changeset 10122:fe081db7c09c libavcodec

add missing wmapro decoder parts
author faust3
date Wed, 02 Sep 2009 16:41:49 +0000
parents 3ae2df444cdb
children e1bd75a1eab2
files Makefile allcodecs.c avcodec.h wmaprodec.c
diffstat 4 files changed, 1277 insertions(+), 9 deletions(-) [+]
line wrap: on
line diff
--- a/Makefile	Sun Aug 30 21:19:37 2009 +0000
+++ b/Makefile	Wed Sep 02 16:41:49 2009 +0000
@@ -249,6 +249,7 @@
 OBJS-$(CONFIG_VP6_DECODER)             += vp6.o vp56.o vp56data.o vp3dsp.o vp6dsp.o huffman.o
 OBJS-$(CONFIG_VQA_DECODER)             += vqavideo.o
 OBJS-$(CONFIG_WAVPACK_DECODER)         += wavpack.o
+OBJS-$(CONFIG_WMAPRO_DECODER)          += wmaprodec.o wma.o
 OBJS-$(CONFIG_WMAV1_DECODER)           += wmadec.o wma.o
 OBJS-$(CONFIG_WMAV1_ENCODER)           += wmaenc.o wma.o
 OBJS-$(CONFIG_WMAV2_DECODER)           += wmadec.o wma.o
--- a/allcodecs.c	Sun Aug 30 21:19:37 2009 +0000
+++ b/allcodecs.c	Wed Sep 02 16:41:49 2009 +0000
@@ -232,6 +232,7 @@
     REGISTER_DECODER (VMDAUDIO, vmdaudio);
     REGISTER_ENCDEC  (VORBIS, vorbis);
     REGISTER_DECODER (WAVPACK, wavpack);
+    REGISTER_DECODER (WMAPRO, wmapro);
     REGISTER_ENCDEC  (WMAV1, wmav1);
     REGISTER_ENCDEC  (WMAV2, wmav2);
     REGISTER_DECODER (WS_SND1, ws_snd1);
--- a/avcodec.h	Sun Aug 30 21:19:37 2009 +0000
+++ b/avcodec.h	Wed Sep 02 16:41:49 2009 +0000
@@ -30,7 +30,7 @@
 #include "libavutil/avutil.h"
 
 #define LIBAVCODEC_VERSION_MAJOR 52
-#define LIBAVCODEC_VERSION_MINOR 34
+#define LIBAVCODEC_VERSION_MINOR 35
 #define LIBAVCODEC_VERSION_MICRO  0
 
 #define LIBAVCODEC_VERSION_INT  AV_VERSION_INT(LIBAVCODEC_VERSION_MAJOR, \
--- a/wmaprodec.c	Sun Aug 30 21:19:37 2009 +0000
+++ b/wmaprodec.c	Wed Sep 02 16:41:49 2009 +0000
@@ -86,6 +86,158 @@
  * subframe in order to reconstruct the output samples.
  */
 
+#include "avcodec.h"
+#include "internal.h"
+#include "get_bits.h"
+#include "put_bits.h"
+#include "wmaprodata.h"
+#include "dsputil.h"
+#include "wma.h"
+
+/** current decoder limitations */
+#define WMAPRO_MAX_CHANNELS    8                             ///< max number of handled channels
+#define MAX_SUBFRAMES  32                                    ///< max number of subframes per channel
+#define MAX_BANDS      29                                    ///< max number of scale factor bands
+#define MAX_FRAMESIZE  16384                                 ///< maximum compressed frame size
+
+#define WMAPRO_BLOCK_MAX_BITS 12                                           ///< log2 of max block size
+#define WMAPRO_BLOCK_MAX_SIZE (1 << WMAPRO_BLOCK_MAX_BITS)                 ///< maximum block size
+#define WMAPRO_BLOCK_SIZES    (WMAPRO_BLOCK_MAX_BITS - BLOCK_MIN_BITS + 1) ///< possible block sizes
+
+
+#define VLCBITS            9
+#define SCALEVLCBITS       8
+#define VEC4MAXDEPTH    ((HUFF_VEC4_MAXBITS+VLCBITS-1)/VLCBITS)
+#define VEC2MAXDEPTH    ((HUFF_VEC2_MAXBITS+VLCBITS-1)/VLCBITS)
+#define VEC1MAXDEPTH    ((HUFF_VEC1_MAXBITS+VLCBITS-1)/VLCBITS)
+#define SCALEMAXDEPTH   ((HUFF_SCALE_MAXBITS+SCALEVLCBITS-1)/SCALEVLCBITS)
+#define SCALERLMAXDEPTH ((HUFF_SCALE_RL_MAXBITS+VLCBITS-1)/VLCBITS)
+
+static VLC              sf_vlc;           ///< scale factor DPCM vlc
+static VLC              sf_rl_vlc;        ///< scale factor run length vlc
+static VLC              vec4_vlc;         ///< 4 coefficients per symbol
+static VLC              vec2_vlc;         ///< 2 coefficients per symbol
+static VLC              vec1_vlc;         ///< 1 coefficient per symbol
+static VLC              coef_vlc[2];      ///< coefficient run length vlc codes
+static float            sin64[33];        ///< sinus table for decorrelation
+
+/**
+ * @brief frame specific decoder context for a single channel
+ */
+typedef struct {
+    int16_t  prev_block_len;                          ///< length of the previous block
+    uint8_t  transmit_coefs;
+    uint8_t  num_subframes;
+    uint16_t subframe_len[MAX_SUBFRAMES];             ///< subframe length in samples
+    uint16_t subframe_offset[MAX_SUBFRAMES];          ///< subframe positions in the current frame
+    uint8_t  cur_subframe;                            ///< current subframe number
+    uint16_t decoded_samples;                         ///< number of already processed samples
+    uint8_t  grouped;                                 ///< channel is part of a group
+    int      quant_step;                              ///< quantization step for the current subframe
+    int8_t   reuse_sf;                                ///< share scale factors between subframes
+    int8_t   scale_factor_step;                       ///< scaling step for the current subframe
+    int      max_scale_factor;                        ///< maximum scale factor for the current subframe
+    int      scale_factors[MAX_BANDS];                ///< scale factor values for the current subframe
+    int      saved_scale_factors[MAX_BANDS];          ///< scale factors from a previous subframe
+    uint8_t  table_idx;                               ///< index in sf_offsets for the scale factor reference block
+    float*   coeffs;                                  ///< pointer to the subframe decode buffer
+    DECLARE_ALIGNED_16(float, out[WMAPRO_BLOCK_MAX_SIZE + WMAPRO_BLOCK_MAX_SIZE / 2]); ///< output buffer
+} WMAProChannelCtx;
+
+/**
+ * @brief channel group for channel transformations
+ */
+typedef struct {
+    uint8_t num_channels;                                     ///< number of channels in the group
+    int8_t  transform;                                        ///< transform on / off
+    int8_t  transform_band[MAX_BANDS];                        ///< controls if the transform is enabled for a certain band
+    float   decorrelation_matrix[WMAPRO_MAX_CHANNELS*WMAPRO_MAX_CHANNELS];
