* subframe in order to reconstruct the output samples.
*/
+#include "libavutil/float_dsp.h"
+#include "libavutil/intfloat.h"
+#include "libavutil/intreadwrite.h"
#include "avcodec.h"
#include "internal.h"
#include "get_bits.h"
#include "put_bits.h"
#include "wmaprodata.h"
-#include "dsputil.h"
-#include "fmtconvert.h"
#include "sinewin.h"
#include "wma.h"
+#include "wma_common.h"
/** current decoder limitations */
#define WMAPRO_MAX_CHANNELS 8 ///< max number of handled channels
#define MAX_FRAMESIZE 32768 ///< maximum compressed frame size
#define WMAPRO_BLOCK_MIN_BITS 6 ///< log2 of min block size
-#define WMAPRO_BLOCK_MAX_BITS 12 ///< log2 of max block size
+#define WMAPRO_BLOCK_MAX_BITS 13 ///< 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 - WMAPRO_BLOCK_MIN_BITS + 1) ///< possible block sizes
typedef struct WMAProDecodeCtx {
/* generic decoder variables */
AVCodecContext* avctx; ///< codec context for av_log
- DSPContext dsp; ///< accelerated DSP functions
- FmtConvertContext fmt_conv;
+ AVFloatDSPContext fdsp;
uint8_t frame_data[MAX_FRAMESIZE +
FF_INPUT_BUFFER_PADDING_SIZE];///< compressed frame data
PutBitContext pb; ///< context for filling the frame_data buffer
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
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?
*@brief helper function to print the most important members of the context
*@param s context
*/
-static void av_cold dump_context(WMAProDecodeCtx *s)
+static av_cold void 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("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);
+ PRINT("num channels", s->avctx->channels);
}
/**
WMAProDecodeCtx *s = avctx->priv_data;
uint8_t *edata_ptr = avctx->extradata;
unsigned int channel_mask;
- int i;
+ int i, bits;
int log2_max_num_subframes;
int num_possible_block_sizes;
s->avctx = avctx;
- dsputil_init(&s->dsp, avctx);
- ff_fmt_convert_init(&s->fmt_conv, avctx);
+ avpriv_float_dsp_init(&s->fdsp, avctx->flags & CODEC_FLAG_BITEXACT);
+
init_put_bits(&s->pb, s->frame_data, MAX_FRAMESIZE);
- avctx->sample_fmt = AV_SAMPLE_FMT_FLT;
+ avctx->sample_fmt = AV_SAMPLE_FMT_FLTP;
if (avctx->extradata_size >= 18) {
s->decode_flags = AV_RL16(edata_ptr+14);
} else {
av_log_ask_for_sample(avctx, "Unknown extradata size\n");
- return AVERROR_INVALIDDATA;
+ return AVERROR_PATCHWELCOME;
}
/** generic init */
s->len_prefix = (s->decode_flags & 0x40);
/** get frame len */
- s->samples_per_frame = 1 << ff_wma_get_frame_len_bits(avctx->sample_rate,
- 3, s->decode_flags);
+ bits = ff_wma_get_frame_len_bits(avctx->sample_rate, 3, s->decode_flags);
+ if (bits > WMAPRO_BLOCK_MAX_BITS) {
+ av_log_missing_feature(avctx, "14-bits block sizes", 1);
+ return AVERROR_PATCHWELCOME;
+ }
+ s->samples_per_frame = 1 << bits;
/** subframe info */
log2_max_num_subframes = ((s->decode_flags & 0x38) >> 3);
return AVERROR_INVALIDDATA;
}
- s->num_channels = avctx->channels;
+ if (s->avctx->sample_rate <= 0) {
+ av_log(avctx, AV_LOG_ERROR, "invalid sample rate\n");
+ return AVERROR_INVALIDDATA;
+ }
- if (s->num_channels < 0) {
- av_log(avctx, AV_LOG_ERROR, "invalid number of channels %d\n", s->num_channels);
+ if (avctx->channels < 0) {
+ av_log(avctx, AV_LOG_ERROR, "invalid number of channels %d\n",
