
+    Q31(0.0f),  Q31(0.01899487526049f), Q31(0.0f), Q31(-0.07293139167538f),
+    Q31(0.0f),  Q31(0.30596630545168f), Q31(0.5f)
 };
diff --git a/libavcodec/aacpsdsp.c b/libavcodec/aacpsdsp.c
deleted file mode 100644
index 5dc1a6a..0000000
--- a/libavcodec/aacpsdsp.c
+++ /dev/null
@@ -1,216 +0,0 @@
-/*
- * Copyright (c) 2010 Alex Converse <alex.converse at gmail.com>
- *
- * This file is part of FFmpeg.
- *
- * FFmpeg is free software; you can redistribute it and/or
- * modify it under the terms of the GNU Lesser General Public
- * License as published by the Free Software Foundation; either
- * version 2.1 of the License, or (at your option) any later version.
- *
- * FFmpeg is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
- * Lesser General Public License for more details.
- *
- * You should have received a copy of the GNU Lesser General Public
- * License along with FFmpeg; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
- */
-
-#include "config.h"
-#include "libavutil/attributes.h"
-#include "aacpsdsp.h"
-
-static void ps_add_squares_c(float *dst, const float (*src)[2], int n)
-{
-    int i;
-    for (i = 0; i < n; i++)
-        dst[i] += src[i][0] * src[i][0] + src[i][1] * src[i][1];
-}
-
-static void ps_mul_pair_single_c(float (*dst)[2], float (*src0)[2], float *src1,
-                                 int n)
-{
-    int i;
-    for (i = 0; i < n; i++) {
-        dst[i][0] = src0[i][0] * src1[i];
-        dst[i][1] = src0[i][1] * src1[i];
-    }
-}
-
-static void ps_hybrid_analysis_c(float (*out)[2], float (*in)[2],
-                                 const float (*filter)[8][2],
-                                 int stride, int n)
-{
-    int i, j;
-
-    for (i = 0; i < n; i++) {
-        float sum_re = filter[i][6][0] * in[6][0];
-        float sum_im = filter[i][6][0] * in[6][1];
-
-        for (j = 0; j < 6; j++) {
-            float in0_re = in[j][0];
-            float in0_im = in[j][1];
-            float in1_re = in[12-j][0];
-            float in1_im = in[12-j][1];
-            sum_re += filter[i][j][0] * (in0_re + in1_re) -
-                      filter[i][j][1] * (in0_im - in1_im);
-            sum_im += filter[i][j][0] * (in0_im + in1_im) +
-                      filter[i][j][1] * (in0_re - in1_re);
-        }
-        out[i * stride][0] = sum_re;
-        out[i * stride][1] = sum_im;
-    }
-}
-
-static void ps_hybrid_analysis_ileave_c(float (*out)[32][2], float L[2][38][64],
-                                        int i, int len)
-{
-    int j;
-
-    for (; i < 64; i++) {
-        for (j = 0; j < len; j++) {
-            out[i][j][0] = L[0][j][i];
-            out[i][j][1] = L[1][j][i];
-        }
-    }
-}
-
-static void ps_hybrid_synthesis_deint_c(float out[2][38][64],
-                                        float (*in)[32][2],
-                                        int i, int len)
-{
-    int n;
-
-    for (; i < 64; i++) {
-        for (n = 0; n < len; n++) {
-            out[0][n][i] = in[i][n][0];
-            out[1][n][i] = in[i][n][1];
-        }
-    }
-}
-
-static void ps_decorrelate_c(float (*out)[2], float (*delay)[2],