+    float*  channel_data[WMAPRO_MAX_CHANNELS];                ///< transformation coefficients
+} WMAProChannelGrp;
+
+/**
+ * @brief main decoder context
+ */
+typedef struct WMAProDecodeCtx {
+    /* generic decoder variables */
+    AVCodecContext*  avctx;                         ///< codec context for av_log
+    DSPContext       dsp;                           ///< accelerated DSP functions
+    uint8_t          frame_data[MAX_FRAMESIZE +
+                      FF_INPUT_BUFFER_PADDING_SIZE];///< compressed frame data
+    PutBitContext    pb;                            ///< context for filling the frame_data buffer
+    MDCTContext      mdct_ctx[WMAPRO_BLOCK_SIZES];  ///< MDCT context per block size
+    DECLARE_ALIGNED_16(float, tmp[WMAPRO_BLOCK_MAX_SIZE]); ///< IMDCT output buffer
+    float*           windows[WMAPRO_BLOCK_SIZES];   ///< windows for the different block sizes
+
+    /* frame size dependent frame information (set during initialization) */
+    uint32_t         decode_flags;                  ///< used compression features
+    uint8_t          len_prefix;                    ///< frame is prefixed with its length
+    uint8_t          dynamic_range_compression;     ///< frame contains DRC data
+    uint8_t          bits_per_sample;               ///< integer audio sample size for the unscaled IMDCT output (used to scale to [-1.0, 1.0])
+    uint16_t         samples_per_frame;             ///< number of samples to output
+    uint16_t         log2_frame_size;
+    int8_t           num_channels;                  ///< number of channels in the stream (same as AVCodecContext.num_channels)
+    int8_t           lfe_channel;                   ///< lfe channel index
+    uint8_t          max_num_subframes;
+    uint8_t          subframe_len_bits;             ///< number of bits used for the subframe length
+    uint8_t          max_subframe_len_bit;          ///< flag indicating that the subframe is of maximum size when the first subframe length bit is 1
+    uint16_t         min_samples_per_subframe;
+    int8_t           num_sfb[WMAPRO_BLOCK_SIZES];   ///< scale factor bands per block size
+    int16_t          sfb_offsets[WMAPRO_BLOCK_SIZES][MAX_BANDS];                    ///< scale factor band offsets (multiples of 4)
+    int8_t           sf_offsets[WMAPRO_BLOCK_SIZES][WMAPRO_BLOCK_SIZES][MAX_BANDS]; ///< scale factor resample matrix
+    int16_t          subwoofer_cutoffs[WMAPRO_BLOCK_SIZES]; ///< subwoofer cutoff values
+
+    /* packet decode state */
+    uint8_t          packet_sequence_number;        ///< current packet number
+    int              num_saved_bits;                ///< saved number of bits
+    int              frame_offset;                  ///< frame offset in the bit reservoir
+    int              subframe_offset;               ///< subframe offset in the bit reservoir
+    uint8_t          packet_loss;                   ///< set in case of bitstream error
+
+    /* frame decode state */
+    uint32_t         frame_num;                     ///< current frame number (not used for decoding)
+    GetBitContext    gb;                            ///< bitstream reader context
+    int              buf_bit_size;                  ///< buffer size in bits
+    float*           samples;                       ///< current samplebuffer pointer
+    float*           samples_end;                   ///< maximum samplebuffer pointer
+    uint8_t          drc_gain;                      ///< gain for the DRC tool
+    int8_t           skip_frame;                    ///< skip output step
+    int8_t           parsed_all_subframes;          ///< all subframes decoded?
+
+    /* subframe/block decode state */
+    int16_t          subframe_len;                  ///< current subframe length
+    int8_t           channels_for_cur_subframe;     ///< number of channels that contain the subframe
+    int8_t           channel_indexes_for_cur_subframe[WMAPRO_MAX_CHANNELS];
+    int8_t           num_bands;                     ///< number of scale factor bands
+    int16_t*         cur_sfb_offsets;               ///< sfb offsets for the current block
+    uint8_t          table_idx;                     ///< index for the num_sfb, sfb_offsets, sf_offsets and subwoofer_cutoffs tables
+    int8_t           esc_len;                       ///< length of escaped coefficients
+
+    uint8_t          num_chgroups;                  ///< number of channel groups
+    WMAProChannelGrp chgroup[WMAPRO_MAX_CHANNELS];  ///< channel group information
+
+    WMAProChannelCtx channel[WMAPRO_MAX_CHANNELS];  ///< per channel data
+} WMAProDecodeCtx;
+
+
+/**
+ *@brief helper function to print the most important members of the context
+ *@param s context
+ */
+static void av_cold dump_context(WMAProDecodeCtx *s)
+{
+#define PRINT(a, b)     av_log(s->avctx, AV_LOG_DEBUG, " %s = %d\n", a, b);
+#define PRINT_HEX(a, b) av_log(s->avctx, AV_LOG_DEBUG, " %s = %x\n", a, b);
+
+    PRINT("ed sample bit depth", s->bits_per_sample);
+    PRINT_HEX("ed decode flags", s->decode_flags);
+    PRINT("samples per frame",   s->samples_per_frame);
+    PRINT("log2 frame size",     s->log2_frame_size);
+    PRINT("max num subframes",   s->max_num_subframes);
+    PRINT("len prefix",          s->len_prefix);
+    PRINT("num channels",        s->num_channels);
+}
+
 /**
  *@brief Uninitialize the decoder and free all resources.
  *@param avctx codec context
@@ -93,7 +245,7 @@
  */
 static av_cold int decode_end(AVCodecContext *avctx)
 {
-    WMA3DecodeContext *s = avctx->priv_data;
+    WMAProDecodeCtx *s = avctx->priv_data;
     int i;
 
     for (i = 0; i < WMAPRO_BLOCK_SIZES; i++)
@@ -103,18 +255,351 @@
 }
 
 /**
+ *@brief Initialize the decoder.