+ avctx->channels);
return AVERROR_INVALIDDATA;
- } else if (s->num_channels > WMAPRO_MAX_CHANNELS) {
+ } else if (avctx->channels > WMAPRO_MAX_CHANNELS) {
av_log_ask_for_sample(avctx, "unsupported number of channels\n");
return AVERROR_PATCHWELCOME;
}
/** init previous block len */
- for (i = 0; i < s->num_channels; i++)
+ for (i = 0; i < avctx->channels; i++)
s->channel[i].prev_block_len = s->samples_per_frame;
/** extract lfe channel position */
dump_context(s);
avctx->channel_layout = channel_mask;
+
return 0;
}
*/
static int decode_tilehdr(WMAProDecodeCtx *s)
{
- uint16_t num_samples[WMAPRO_MAX_CHANNELS]; /**< sum of samples for all currently known subframes of a channel */
+ uint16_t num_samples[WMAPRO_MAX_CHANNELS] = { 0 };/**< 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 channels_for_cur_subframe = s->avctx->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
+ * while loop when always the minimum amount of 128 samples is subtracted
* 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++)
+ for (c = 0; c < s->avctx->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;
int subframe_len;
/** check which channels contain the subframe */
- for (c = 0; c < s->num_channels; c++) {
+ for (c = 0; c < s->avctx->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))
/** 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++) {
+ for (c = 0; c < s->avctx->channels; c++) {
WMAProChannelCtx* chan = &s->channel[c];
if (contains_subframe[c]) {
}
} while (min_channel_len < s->samples_per_frame);
- for (c = 0; c < s->num_channels; c++) {
+ for (c = 0; c < s->avctx->channels; c++) {
int i;
int offset = 0;
for (i = 0; i < s->channel[c].num_subframes; i++) {
int i;
int offset = 0;
int8_t rotation_offset[WMAPRO_MAX_CHANNELS * WMAPRO_MAX_CHANNELS];
- memset(chgroup->decorrelation_matrix, 0, s->num_channels *
- s->num_channels * sizeof(*chgroup->decorrelation_matrix));
+ memset(chgroup->decorrelation_matrix, 0, s->avctx->channels *
+ s->avctx->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);
/** in the one channel case channel transforms are pointless */
s->num_chgroups = 0;
- if (s->num_channels > 1) {
+ if (s->avctx->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;
+ return AVERROR_PATCHWELCOME;
}
for (s->num_chgroups = 0; remaining_channels &&
}
} else {
chgroup->transform = 1;
- if (s->num_channels == 2) {
+ if (s->avctx->channels == 2) {
chgroup->decorrelation_matrix[0] = 1.0;
chgroup->decorrelation_matrix[1] = -1.0;
chgroup->decorrelation_matrix[2] = 1.0;
/* Integers 0..15 as single-precision floats. The table saves a
costly int to float conversion, and storing the values as
integers allows fast sign-flipping. */
- static const int fval_tab[16] = {
+ static const uint32_t fval_tab[16] = {
0x00000000, 0x3f800000, 0x40000000, 0x40400000,
0x40800000, 0x40a00000, 0x40c00000, 0x40e00000,
0x41000000, 0x41100000, 0x41200000, 0x41300000,
4 vector coded large values) */
while ((s->transmit_num_vec_coeffs || !rl_mode) &&
(cur_coeff + 3 < ci->num_vec_coeffs)) {
- int vals[4];
+ uint32_t vals[4];
int i;
unsigned int idx;
for (i = 0; i < 4; i += 2) {
idx = get_vlc2(&s->gb, vec2_vlc.table, VLCBITS, VEC2MAXDEPTH);