-                             float (*ap_delay)[PS_QMF_TIME_SLOTS + PS_MAX_AP_DELAY][2],
-                             const float phi_fract[2], const float (*Q_fract)[2],
-                             const float *transient_gain,
-                             float g_decay_slope,
-                             int len)
-{
-    static const float a[] = { 0.65143905753106f,
-                               0.56471812200776f,
-                               0.48954165955695f };
-    float ag[PS_AP_LINKS];
-    int m, n;
-
-    for (m = 0; m < PS_AP_LINKS; m++)
-        ag[m] = a[m] * g_decay_slope;
-
-    for (n = 0; n < len; n++) {
-        float in_re = delay[n][0] * phi_fract[0] - delay[n][1] * phi_fract[1];
-        float in_im = delay[n][0] * phi_fract[1] + delay[n][1] * phi_fract[0];
-        for (m = 0; m < PS_AP_LINKS; m++) {
-            float a_re                = ag[m] * in_re;
-            float a_im                = ag[m] * in_im;
-            float link_delay_re       = ap_delay[m][n+2-m][0];
-            float link_delay_im       = ap_delay[m][n+2-m][1];
-            float fractional_delay_re = Q_fract[m][0];
-            float fractional_delay_im = Q_fract[m][1];
-            float apd_re = in_re;
-            float apd_im = in_im;
-            in_re = link_delay_re * fractional_delay_re -
-                    link_delay_im * fractional_delay_im - a_re;
-            in_im = link_delay_re * fractional_delay_im +
-                    link_delay_im * fractional_delay_re - a_im;
-            ap_delay[m][n+5][0] = apd_re + ag[m] * in_re;
-            ap_delay[m][n+5][1] = apd_im + ag[m] * in_im;
-        }
-        out[n][0] = transient_gain[n] * in_re;
-        out[n][1] = transient_gain[n] * in_im;
-    }
-}
-
-static void ps_stereo_interpolate_c(float (*l)[2], float (*r)[2],
-                                    float h[2][4], float h_step[2][4],
-                                    int len)
-{
-    float h0 = h[0][0];
-    float h1 = h[0][1];
-    float h2 = h[0][2];
-    float h3 = h[0][3];
-    float hs0 = h_step[0][0];
-    float hs1 = h_step[0][1];
-    float hs2 = h_step[0][2];
-    float hs3 = h_step[0][3];
-    int n;
-
-    for (n = 0; n < len; n++) {
-        //l is s, r is d
-        float l_re = l[n][0];
-        float l_im = l[n][1];
-        float r_re = r[n][0];
-        float r_im = r[n][1];
-        h0 += hs0;
-        h1 += hs1;
-        h2 += hs2;
-        h3 += hs3;
-        l[n][0] = h0 * l_re + h2 * r_re;
-        l[n][1] = h0 * l_im + h2 * r_im;
-        r[n][0] = h1 * l_re + h3 * r_re;
-        r[n][1] = h1 * l_im + h3 * r_im;
-    }
-}
-
-static void ps_stereo_interpolate_ipdopd_c(float (*l)[2], float (*r)[2],
-                                           float h[2][4], float h_step[2][4],
-                                           int len)
-{
-    float h00  = h[0][0],      h10  = h[1][0];
-    float h01  = h[0][1],      h11  = h[1][1];
-    float h02  = h[0][2],      h12  = h[1][2];
-    float h03  = h[0][3],      h13  = h[1][3];
-    float hs00 = h_step[0][0], hs10 = h_step[1][0];
-    float hs01 = h_step[0][1], hs11 = h_step[1][1];
-    float hs02 = h_step[0][2], hs12 = h_step[1][2];