+ *@param avctx codec context
+ *@return 0 on success, -1 otherwise
+ */
+static av_cold int decode_init(AVCodecContext *avctx)
+{
+    WMAProDecodeCtx *s = avctx->priv_data;
+    uint8_t *edata_ptr   = avctx->extradata;
+    unsigned int channel_mask;
+    int i;
+    int log2_max_num_subframes;
+    int num_possible_block_sizes;
+
+    s->avctx = avctx;
+    dsputil_init(&s->dsp, avctx);
+    init_put_bits(&s->pb, s->frame_data, MAX_FRAMESIZE);
+
+    avctx->sample_fmt = SAMPLE_FMT_FLT;
+
+    if (avctx->extradata_size >= 18) {
+        s->decode_flags    = AV_RL16(edata_ptr+14);
+        channel_mask       = AV_RL32(edata_ptr+2);
+        s->bits_per_sample = AV_RL16(edata_ptr);
+        /** dump the extradata */
+        for (i = 0; i < avctx->extradata_size; i++)
+            dprintf(avctx, "[%x] ", avctx->extradata[i]);
+        dprintf(avctx, "\n");
+
+    } else {
+        av_log_ask_for_sample(avctx, "Unknown extradata size\n");
+        return AVERROR_INVALIDDATA;
+    }
+
+    /** generic init */
+    s->log2_frame_size = av_log2(avctx->block_align) + 4;
+
+    /** frame info */
+    s->skip_frame  = 1; /** skip first frame */
+    s->packet_loss = 1;
+    s->len_prefix  = (s->decode_flags & 0x40);
+
+    if (!s->len_prefix) {
+        av_log_ask_for_sample(avctx, "no length prefix\n");
+        return AVERROR_INVALIDDATA;
+    }
+
+    /** get frame len */
+    s->samples_per_frame = 1 << ff_wma_get_frame_len_bits(avctx->sample_rate,
+                                                          3, s->decode_flags);
+
+    /** init previous block len */
+    for (i = 0; i < avctx->channels; i++)
+        s->channel[i].prev_block_len = s->samples_per_frame;
+
+    /** subframe info */
+    log2_max_num_subframes       = ((s->decode_flags & 0x38) >> 3);
+    s->max_num_subframes         = 1 << log2_max_num_subframes;
+    if (s->max_num_subframes == 16)
+        s->max_subframe_len_bit = 1;
+    s->subframe_len_bits = av_log2(log2_max_num_subframes) + 1;
+
+    num_possible_block_sizes     = log2_max_num_subframes + 1;
+    s->min_samples_per_subframe  = s->samples_per_frame / s->max_num_subframes;
+    s->dynamic_range_compression = (s->decode_flags & 0x80);
+
+    if (s->max_num_subframes > MAX_SUBFRAMES) {
+        av_log(avctx, AV_LOG_ERROR, "invalid number of subframes %i\n",
+               s->max_num_subframes);
+        return AVERROR_INVALIDDATA;
+    }
+
+    s->num_channels = avctx->channels;
+
+    /** extract lfe channel position */
+    s->lfe_channel = -1;
+
+    if (channel_mask & 8) {
+        unsigned int mask;
+        for (mask = 1; mask < 16; mask <<= 1) {
+            if (channel_mask & mask)
+                ++s->lfe_channel;
+        }
+    }
+
+    if (s->num_channels < 0 || s->num_channels > WMAPRO_MAX_CHANNELS) {
+        av_log_ask_for_sample(avctx, "invalid number of channels\n");
+        return AVERROR_NOTSUPP;
+    }
+
+    INIT_VLC_STATIC(&sf_vlc, SCALEVLCBITS, HUFF_SCALE_SIZE,
+                    scale_huffbits, 1, 1,
+                    scale_huffcodes, 2, 2, 616);
+
+    INIT_VLC_STATIC(&sf_rl_vlc, VLCBITS, HUFF_SCALE_RL_SIZE,
+                    scale_rl_huffbits, 1, 1,
+                    scale_rl_huffcodes, 4, 4, 1406);
+
+    INIT_VLC_STATIC(&coef_vlc[0], VLCBITS, HUFF_COEF0_SIZE,
+                    coef0_huffbits, 1, 1,
+                    coef0_huffcodes, 4, 4, 2108);
+
+    INIT_VLC_STATIC(&coef_vlc[1], VLCBITS, HUFF_COEF1_SIZE,
+                    coef1_huffbits, 1, 1,
+                    coef1_huffcodes, 4, 4, 3912);
+
+    INIT_VLC_STATIC(&vec4_vlc, VLCBITS, HUFF_VEC4_SIZE,
+                    vec4_huffbits, 1, 1,
+                    vec4_huffcodes, 2, 2, 604);
+
+    INIT_VLC_STATIC(&vec2_vlc, VLCBITS, HUFF_VEC2_SIZE,
+                    vec2_huffbits, 1, 1,
+                    vec2_huffcodes, 2, 2, 562);
+
+    INIT_VLC_STATIC(&vec1_vlc, VLCBITS, HUFF_VEC1_SIZE,
+                    vec1_huffbits, 1, 1,
+                    vec1_huffcodes, 2, 2, 562);
+
+    /** calculate number of scale factor bands and their offsets
+        for every possible block size */
+    for (i = 0; i < num_possible_block_sizes; i++) {
+        int subframe_len = s->samples_per_frame >> i;
+        int x;
+        int band = 1;
+
+        s->sfb_offsets[i][0] = 0;
+
+        for (x = 0; x < MAX_BANDS-1 && s->sfb_offsets[i][band-1] < subframe_len; x++) {
+            int offset = (subframe_len * 2 * critical_freq[x])
+                          / s->avctx->sample_rate + 2;
+            offset &= ~3;
+            if ( offset > s->sfb_offsets[i][band - 1] )
+                s->sfb_offsets[i][band++] = offset;
+        }
+        s->sfb_offsets[i][band - 1] = subframe_len;
+        s->num_sfb[i]               = band - 1;
+    }
+
+
+    /** Scale factors can be shared between blocks of different size
+        as every block has a different scale factor band layout.
+        The matrix sf_offsets is needed to find the correct scale factor.
+     */
+
+    for (i = 0; i < num_possible_block_sizes; i++) {
+        int b;
+        for (b = 0; b < s->num_sfb[i]; b++) {
+            int x;
+            int offset = ((s->sfb_offsets[i][b]
+                         + s->sfb_offsets[i][b + 1] - 1)<<i) >> 1;
+            for (x = 0; x < num_possible_block_sizes; x++) {
+                int v = 0;
+                while (s->sfb_offsets[x][v + 1] << x < offset)
+                    ++v;
+                s->sf_offsets[i][x][b] = v;
+            }
+        }
+    }
+
+    /** init MDCT, FIXME: only init needed sizes */
+    for (i = 0; i < WMAPRO_BLOCK_SIZES; i++)
+        ff_mdct_init(&s->mdct_ctx[i], BLOCK_MIN_BITS+1+i, 1,
+                     1.0 / (1 <<(BLOCK_MIN_BITS + i - 1))
+                     / (1 << (s->bits_per_sample - 1)));
+
+    /** init MDCT windows: simple sinus window */
+    for (i = 0; i < WMAPRO_BLOCK_SIZES; i++) {
+        const int n       = 1 << (WMAPRO_BLOCK_MAX_BITS - i);
+        const int win_idx = WMAPRO_BLOCK_MAX_BITS - i - 7;
+        ff_sine_window_init(ff_sine_windows[win_idx], n);
+        s->windows[WMAPRO_BLOCK_SIZES-i-1] = ff_sine_windows[win_idx];
+    }
+
+    /** calculate subwoofer cutoff values */
+    for (i = 0; i < num_possible_block_sizes; i++) {
+        int block_size = s->samples_per_frame >> i;
+        int cutoff = (440*block_size + 3 * (s->avctx->sample_rate >> 1) - 1)
+                     / s->avctx->sample_rate;
+        s->subwoofer_cutoffs[i] = av_clip(cutoff, 4, block_size);
+    }
+
+    /** calculate sine values for the decorrelation matrix */
+    for (i = 0; i < 33; i++)
+        sin64[i] = sin(i*M_PI / 64.0);
+
+    if (avctx->debug & FF_DEBUG_BITSTREAM)
+        dump_context(s);
+
+    avctx->channel_layout = channel_mask;
+    return 0;
+}
+
+/**
+ *@brief Decode the subframe length.
+ *@param s context
+ *@param offset sample offset in the frame
+ *@return decoded subframe length on success, < 0 in case of an error
+ */
+static int decode_subframe_length(WMAProDecodeCtx *s, int offset)
+{
+    int frame_len_shift = 0;
+    int subframe_len;
+
+    /** no need to read from the bitstream when only one length is possible */
+    if (offset == s->samples_per_frame - s->min_samples_per_subframe)
+        return s->min_samples_per_subframe;
+
+    /** 1 bit indicates if the subframe is of maximum length */
+    if (s->max_subframe_len_bit) {
+        if (get_bits1(&s->gb))
+            frame_len_shift = 1 + get_bits(&s->gb, s->subframe_len_bits-1);
+    } else
+        frame_len_shift = get_bits(&s->gb, s->subframe_len_bits);
+
+    subframe_len = s->samples_per_frame >> frame_len_shift;
+
+    /** sanity check the length */
+    if (subframe_len < s->min_samples_per_subframe
+              || subframe_len > s->samples_per_frame) {
+        av_log(s->avctx, AV_LOG_ERROR, "broken frame: subframe_len %i\n",
+               subframe_len);
+        return AVERROR_INVALIDDATA;
+    }
+    return subframe_len;
+}
+
+/**
+ *@brief Decode how the data in the frame is split into subframes.