if (idx == HUFF_VEC2_SIZE - 1) {
- int v0, v1;
+ uint32_t v0, v1;
v0 = get_vlc2(&s->gb, vec1_vlc.table, VLCBITS, VEC1MAXDEPTH);
if (v0 == HUFF_VEC1_SIZE - 1)
v0 += ff_wma_get_large_val(&s->gb);
v1 = get_vlc2(&s->gb, vec1_vlc.table, VLCBITS, VEC1MAXDEPTH);
if (v1 == HUFF_VEC1_SIZE - 1)
v1 += ff_wma_get_large_val(&s->gb);
- ((float*)vals)[i ] = v0;
- ((float*)vals)[i+1] = v1;
+ vals[i ] = av_float2int(v0);
+ vals[i+1] = av_float2int(v1);
} else {
vals[i] = fval_tab[symbol_to_vec2[idx] >> 4 ];
vals[i+1] = fval_tab[symbol_to_vec2[idx] & 0xF];
/** decode sign */
for (i = 0; i < 4; i++) {
if (vals[i]) {
- int sign = get_bits1(&s->gb) - 1;
- *(uint32_t*)&ci->coeffs[cur_coeff] = vals[i] ^ sign<<31;
+ uint32_t sign = get_bits1(&s->gb) - 1;
+ AV_WN32A(&ci->coeffs[cur_coeff], vals[i] ^ sign << 31);
num_zeros = 0;
} else {
ci->coeffs[cur_coeff] = 0;
(*ch)[y] = sum;
}
}
- } else if (s->num_channels == 2) {
+ } else if (s->avctx->channels == 2) {
int len = FFMIN(sfb[1], s->subframe_len) - sfb[0];
- s->dsp.vector_fmul_scalar(ch_data[0] + sfb[0],
- ch_data[0] + sfb[0],
- 181.0 / 128, len);
- s->dsp.vector_fmul_scalar(ch_data[1] + sfb[0],
- ch_data[1] + sfb[0],
- 181.0 / 128, len);
+ s->fdsp.vector_fmul_scalar(ch_data[0] + sfb[0],
+ ch_data[0] + sfb[0],
+ 181.0 / 128, len);
+ s->fdsp.vector_fmul_scalar(ch_data[1] + sfb[0],
+ ch_data[1] + sfb[0],
+ 181.0 / 128, len);
}
}
}
winlen >>= 1;
- s->dsp.vector_fmul_window(start, start, start + winlen,
- window, winlen);
+ s->fdsp.vector_fmul_window(start, start, start + winlen,
+ window, winlen);
s->channel[c].prev_block_len = s->subframe_len;
}
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 total_samples = s->samples_per_frame * s->avctx->channels;
int transmit_coeffs = 0;
int cur_subwoofer_cutoff;
== the next block of the channel with the smallest number of
decoded samples
*/
- for (i = 0; i < s->num_channels; i++) {
+ for (i = 0; i < s->avctx->channels; i++) {
s->channel[i].grouped = 0;
if (offset > s->channel[i].decoded_samples) {
offset = s->channel[i].decoded_samples;
/** 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++) {
+ for (i = 0; i < s->avctx->channels; i++) {
const int cur_subframe = s->channel[i].cur_subframe;
- /** substract already processed samples */
+ /** subtract already processed samples */
total_samples -= s->channel[i].decoded_samples;
/** and count if there are multiple subframes that match our profile */
/** 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;
+ return AVERROR_PATCHWELCOME;
}
int num_bits = av_log2((s->subframe_len + 3)/4) + 1;
for (i = 0; i < s->channels_for_cur_subframe; i++) {
int c = s->channel_indexes_for_cur_subframe[i];
- s->channel[c].num_vec_coeffs = get_bits(&s->gb, num_bits) << 2;
+ int num_vec_coeffs = get_bits(&s->gb, num_bits) << 2;
+ if (num_vec_coeffs > WMAPRO_BLOCK_MAX_SIZE) {
+ av_log(s->avctx, AV_LOG_ERROR, "num_vec_coeffs %d is too large\n", num_vec_coeffs);
+ return AVERROR_INVALIDDATA;
+ }
+ s->channel[c].num_vec_coeffs = num_vec_coeffs;
}
} else {
for (i = 0; i < s->channels_for_cur_subframe; i++) {
s->channel[c].scale_factor_step;
const float quant = pow(10.0, exp / 20.0);
int start = s->cur_sfb_offsets[b];
- s->dsp.vector_fmul_scalar(s->tmp + start,