-    float hs03 = h_step[0][3], hs13 = h_step[1][3];
-    int n;
-
-    for (n = 0; n < len; n++) {
-        //l is s, r is d
-        float l_re = l[n][0];
-        float l_im = l[n][1];
-        float r_re = r[n][0];
-        float r_im = r[n][1];
-        h00 += hs00;
-        h01 += hs01;
-        h02 += hs02;
-        h03 += hs03;
-        h10 += hs10;
-        h11 += hs11;
-        h12 += hs12;
-        h13 += hs13;
-
-        l[n][0] = h00 * l_re + h02 * r_re - h10 * l_im - h12 * r_im;
-        l[n][1] = h00 * l_im + h02 * r_im + h10 * l_re + h12 * r_re;
-        r[n][0] = h01 * l_re + h03 * r_re - h11 * l_im - h13 * r_im;
-        r[n][1] = h01 * l_im + h03 * r_im + h11 * l_re + h13 * r_re;
-    }
-}
-
-av_cold void ff_psdsp_init(PSDSPContext *s)
-{
-    s->add_squares            = ps_add_squares_c;
-    s->mul_pair_single        = ps_mul_pair_single_c;
-    s->hybrid_analysis        = ps_hybrid_analysis_c;
-    s->hybrid_analysis_ileave = ps_hybrid_analysis_ileave_c;
-    s->hybrid_synthesis_deint = ps_hybrid_synthesis_deint_c;
-    s->decorrelate            = ps_decorrelate_c;
-    s->stereo_interpolate[0]  = ps_stereo_interpolate_c;
-    s->stereo_interpolate[1]  = ps_stereo_interpolate_ipdopd_c;
-
-    if (ARCH_ARM)
-        ff_psdsp_init_arm(s);
-    if (ARCH_MIPS)
-        ff_psdsp_init_mips(s);
-}
diff --git a/libavcodec/aacpsdsp.h b/libavcodec/aacpsdsp.h
index 0ef3023..9e3c5aa 100644
--- a/libavcodec/aacpsdsp.h
+++ b/libavcodec/aacpsdsp.h
@@ -21,33 +21,35 @@
 #ifndef LIBAVCODEC_AACPSDSP_H
 #define LIBAVCODEC_AACPSDSP_H
 
+#include "aac_defines.h"
+
 #define PS_QMF_TIME_SLOTS 32
 #define PS_AP_LINKS 3
 #define PS_MAX_AP_DELAY 5
 
 typedef struct PSDSPContext {
-    void (*add_squares)(float *dst, const float (*src)[2], int n);
-    void (*mul_pair_single)(float (*dst)[2], float (*src0)[2], float *src1,
+    void (*add_squares)(INTFLOAT *dst, const INTFLOAT (*src)[2], int n);
+    void (*mul_pair_single)(INTFLOAT (*dst)[2], INTFLOAT (*src0)[2], INTFLOAT *src1,
                             int n);
-    void (*hybrid_analysis)(float (*out)[2], float (*in)[2],
-                            const float (*filter)[8][2],
+    void (*hybrid_analysis)(INTFLOAT (*out)[2], INTFLOAT (*in)[2],
+                            const INTFLOAT (*filter)[8][2],
                             int stride, int n);
-    void (*hybrid_analysis_ileave)(float (*out)[32][2], float L[2][38][64],
+    void (*hybrid_analysis_ileave)(INTFLOAT (*out)[32][2], INTFLOAT L[2][38][64],
                                    int i, int len);
-    void (*hybrid_synthesis_deint)(float out[2][38][64], float (*in)[32][2],
+    void (*hybrid_synthesis_deint)(INTFLOAT out[2][38][64], INTFLOAT (*in)[32][2],
                                    int i, int len);
-    void (*decorrelate)(float (*out)[2], float (*delay)[2],
-                        float (*ap_delay)[PS_QMF_TIME_SLOTS+PS_MAX_AP_DELAY][2],
-                        const float phi_fract[2], const float (*Q_fract)[2],
-                        const float *transient_gain,
-                        float g_decay_slope,
+    void (*decorrelate)(INTFLOAT (*out)[2], INTFLOAT (*delay)[2],