+ *       Every WMA frame contains the encoded data for a fixed number of
+ *       samples per channel. The data for every channel might be split
+ *       into several subframes. This function will reconstruct the list of
+ *       subframes for every channel.
+ *
+ *       If the subframes are not evenly split, the algorithm estimates the
+ *       channels with the lowest number of total samples.
+ *       Afterwards, for each of these channels a bit is read from the
+ *       bitstream that indicates if the channel contains a subframe with the
+ *       next subframe size that is going to be read from the bitstream or not.
+ *       If a channel contains such a subframe, the subframe size gets added to
+ *       the channel's subframe list.
+ *       The algorithm repeats these steps until the frame is properly divided
+ *       between the individual channels.
+ *
+ *@param s context
+ *@return 0 on success, < 0 in case of an error
+ */
+static int decode_tilehdr(WMAProDecodeCtx *s)
+{
+    uint16_t num_samples[WMAPRO_MAX_CHANNELS];        /** sum of samples for all currently known subframes of a channel */
+    uint8_t  contains_subframe[WMAPRO_MAX_CHANNELS];  /** flag indicating if a channel contains the current subframe */
+    int channels_for_cur_subframe = s->num_channels;  /** number of channels that contain the current subframe */
+    int fixed_channel_layout = 0;                     /** flag indicating that all channels use the same subframe offsets and sizes */
+    int min_channel_len = 0;                          /** smallest sum of samples (channels with this length will be processed first) */
+    int c;
+
+    /* Should never consume more than 3073 bits (256 iterations for the
+     * while loop when always the minimum amount of 128 samples is substracted
+     * from missing samples in the 8 channel case).
+     * 1 + BLOCK_MAX_SIZE * MAX_CHANNELS / BLOCK_MIN_SIZE * (MAX_CHANNELS  + 4)
+     */
+
+    /** reset tiling information */
+    for (c = 0; c < s->num_channels; c++)
+        s->channel[c].num_subframes = 0;
+
+    memset(num_samples, 0, sizeof(num_samples));
+
+    if (s->max_num_subframes == 1 || get_bits1(&s->gb))
+        fixed_channel_layout = 1;
+
+    /** loop until the frame data is split between the subframes */
+    do {
+        int subframe_len;
+
+        /** check which channels contain the subframe */
+        for (c = 0; c < s->num_channels; c++) {
+            if (num_samples[c] == min_channel_len) {
+                if (fixed_channel_layout || channels_for_cur_subframe == 1 ||
+                   (min_channel_len == s->samples_per_frame - s->min_samples_per_subframe))
+                    contains_subframe[c] = 1;
+                else
+                    contains_subframe[c] = get_bits1(&s->gb);
+            } else
+                contains_subframe[c] = 0;
+        }
+
+        /** get subframe length, subframe_len == 0 is not allowed */
+        if ((subframe_len = decode_subframe_length(s, min_channel_len)) <= 0)
+            return AVERROR_INVALIDDATA;
+
+        /** add subframes to the individual channels and find new min_channel_len */
+        min_channel_len += subframe_len;
+        for (c = 0; c < s->num_channels; c++) {
+            WMAProChannelCtx* chan = &s->channel[c];
+
+            if (contains_subframe[c]) {
+                if (chan->num_subframes >= MAX_SUBFRAMES) {
+                    av_log(s->avctx, AV_LOG_ERROR,
+                           "broken frame: num subframes > 31\n");
+                    return AVERROR_INVALIDDATA;
+                }
+                chan->subframe_len[chan->num_subframes] = subframe_len;
+                num_samples[c] += subframe_len;
+                ++chan->num_subframes;
+                if (num_samples[c] > s->samples_per_frame) {
+                    av_log(s->avctx, AV_LOG_ERROR,"broken frame: "
+                           "channel len > samples_per_frame\n");
+                    return AVERROR_INVALIDDATA;
+                }
+            } else if(num_samples[c] <= min_channel_len) {
+                if (num_samples[c] < min_channel_len) {
+                    channels_for_cur_subframe = 0;
+                    min_channel_len = num_samples[c];
+                }
+                ++channels_for_cur_subframe;
+            }
+        }
+    } while (min_channel_len < s->samples_per_frame);
+
+    for (c = 0; c < s->num_channels; c++) {
+        int i;
+        int offset = 0;
+        for (i = 0; i < s->channel[c].num_subframes; i++) {
+            dprintf(s->avctx, "frame[%i] channel[%i] subframe[%i]"
+                   " len %i\n", s->frame_num, c, i, s->channel[c].subframe_len[i]);
+            s->channel[c].subframe_offset[i] = offset;
+            offset += s->channel[c].subframe_len[i];
+        }
+    }
+
+    return 0;
+}
+
+/**
  *@brief Calculate a decorrelation matrix from the bitstream parameters.
  *@param s codec context
  *@param chgroup channel group for which the matrix needs to be calculated
  */
-static void decode_decorrelation_matrix(WMA3DecodeContext *s,
-                                        WMA3ChannelGroup *chgroup)
+static void decode_decorrelation_matrix(WMAProDecodeCtx *s,
+                                        WMAProChannelGrp *chgroup)
 {
     int i;
     int offset = 0;
     int8_t rotation_offset[WMAPRO_MAX_CHANNELS * WMAPRO_MAX_CHANNELS];
-    memset(chgroup->decorrelation_matrix, 0,
-           sizeof(float) *s->num_channels * s->num_channels);
+    memset(chgroup->decorrelation_matrix, 0, s->num_channels *
+           s->num_channels * sizeof(*chgroup->decorrelation_matrix));
 
     for (i = 0; i < chgroup->num_channels * (chgroup->num_channels - 1) >> 1; i++)
         rotation_offset[i] = get_bits(&s->gb, 6);
@@ -153,16 +638,128 @@
 }
 
 /**
+ *@brief Decode channel transformation parameters
+ *@param s codec context
+ *@return 0 in case of success, < 0 in case of bitstream errors
+ */
+static int decode_channel_transform(WMAProDecodeCtx* s)
+{
+    int i;
+    /* should never consume more than 1921 bits for the 8 channel case
+     * 1 + MAX_CHANNELS * ( MAX_CHANNELS + 2 + 3 * MAX_CHANNELS * MAX_CHANNELS
+     * + MAX_CHANNELS + MAX_BANDS + 1)
+     */
+
+    /** in the one channel case channel transforms are pointless */
+    s->num_chgroups = 0;
+    if (s->num_channels > 1) {
+        int remaining_channels = s->channels_for_cur_subframe;
+
+        if (get_bits1(&s->gb)) {
+            av_log_ask_for_sample(s->avctx,
+                                  "unsupported channel transform bit\n");
+            return AVERROR_INVALIDDATA;
+        }
+
+        for (s->num_chgroups = 0; remaining_channels &&
+            s->num_chgroups < s->channels_for_cur_subframe; s->num_chgroups++) {
+            WMAProChannelGrp* chgroup = &s->chgroup[s->num_chgroups];
+            float** channel_data = chgroup->channel_data;
+            chgroup->num_channels = 0;
+            chgroup->transform = 0;
+
+            /** decode channel mask */
+            if (remaining_channels > 2) {
+                for (i = 0; i < s->channels_for_cur_subframe; i++) {
+                    int channel_idx = s->channel_indexes_for_cur_subframe[i];
+                    if (!s->channel[channel_idx].grouped
+                        && get_bits1(&s->gb)) {
+                        ++chgroup->num_channels;
+                        s->channel[channel_idx].grouped = 1;
+                        *channel_data++ = s->channel[channel_idx].coeffs;
+                    }
+                }
+            } else {
+                chgroup->num_channels = remaining_channels;
+                for (i = 0; i < s->channels_for_cur_subframe; i++) {
+                    int channel_idx = s->channel_indexes_for_cur_subframe[i];
+                    if (!s->channel[channel_idx].grouped)
+                        *channel_data++ = s->channel[channel_idx].coeffs;