+                        INTFLOAT (*ap_delay)[PS_QMF_TIME_SLOTS+PS_MAX_AP_DELAY][2],
+                        const INTFLOAT phi_fract[2], const INTFLOAT (*Q_fract)[2],
+                        const INTFLOAT *transient_gain,
+                        INTFLOAT g_decay_slope,
                         int len);
-    void (*stereo_interpolate[2])(float (*l)[2], float (*r)[2],
-                                  float h[2][4], float h_step[2][4],
+    void (*stereo_interpolate[2])(INTFLOAT (*l)[2], INTFLOAT (*r)[2],
+                                  INTFLOAT h[2][4], INTFLOAT h_step[2][4],
                                   int len);
 } PSDSPContext;
 
-void ff_psdsp_init(PSDSPContext *s);
+void AAC_RENAME(ff_psdsp_init)(PSDSPContext *s);
 void ff_psdsp_init_arm(PSDSPContext *s);
 void ff_psdsp_init_mips(PSDSPContext *s);
 
diff --git a/libavcodec/aacpsdsp_fixed.c b/libavcodec/aacpsdsp_fixed.c
new file mode 100644
index 0000000..2413295
--- /dev/null
+++ b/libavcodec/aacpsdsp_fixed.c
@@ -0,0 +1,23 @@
+/*
+ * Copyright (c) 2010 Alex Converse <alex.converse at gmail.com>
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#define USE_FIXED 1
+
+#include "aacpsdsp_template.c"
diff --git a/libavcodec/aacpsdsp_float.c b/libavcodec/aacpsdsp_float.c
new file mode 100644
index 0000000..99aa650
--- /dev/null
+++ b/libavcodec/aacpsdsp_float.c
@@ -0,0 +1,23 @@
+/*
+ * Copyright (c) 2010 Alex Converse <alex.converse at gmail.com>
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#define USE_FIXED 0
+
+#include "aacpsdsp_template.c"
diff --git a/libavcodec/aacpsdsp_template.c b/libavcodec/aacpsdsp_template.c
new file mode 100644
index 0000000..bfec828
--- /dev/null
+++ b/libavcodec/aacpsdsp_template.c
@@ -0,0 +1,228 @@
+/*
+ * Copyright (c) 2010 Alex Converse <alex.converse at gmail.com>
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ *
+ * Note: Rounding-to-nearest used unless otherwise stated
+ *
+ */
+#include <stdint.h>
+
+#include "config.h"
+#include "libavutil/attributes.h"
+#include "aacpsdsp.h"
+
+static void ps_add_squares_c(INTFLOAT *dst, const INTFLOAT (*src)[2], int n)
+{
+    int i;
+    for (i = 0; i < n; i++)
+        dst[i] += AAC_MADD28(src[i][0], src[i][0], src[i][1], src[i][1]);
+}
+
+static void ps_mul_pair_single_c(INTFLOAT (*dst)[2], INTFLOAT (*src0)[2], INTFLOAT *src1,
+                                 int n)
+{
+    int i;
+    for (i = 0; i < n; i++) {
+        dst[i][0] = AAC_MUL16(src0[i][0], src1[i]);
+        dst[i][1] = AAC_MUL16(src0[i][1], src1[i]);
+    }
+}
+
+static void ps_hybrid_analysis_c(INTFLOAT (*out)[2], INTFLOAT (*in)[2],
+                                 const INTFLOAT (*filter)[8][2],
+                                 int stride, int n)
+{
+    int i, j;
+
+    for (i = 0; i < n; i++) {
+        INT64FLOAT sum_re = (INT64FLOAT)filter[i][6][0] * in[6][0];
+        INT64FLOAT sum_im = (INT64FLOAT)filter[i][6][0] * in[6][1];
+
+        for (j = 0; j < 6; j++) {
+            INTFLOAT in0_re = in[j][0];
+            INTFLOAT in0_im = in[j][1];
+            INTFLOAT in1_re = in[12-j][0];
+            INTFLOAT in1_im = in[12-j][1];
+            sum_re += (INT64FLOAT)filter[i][j][0] * (in0_re + in1_re) -