+                    s->channel[channel_idx].grouped = 1;
+                }
+            }
+
+            /** decode transform type */
+            if (chgroup->num_channels == 2) {
+                if (get_bits1(&s->gb)) {
+                    if (get_bits1(&s->gb)) {
+                        av_log_ask_for_sample(s->avctx,
+                               "unsupported channel transform type\n");
+                    }
+                } else {
+                    chgroup->transform = 1;
+                    if (s->num_channels == 2) {
+                        chgroup->decorrelation_matrix[0] =  1.0;
+                        chgroup->decorrelation_matrix[1] = -1.0;
+                        chgroup->decorrelation_matrix[2] =  1.0;
+                        chgroup->decorrelation_matrix[3] =  1.0;
+                    } else {
+                        /** cos(pi/4) */
+                        chgroup->decorrelation_matrix[0] =  0.70703125;
+                        chgroup->decorrelation_matrix[1] = -0.70703125;
+                        chgroup->decorrelation_matrix[2] =  0.70703125;
+                        chgroup->decorrelation_matrix[3] =  0.70703125;
+                    }
+                }
+            } else if (chgroup->num_channels > 2) {
+                if (get_bits1(&s->gb)) {
+                    chgroup->transform = 1;
+                    if (get_bits1(&s->gb)) {
+                        decode_decorrelation_matrix(s, chgroup);
+                    } else {
+                        /** FIXME: more than 6 coupled channels not supported */
+                        if (chgroup->num_channels > 6) {
+                            av_log_ask_for_sample(s->avctx,
+                                   "coupled channels > 6\n");
+                        } else {
+                            memcpy(chgroup->decorrelation_matrix,
+                              default_decorrelation[chgroup->num_channels],
+                              chgroup->num_channels * chgroup->num_channels *
+                              sizeof(*chgroup->decorrelation_matrix));
+                        }
+                    }
+                }
+            }
+
+            /** decode transform on / off */
+            if (chgroup->transform) {
+                if (!get_bits1(&s->gb)) {
+                    int i;
+                    /** transform can be enabled for individual bands */
+                    for (i = 0; i < s->num_bands; i++) {
+                        chgroup->transform_band[i] = get_bits1(&s->gb);
+                    }
+                } else {
+                    memset(chgroup->transform_band, 1, s->num_bands);
+                }
+            }
+            remaining_channels -= chgroup->num_channels;
+        }
+    }
+    return 0;
+}
+
+/**
  *@brief Extract the coefficients from the bitstream.
  *@param s codec context
  *@param c current channel number
  *@return 0 on success, < 0 in case of bitstream errors
  */
-static int decode_coeffs(WMA3DecodeContext *s, int c)
+static int decode_coeffs(WMAProDecodeCtx *s, int c)
 {
     int vlctable;
     VLC* vlc;
-    WMA3ChannelCtx* ci = &s->channel[c];
+    WMAProChannelCtx* ci = &s->channel[c];
     int rl_mode = 0;
     int cur_coeff = 0;
     int num_zeros = 0;
@@ -220,6 +817,7 @@
                 ci->coeffs[cur_coeff] = (vals[i]^sign) - sign;
                 num_zeros = 0;
             } else {
+                ci->coeffs[cur_coeff] = 0;
                 /** switch to run level mode when subframe_len / 128 zeros
                    were found in a row */
                 rl_mode |= (++num_zeros > s->subframe_len>>8);
@@ -230,6 +828,8 @@
 
     /** decode run level coded coefficients */
     if (rl_mode) {
+        memset(&ci->coeffs[cur_coeff], 0,
+               sizeof(*ci->coeffs) * (s->subframe_len - cur_coeff));
         if (ff_wma_run_level_decode(s->avctx, &s->gb, vlc,
                                     level, run, 1, ci->coeffs,
                                     cur_coeff, s->subframe_len,
@@ -241,10 +841,106 @@
 }
 
 /**
+ *@brief Extract scale factors from the bitstream.
+ *@param s codec context
+ *@return 0 on success, < 0 in case of bitstream errors
+ */
+static int decode_scale_factors(WMAProDecodeCtx* s)
+{
+    int i;
+
+    /** should never consume more than 5344 bits
+     *  MAX_CHANNELS * (1 +  MAX_BANDS * 23)
+     */
+
+    for (i = 0; i < s->channels_for_cur_subframe; i++) {
+        int c = s->channel_indexes_for_cur_subframe[i];
+        int* sf;
+        int* sf_end = s->channel[c].scale_factors + s->num_bands;
+
+        /** resample scale factors for the new block size
+         *  as the scale factors might need to be resampled several times
+         *  before some  new values are transmitted, a backup of the last
+         *  transmitted scale factors is kept in saved_scale_factors
+         */
+        if (s->channel[c].reuse_sf) {
+            const int8_t* sf_offsets = s->sf_offsets[s->table_idx][s->channel[c].table_idx];
+            int b;
+            for (b = 0; b < s->num_bands; b++)
+                s->channel[c].scale_factors[b] =
+                                   s->channel[c].saved_scale_factors[*sf_offsets++];
+        }
+
+        if (!s->channel[c].cur_subframe || get_bits1(&s->gb)) {
+
+            if (!s->channel[c].reuse_sf) {
+                int val;
+                /** decode DPCM coded scale factors */
+                s->channel[c].scale_factor_step = get_bits(&s->gb, 2) + 1;
+                val = 45 / s->channel[c].scale_factor_step;
+                for (sf = s->channel[c].scale_factors; sf < sf_end; sf++) {
+                    val += get_vlc2(&s->gb, sf_vlc.table, SCALEVLCBITS, SCALEMAXDEPTH) - 60;
+                    *sf = val;
+                }
+            } else {
+                int i;
+                /** run level decode differences to the resampled factors */
+                for (i = 0; i < s->num_bands; i++) {
+                    int idx;
+                    int skip;
+                    int val;
+                    int sign;
+
+                    idx = get_vlc2(&s->gb, sf_rl_vlc.table, VLCBITS, SCALERLMAXDEPTH);
+
+                    if ( !idx ) {
+                        uint32_t code = get_bits(&s->gb, 14);
+                        val  =  code >> 6;
+                        sign = (code & 1) - 1;
+                        skip = (code & 0x3f) >> 1;
+                    } else if (idx == 1) {
+                        break;
+                    } else {
+                        skip = scale_rl_run[idx];
+                        val  = scale_rl_level[idx];
+                        sign = get_bits1(&s->gb)-1;
+                    }
+
+                    i += skip;
+                    if (i >= s->num_bands) {
+                        av_log(s->avctx,AV_LOG_ERROR,
+                               "invalid scale factor coding\n");
+                        return AVERROR_INVALIDDATA;
+                    }
+                    s->channel[c].scale_factors[i] += (val ^ sign) - sign;
+                }
+            }
+
+            /** save transmitted scale factors so that they can be reused for
+                the next subframe */
+            memcpy(s->channel[c].saved_scale_factors,
+                   s->channel[c].scale_factors, s->num_bands *
+                   sizeof(*s->channel[c].saved_scale_factors));
+            s->channel[c].table_idx = s->table_idx;
+            s->channel[c].reuse_sf  = 1;
+        }
+
+        /** calculate new scale factor maximum */
+        s->channel[c].max_scale_factor = s->channel[c].scale_factors[0];
+        for (sf = s->channel[c].scale_factors + 1; sf < sf_end; sf++) {
+            s->channel[c].max_scale_factor =
+                FFMAX(s->channel[c].max_scale_factor, *sf);
+        }
+
+    }
+    return 0;
+}
+
+/**
  *@brief Reconstruct the individual channel data.