+                      (INT64FLOAT)filter[i][j][1] * (in0_im - in1_im);
+            sum_im += (INT64FLOAT)filter[i][j][0] * (in0_im + in1_im) +
+                      (INT64FLOAT)filter[i][j][1] * (in0_re - in1_re);
+        }
+#if USE_FIXED
+        out[i * stride][0] = (int)((sum_re + 0x40000000) >> 31);
+        out[i * stride][1] = (int)((sum_im + 0x40000000) >> 31);
+#else
+        out[i * stride][0] = sum_re;
+        out[i * stride][1] = sum_im;
+#endif /* USE_FIXED */
+    }
+}
+static void ps_hybrid_analysis_ileave_c(INTFLOAT (*out)[32][2], INTFLOAT L[2][38][64],
+                                      int i, int len)
+{
+    int j;
+
+    for (; i < 64; i++) {
+        for (j = 0; j < len; j++) {
+            out[i][j][0] = L[0][j][i];
+            out[i][j][1] = L[1][j][i];
+        }
+    }
+}
+
+static void ps_hybrid_synthesis_deint_c(INTFLOAT out[2][38][64],
+                                      INTFLOAT (*in)[32][2],
+                                      int i, int len)
+{
+    int n;
+
+    for (; i < 64; i++) {
+        for (n = 0; n < len; n++) {
+            out[0][n][i] = in[i][n][0];
+            out[1][n][i] = in[i][n][1];
+        }
+    }
+}
+
+static void ps_decorrelate_c(INTFLOAT (*out)[2], INTFLOAT (*delay)[2],
+                             INTFLOAT (*ap_delay)[PS_QMF_TIME_SLOTS + PS_MAX_AP_DELAY][2],
+                             const INTFLOAT phi_fract[2], const INTFLOAT (*Q_fract)[2],
+                             const INTFLOAT *transient_gain,
+                             INTFLOAT g_decay_slope,
+                             int len)
+{
+    static const INTFLOAT a[] = { Q31(0.65143905753106f),
+                               Q31(0.56471812200776f),
+                               Q31(0.48954165955695f) };
+    INTFLOAT ag[PS_AP_LINKS];
+    int m, n;
+
+    for (m = 0; m < PS_AP_LINKS; m++)
+        ag[m] = AAC_MUL30(a[m], g_decay_slope);
+
+    for (n = 0; n < len; n++) {
+        INTFLOAT in_re = AAC_MSUB30(delay[n][0], phi_fract[0], delay[n][1], phi_fract[1]);
+        INTFLOAT in_im = AAC_MADD30(delay[n][0], phi_fract[1], delay[n][1], phi_fract[0]);
+        for (m = 0; m < PS_AP_LINKS; m++) {
+            INTFLOAT a_re                = AAC_MUL31(ag[m], in_re);
+            INTFLOAT a_im                = AAC_MUL31(ag[m], in_im);
+            INTFLOAT link_delay_re       = ap_delay[m][n+2-m][0];
+            INTFLOAT link_delay_im       = ap_delay[m][n+2-m][1];
+            INTFLOAT fractional_delay_re = Q_fract[m][0];
+            INTFLOAT fractional_delay_im = Q_fract[m][1];
+            INTFLOAT apd_re = in_re;
+            INTFLOAT apd_im = in_im;
+            in_re = AAC_MSUB30(link_delay_re, fractional_delay_re,
+                    link_delay_im, fractional_delay_im);
+            in_re -= a_re;
+            in_im = AAC_MADD30(link_delay_re, fractional_delay_im,
+                    link_delay_im, fractional_delay_re);
+            in_im -= a_im;
+            ap_delay[m][n+5][0] = apd_re + AAC_MUL31(ag[m], in_re);
+            ap_delay[m][n+5][1] = apd_im + AAC_MUL31(ag[m], in_im);
+        }
+        out[n][0] = AAC_MUL16(transient_gain[n], in_re);
+        out[n][1] = AAC_MUL16(transient_gain[n], in_im);
+    }
+}
+
+static void ps_stereo_interpolate_c(INTFLOAT (*l)[2], INTFLOAT (*r)[2],
+                                    INTFLOAT h[2][4], INTFLOAT h_step[2][4],
+                                    int len)
+{
+    INTFLOAT h0 = h[0][0];
+    INTFLOAT h1 = h[0][1];
+    INTFLOAT h2 = h[0][2];
+    INTFLOAT h3 = h[0][3];
+    INTFLOAT hs0 = h_step[0][0];
+    INTFLOAT hs1 = h_step[0][1];
+    INTFLOAT hs2 = h_step[0][2];
+    INTFLOAT hs3 = h_step[0][3];
+    int n;
+
+    for (n = 0; n < len; n++) {
+        //l is s, r is d
+        INTFLOAT l_re = l[n][0];
+        INTFLOAT l_im = l[n][1];
+        INTFLOAT r_re = r[n][0];
+        INTFLOAT r_im = r[n][1];
+        h0 += hs0;
+        h1 += hs1;
+        h2 += hs2;
+        h3 += hs3;
+        l[n][0] = AAC_MADD30(h0,  l_re,  h2, r_re);
+        l[n][1] = AAC_MADD30(h0,  l_im,  h2,  r_im);
+        r[n][0] = AAC_MADD30(h1,  l_re,  h3,  r_re);
+        r[n][1] = AAC_MADD30(h1,  l_im,  h3,  r_im);
+    }
+}
+
+static void ps_stereo_interpolate_ipdopd_c(INTFLOAT (*l)[2], INTFLOAT (*r)[2],
+                                           INTFLOAT h[2][4], INTFLOAT h_step[2][4],
+                                           int len)
+{
+    INTFLOAT h00  = h[0][0],      h10  = h[1][0];
+    INTFLOAT h01  = h[0][1],      h11  = h[1][1];
+    INTFLOAT h02  = h[0][2],      h12  = h[1][2];
+    INTFLOAT h03  = h[0][3],      h13  = h[1][3];
+    INTFLOAT hs00 = h_step[0][0], hs10 = h_step[1][0];
+    INTFLOAT hs01 = h_step[0][1], hs11 = h_step[1][1];
+    INTFLOAT hs02 = h_step[0][2], hs12 = h_step[1][2];
+    INTFLOAT hs03 = h_step[0][3], hs13 = h_step[1][3];
+    int n;
+
+    for (n = 0; n < len; n++) {
+        //l is s, r is d
+        INTFLOAT l_re = l[n][0];
+        INTFLOAT l_im = l[n][1];
+        INTFLOAT r_re = r[n][0];
+        INTFLOAT r_im = r[n][1];
+        h00 += hs00;
+        h01 += hs01;
+        h02 += hs02;
+        h03 += hs03;
+        h10 += hs10;
+        h11 += hs11;
+        h12 += hs12;
+        h13 += hs13;
+
+        l[n][0] = AAC_MSUB30_V8(h00, l_re, h02, r_re, h10, l_im, h12, r_im);
+        l[n][1] = AAC_MADD30_V8(h00, l_im, h02, r_im, h10, l_re, h12, r_re);
+        r[n][0] = AAC_MSUB30_V8(h01, l_re, h03, r_re, h11, l_im, h13, r_im);
+        r[n][1] = AAC_MADD30_V8(h01, l_im, h03, r_im, h11, l_re, h13, r_re);
+    }
+}
+
+av_cold void AAC_RENAME(ff_psdsp_init)(PSDSPContext *s)
+{
+    s->add_squares            = ps_add_squares_c;
+    s->mul_pair_single        = ps_mul_pair_single_c;
+    s->hybrid_analysis        = ps_hybrid_analysis_c;
+    s->hybrid_analysis_ileave = ps_hybrid_analysis_ileave_c;
+    s->hybrid_synthesis_deint = ps_hybrid_synthesis_deint_c;
+    s->decorrelate            = ps_decorrelate_c;
+    s->stereo_interpolate[0]  = ps_stereo_interpolate_c;
+    s->stereo_interpolate[1]  = ps_stereo_interpolate_ipdopd_c;
+
+#if !USE_FIXED
+    if (ARCH_ARM)
+        ff_psdsp_init_arm(s);
+    if (ARCH_MIPS)
+        ff_psdsp_init_mips(s);
+#endif /* !USE_FIXED */
+}
-- 
1.8.2.1



More information about the ffmpeg-devel mailing list