  *@param s codec context
  */
-static void inverse_channel_transform(WMA3DecodeContext *s)
+static void inverse_channel_transform(WMAProDecodeCtx *s)
 {
     int i;
 
@@ -292,3 +988,573 @@
     }
 }
 
+/**
+ *@brief Apply sine window and reconstruct the output buffer.
+ *@param s codec context
+ */
+static void wmapro_window(WMAProDecodeCtx *s)
+{
+    int i;
+    for (i = 0; i< s->channels_for_cur_subframe; i++) {
+        int c = s->channel_indexes_for_cur_subframe[i];
+        float* window;
+        int winlen = s->channel[c].prev_block_len;
+        float* start = s->channel[c].coeffs - (winlen >> 1);
+
+        if (s->subframe_len < winlen) {
+            start += (winlen - s->subframe_len)>>1;
+            winlen = s->subframe_len;
+        }
+
+        window = s->windows[av_log2(winlen)-BLOCK_MIN_BITS];
+
+        winlen >>= 1;
+
+        s->dsp.vector_fmul_window(start, start, start + winlen,
+                                  window, 0, winlen);
+
+        s->channel[c].prev_block_len = s->subframe_len;
+    }
+}
+
+/**
+ *@brief Decode a single subframe (block).
+ *@param s codec context
+ *@return 0 on success, < 0 when decoding failed
+ */
+static int decode_subframe(WMAProDecodeCtx *s)
+{
+    int offset = s->samples_per_frame;
+    int subframe_len = s->samples_per_frame;
+    int i;
+    int total_samples   = s->samples_per_frame * s->num_channels;
+    int transmit_coeffs = 0;
+    int cur_subwoofer_cutoff;
+
+    s->subframe_offset = get_bits_count(&s->gb);
+
+    /** reset channel context and find the next block offset and size
+        == the next block of the channel with the smallest number of
+        decoded samples
+    */
+    for (i = 0; i < s->num_channels; i++) {
+        s->channel[i].grouped = 0;
+        if (offset > s->channel[i].decoded_samples) {
+            offset = s->channel[i].decoded_samples;
+            subframe_len =
+                s->channel[i].subframe_len[s->channel[i].cur_subframe];
+        }
+    }
+
+    dprintf(s->avctx,
+           "processing subframe with offset %i len %i\n", offset, subframe_len);
+
+    /** get a list of all channels that contain the estimated block */
+    s->channels_for_cur_subframe = 0;
+    for (i = 0; i < s->num_channels; i++) {
+        const int cur_subframe = s->channel[i].cur_subframe;
+        /** substract already processed samples */
+        total_samples -= s->channel[i].decoded_samples;
+
+        /** and count if there are multiple subframes that match our profile */
+        if (offset == s->channel[i].decoded_samples &&
+           subframe_len == s->channel[i].subframe_len[cur_subframe]) {
+            total_samples -= s->channel[i].subframe_len[cur_subframe];
+            s->channel[i].decoded_samples +=
+                s->channel[i].subframe_len[cur_subframe];
+            s->channel_indexes_for_cur_subframe[s->channels_for_cur_subframe] = i;
+            ++s->channels_for_cur_subframe;
+        }
+    }
+
+    /** check if the frame will be complete after processing the
+        estimated block */
+    if (!total_samples)
+        s->parsed_all_subframes = 1;
+
+
+    dprintf(s->avctx, "subframe is part of %i channels\n",
+           s->channels_for_cur_subframe);
+
+    /** calculate number of scale factor bands and their offsets */
+    s->table_idx         = av_log2(s->samples_per_frame/subframe_len);
+    s->num_bands         = s->num_sfb[s->table_idx];
+    s->cur_sfb_offsets   = s->sfb_offsets[s->table_idx];
+    cur_subwoofer_cutoff = s->subwoofer_cutoffs[s->table_idx];
+
+    /** configure the decoder for the current subframe */
+    for (i = 0; i < s->channels_for_cur_subframe; i++) {
+        int c = s->channel_indexes_for_cur_subframe[i];
+
+        s->channel[c].coeffs = &s->channel[c].out[(s->samples_per_frame>>1)
+                                                  + offset];
+    }
+
+    s->subframe_len = subframe_len;
+    s->esc_len = av_log2(s->subframe_len - 1) + 1;
+
+    /** skip extended header if any */
+    if (get_bits1(&s->gb)) {
+        int num_fill_bits;
+        if (!(num_fill_bits = get_bits(&s->gb, 2))) {
+            int len = get_bits(&s->gb, 4);
+            num_fill_bits = get_bits(&s->gb, len) + 1;
+        }
+
+        if (num_fill_bits >= 0) {
+            if (get_bits_count(&s->gb) + num_fill_bits > s->num_saved_bits) {
+                av_log(s->avctx,AV_LOG_ERROR,"invalid number of fill bits\n");
+                return AVERROR_INVALIDDATA;
+            }
+
+            skip_bits_long(&s->gb, num_fill_bits);
+        }
+    }
+
+    /** no idea for what the following bit is used */
+    if (get_bits1(&s->gb)) {
+        av_log_ask_for_sample(s->avctx, "reserved bit set\n");
+        return AVERROR_INVALIDDATA;
+    }
+
+
+    if (decode_channel_transform(s) < 0)
+        return AVERROR_INVALIDDATA;
+
+
+    for (i = 0; i < s->channels_for_cur_subframe; i++) {
+        int c = s->channel_indexes_for_cur_subframe[i];
+        if ((s->channel[c].transmit_coefs = get_bits1(&s->gb)))
+            transmit_coeffs = 1;
+    }
+
+    if (transmit_coeffs) {
+        int step;
+        int quant_step = 90 * s->bits_per_sample >> 4;
+        if ((get_bits1(&s->gb))) {
+            /** FIXME: might change run level mode decision */
+            av_log_ask_for_sample(s->avctx, "unsupported quant step coding\n");
+            return AVERROR_INVALIDDATA;
+        }
+        /** decode quantization step */
+        step = get_sbits(&s->gb, 6);
+        quant_step += step;
+        if (step == -32 || step == 31) {
+            const int sign = (step == 31) - 1;
+            int quant = 0;
+            while (get_bits_count(&s->gb) + 5 < s->num_saved_bits &&
+                   (step = get_bits(&s->gb, 5)) == 31 ) {
+                     quant += 31;
+            }
+            quant_step += ((quant + step) ^ sign) - sign;
+        }
+        if (quant_step < 0) {
+            av_log(s->avctx,AV_LOG_DEBUG,"negative quant step\n");
+        }
+
+        /** decode quantization step modifiers for every channel */
+
+        if (s->channels_for_cur_subframe == 1) {
+            s->channel[s->channel_indexes_for_cur_subframe[0]].quant_step = quant_step;
+        } else {
+            int modifier_len = get_bits(&s->gb, 3);
+            for (i = 0; i < s->channels_for_cur_subframe; i++) {
+                int c = s->channel_indexes_for_cur_subframe[i];
+                s->channel[c].quant_step = quant_step;
+                if (get_bits1(&s->gb)) {
+                    if (modifier_len) {
+                        s->channel[c].quant_step +=
+                                get_bits(&s->gb, modifier_len) + 1;
+                    } else
+                        ++s->channel[c].quant_step;
+                }
+            }
+        }
+
+        /** decode scale factors */
+        if (decode_scale_factors(s) < 0)
+            return AVERROR_INVALIDDATA;
+    }
+
+    dprintf(s->avctx, "BITSTREAM: subframe header length was %i\n",
+           get_bits_count(&s->gb) - s->subframe_offset);
+
+    /** parse coefficients */
+    for (i = 0; i < s->channels_for_cur_subframe; i++) {
+        int c = s->channel_indexes_for_cur_subframe[i];
+        if (s->channel[c].transmit_coefs &&
+           get_bits_count(&s->gb) < s->num_saved_bits) {
+                decode_coeffs(s, c);
+        } else
+            memset(s->channel[c].coeffs, 0,
+                   sizeof(*s->channel[c].coeffs) * subframe_len);
+    }
+
+    dprintf(s->avctx, "BITSTREAM: subframe length was %i\n",
+           get_bits_count(&s->gb) - s->subframe_offset);
+
+    if (transmit_coeffs) {
+        /** reconstruct the per channel data */
+        inverse_channel_transform(s);
+        for (i = 0; i < s->channels_for_cur_subframe; i++) {
+            int c = s->channel_indexes_for_cur_subframe[i];
+            const int* sf = s->channel[c].scale_factors;
+            int b;
+
+            if (c == s->lfe_channel)
+                memset(&s->tmp[cur_subwoofer_cutoff], 0, sizeof(*s->tmp) *
+                       (subframe_len - cur_subwoofer_cutoff));
+
+            /** inverse quantization and rescaling */
+            for (b = 0; b < s->num_bands; b++) {
+                const int end = FFMIN(s->cur_sfb_offsets[b+1], s->subframe_len);
+                const int exp = s->channel[c].quant_step -
+                            (s->channel[c].max_scale_factor - *sf++) *
+                            s->channel[c].scale_factor_step;
+                const float quant = pow(10.0, exp / 20.0);
+                int start;
+
+                for (start = s->cur_sfb_offsets[b]; start < end; start++)
+                    s->tmp[start] = s->channel[c].coeffs[start] * quant;
+            }
+
+            /** apply imdct (ff_imdct_half == DCTIV with reverse) */
+            ff_imdct_half(&s->mdct_ctx[av_log2(subframe_len)-BLOCK_MIN_BITS],
+                          s->channel[c].coeffs, s->tmp);
+        }
+    }
+
+    /** window and overlapp-add */
+    wmapro_window(s);
+
+    /** handled one subframe */
+    for (i = 0; i < s->channels_for_cur_subframe; i++) {
+        int c = s->channel_indexes_for_cur_subframe[i];
+        if (s->channel[c].cur_subframe >= s->channel[c].num_subframes) {
+            av_log(s->avctx,AV_LOG_ERROR,"broken subframe\n");
+            return AVERROR_INVALIDDATA;
+        }
+        ++s->channel[c].cur_subframe;
+    }
+
+    return 0;
+}
+
+/**
+ *@brief Decode one WMA frame.
+ *@param s codec context
+ *@return 0 if the trailer bit indicates that this is the last frame,
+ *        1 if there are additional frames
+ */
+static int decode_frame(WMAProDecodeCtx *s)
+{
+    GetBitContext* gb = &s->gb;
+    int more_frames = 0;
+    int len = 0;
+    int i;
+
+    /** check for potential output buffer overflow */
+    if (s->num_channels * s->samples_per_frame > s->samples_end - s->samples) {
+        av_log(s->avctx,AV_LOG_ERROR,
+               "not enough space for the output samples\n");
+        s->packet_loss = 1;
+        return 0;
+    }
+
+    /** get frame length */
+    if (s->len_prefix)
+        len = get_bits(gb, s->log2_frame_size);
+
+    dprintf(s->avctx, "decoding frame with length %x\n", len);
+
+    /** decode tile information */
+    if (decode_tilehdr(s)) {
+        s->packet_loss = 1;
+        return 0;
+    }
+
+    /** read postproc transform */
+    if (s->num_channels > 1 && get_bits1(gb)) {
+        av_log_ask_for_sample(s->avctx, "Unsupported postproc transform found\n");
+        s->packet_loss = 1;
+        return 0;
+    }
+
+    /** read drc info */
+    if (s->dynamic_range_compression) {
+        s->drc_gain = get_bits(gb, 8);
+        dprintf(s->avctx, "drc_gain %i\n", s->drc_gain);
+    }
+
+    /** no idea what these are for, might be the number of samples
+        that need to be skipped at the beginning or end of a stream */
+    if (get_bits1(gb)) {
+        int skip;
+
+        /** usually true for the first frame */
+        if (get_bits1(gb)) {
+            skip = get_bits(gb, av_log2(s->samples_per_frame * 2));
+            dprintf(s->avctx, "start skip: %i\n", skip);
+        }
+
+        /** sometimes true for the last frame */
+        if (get_bits1(gb)) {
+            skip = get_bits(gb, av_log2(s->samples_per_frame * 2));
+            dprintf(s->avctx, "end skip: %i\n", skip);
+        }
+
+    }
+
+    dprintf(s->avctx, "BITSTREAM: frame header length was %i\n",
+           get_bits_count(gb) - s->frame_offset);
+
+    /** reset subframe states */
+    s->parsed_all_subframes = 0;
+    for (i = 0; i < s->num_channels; i++) {
+        s->channel[i].decoded_samples = 0;
+        s->channel[i].cur_subframe    = 0;
+        s->channel[i].reuse_sf        = 0;
+    }
+
+    /** decode all subframes */
+    while (!s->parsed_all_subframes) {
+        if (decode_subframe(s) < 0) {
+            s->packet_loss = 1;
+            return 0;
+        }
+    }
+
+    /** interleave samples and write them to the output buffer */
+    for (i = 0; i < s->num_channels; i++) {
+        float* ptr;
+        int incr = s->num_channels;
+        float* iptr = s->channel[i].out;
+        int x;
+
+        ptr = s->samples + i;
+
+        for (x = 0; x < s->samples_per_frame; x++) {
+            *ptr = av_clipf(*iptr++, -1.0, 32767.0 / 32768.0);
+            ptr += incr;
+        }
+
+        /** reuse second half of the IMDCT output for the next frame */
+        memcpy(&s->channel[i].out[0],
+               &s->channel[i].out[s->samples_per_frame],
+               s->samples_per_frame * sizeof(*s->channel[i].out) >> 1);
+    }
+
+    if (s->skip_frame) {
+        s->skip_frame = 0;
+    } else
+        s->samples += s->num_channels * s->samples_per_frame;
+
+    if (len != (get_bits_count(gb) - s->frame_offset) + 2) {
+        /** FIXME: not sure if this is always an error */
+        av_log(s->avctx,AV_LOG_ERROR,"frame[%i] would have to skip %i bits\n",
+               s->frame_num, len - (get_bits_count(gb) - s->frame_offset) - 1);
+        s->packet_loss = 1;
+        return 0;
+    }
+
+    /** skip the rest of the frame data */
+    skip_bits_long(gb, len - (get_bits_count(gb) - s->frame_offset) - 1);
+
+    /** decode trailer bit */
+    more_frames = get_bits1(gb);
+
+    ++s->frame_num;
+    return more_frames;
+}
+
+/**
+ *@brief Calculate remaining input buffer length.
+ *@param s codec context
+ *@param gb bitstream reader context
+ *@return remaining size in bits
+ */
+static int remaining_bits(WMAProDecodeCtx *s, GetBitContext* gb)
+{
+    return s->buf_bit_size - get_bits_count(gb);
+}
+
+/**
+ *@brief Fill the bit reservoir with a (partial) frame.
+ *@param s codec context
+ *@param gb bitstream reader context
+ *@param len length of the partial frame
+ *@param append decides wether to reset the buffer or not
+ */
+static void save_bits(WMAProDecodeCtx *s, GetBitContext* gb, int len,
+                          int append)
+{
+    int buflen;
+
+    /** when the frame data does not need to be concatenated, the input buffer
+        is resetted and additional bits from the previous frame are copyed
+        and skipped later so that a fast byte copy is possible */
+
+    if (!append) {
+        s->frame_offset = get_bits_count(gb) & 7;
+        s->num_saved_bits = s->frame_offset;
+        init_put_bits(&s->pb, s->frame_data, MAX_FRAMESIZE);
+    }
+
+    buflen = (s->num_saved_bits + len + 8) >> 3;
+
+    if (len <= 0 || buflen > MAX_FRAMESIZE) {
+         av_log_ask_for_sample(s->avctx, "input buffer too small\n");
+         s->packet_loss = 1;
+         return;
+    }
+
+    s->num_saved_bits += len;
+    if (!append) {
+        ff_copy_bits(&s->pb, gb->buffer + (get_bits_count(gb) >> 3), s->num_saved_bits);
+    } else {
+        int align = 8 - (get_bits_count(gb) & 7);
+        align = FFMIN(align, len);
+        put_bits(&s->pb, align, get_bits(gb, align));
+        len -= align;
+        ff_copy_bits(&s->pb, gb->buffer + (get_bits_count(gb) >> 3), len);
+    }
+    skip_bits_long(gb, len);
+
+    {
+    PutBitContext tmp = s->pb;
+    flush_put_bits(&tmp);
+    }
+
+    init_get_bits(&s->gb, s->frame_data, s->num_saved_bits);
+    skip_bits(&s->gb, s->frame_offset);
+}
+
+/**
+ *@brief Decode a single WMA packet.
+ *@param avctx codec context
+ *@param data the output buffer
+ *@param data_size number of bytes that were written to the output buffer
+ *@param avpkt input packet
+ *@return number of bytes that were read from the input buffer
+ */
+static int decode_packet(AVCodecContext *avctx,
+                             void *data, int *data_size, AVPacket* avpkt)
+{
+    GetBitContext gb;
+    WMAProDecodeCtx *s = avctx->priv_data;
+    const uint8_t* buf   = avpkt->data;
+    int buf_size         = avpkt->size;
+    int more_frames      = 1;
+    int num_bits_prev_frame;
+    int packet_sequence_number;
+
+    s->samples      = data;
+    s->samples_end  = (float*)((int8_t*)data + *data_size);
+    s->buf_bit_size = buf_size << 3;
+
+
+    *data_size = 0;
+
+    /** sanity check for the buffer length */
+    if (buf_size < avctx->block_align)
+        return 0;
+
+    buf_size = avctx->block_align;
+
+    /** parse packet header */
+    init_get_bits(&gb, buf, s->buf_bit_size);
+    packet_sequence_number = get_bits(&gb, 4);
+    skip_bits(&gb, 2);
+
+    /** get number of bits that need to be added to the previous frame */
+    num_bits_prev_frame = get_bits(&gb, s->log2_frame_size);
+    dprintf(avctx, "packet[%d]: nbpf %x\n", avctx->frame_number,
+                  num_bits_prev_frame);
+
+    /** check for packet loss */
+    if (!s->packet_loss &&
+        ((s->packet_sequence_number + 1)&0xF) != packet_sequence_number) {
+        s->packet_loss = 1;
+        av_log(avctx, AV_LOG_ERROR, "Packet loss detected! seq %x vs %x\n",
+                      s->packet_sequence_number, packet_sequence_number);
+    }
+    s->packet_sequence_number = packet_sequence_number;
+
+    if (num_bits_prev_frame > 0) {
+        /** append the previous frame data to the remaining data from the
+            previous packet to create a full frame */
+        save_bits(s, &gb, num_bits_prev_frame, 1);
+        dprintf(avctx, "accumulated %x bits of frame data\n",
+                      s->num_saved_bits - s->frame_offset);
+
+        /** decode the cross packet frame if it is valid */
+        if (!s->packet_loss)
+            decode_frame(s);
+    } else if (s->num_saved_bits - s->frame_offset) {
+        dprintf(avctx, "ignoring %x previously saved bits\n",
+                      s->num_saved_bits - s->frame_offset);
+    }
+
+    s->packet_loss = 0;
+    /** decode the rest of the packet */
+    while (!s->packet_loss && more_frames &&
+          remaining_bits(s, &gb) > s->log2_frame_size) {
+        int frame_size = show_bits(&gb, s->log2_frame_size);
+
+        /** there is enough data for a full frame */
+        if (remaining_bits(s,&gb) >= frame_size && frame_size > 0) {
+            save_bits(s, &gb, frame_size, 0);
+
+            /** decode the frame */
+            more_frames = decode_frame(s);
+
+            if (!more_frames) {
+                dprintf(avctx, "no more frames\n");
+            }
+        } else
+            more_frames = 0;
+    }
+
+    if (!s->packet_loss && remaining_bits(s,&gb) > 0) {
+        /** save the rest of the data so that it can be decoded
+            with the next packet */
+        save_bits(s, &gb, remaining_bits(s,&gb), 0);
+    }
+
+    *data_size = (int8_t *)s->samples - (int8_t *)data;
+
+    return avctx->block_align;
+}
+
+/**
+ *@brief Clear decoder buffers (for seeking).
+ *@param avctx codec context
+ */
+static void flush(AVCodecContext *avctx)
+{
+    WMAProDecodeCtx *s = avctx->priv_data;
+    int i;
+    /** reset output buffer as a part of it is used during the windowing of a
+        new frame */
+    for (i = 0; i < s->num_channels; i++)
+        memset(s->channel[i].out, 0, s->samples_per_frame *
+               sizeof(*s->channel[i].out));
+    s->packet_loss = 1;
+}
+
+
+/**
+ *@brief wmapro decoder
+ */
+AVCodec wmapro_decoder = {
+    "wmapro",
+    CODEC_TYPE_AUDIO,
+    CODEC_ID_WMAPRO,
+    sizeof(WMAProDecodeCtx),
+    decode_init,
+    NULL,
+    decode_end,
+    decode_packet,
+    .flush= flush,
+    .long_name = NULL_IF_CONFIG_SMALL("Windows Media Audio 9 Professional"),
+};