- s->channel[c].coeffs + start,
- quant, end - start);
+ s->fdsp.vector_fmul_scalar(s->tmp + start,
+ s->channel[c].coeffs + start,
+ quant, end - start);
}
/** apply imdct (imdct_half == DCTIV with reverse) */
*@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)
+static int decode_frame(WMAProDecodeCtx *s, AVFrame *frame, int *got_frame_ptr)
{
+ AVCodecContext *avctx = s->avctx;
GetBitContext* gb = &s->gb;
int more_frames = 0;
int len = 0;
- int i;
- const float *out_ptr[WMAPRO_MAX_CHANNELS];
-
- /** check for potential output buffer overflow */
- if (s->num_channels * s->samples_per_frame > s->samples_end - s->samples) {
- /** return an error if no frame could be decoded at all */
- av_log(s->avctx, AV_LOG_ERROR,
- "not enough space for the output samples\n");
- s->packet_loss = 1;
- return 0;
- }
+ int i, ret;
/** get frame length */
if (s->len_prefix)
}
/** read postproc transform */
- if (s->num_channels > 1 && get_bits1(gb)) {
+ if (s->avctx->channels > 1 && get_bits1(gb)) {
if (get_bits1(gb)) {
- for (i = 0; i < s->num_channels * s->num_channels; i++)
+ for (i = 0; i < avctx->channels * avctx->channels; i++)
skip_bits(gb, 4);
}
}
/** reset subframe states */
s->parsed_all_subframes = 0;
- for (i = 0; i < s->num_channels; i++) {
+ for (i = 0; i < avctx->channels; i++) {
s->channel[i].decoded_samples = 0;
s->channel[i].cur_subframe = 0;
s->channel[i].reuse_sf = 0;
}
}
- /** interleave samples and write them to the output buffer */
- for (i = 0; i < s->num_channels; i++)
- out_ptr[i] = s->channel[i].out;
- s->fmt_conv.float_interleave(s->samples, out_ptr, s->samples_per_frame,
- s->num_channels);
+ /* get output buffer */
+ frame->nb_samples = s->samples_per_frame;
+ if ((ret = ff_get_buffer(avctx, frame, 0)) < 0) {
+ av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
+ s->packet_loss = 1;
+ return 0;
+ }
+
+ /** copy samples to the output buffer */
+ for (i = 0; i < avctx->channels; i++)
+ memcpy(frame->extended_data[i], s->channel[i].out,
+ s->samples_per_frame * sizeof(*s->channel[i].out));
- for (i = 0; i < s->num_channels; i++) {
+ for (i = 0; i < avctx->channels; i++) {
/** 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],
if (s->skip_frame) {
s->skip_frame = 0;
- } else
- s->samples += s->num_channels * s->samples_per_frame;
+ *got_frame_ptr = 0;
+ } else {
+ *got_frame_ptr = 1;
+ }
if (s->len_prefix) {
if (len != (get_bits_count(gb) - s->frame_offset) + 2) {
*@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
+ *@param append decides whether to reset the buffer or not
*/
static void save_bits(WMAProDecodeCtx *s, GetBitContext* gb, int len,
int append)
*@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)
+static int decode_packet(AVCodecContext *avctx, void *data,
+ int *got_frame_ptr, AVPacket* avpkt)
{
WMAProDecodeCtx *s = avctx->priv_data;
GetBitContext* gb = &s->pgb;
int num_bits_prev_frame;
int packet_sequence_number;
- s->samples = data;
- s->samples_end = (float*)((int8_t*)data + *data_size);
- *data_size = 0;
+ *got_frame_ptr = 0;
if (s->packet_done || s->packet_loss) {
s->packet_done = 0;
/** decode the cross packet frame if it is valid */
if (!s->packet_loss)
- decode_frame(s);
+ decode_frame(s, data, got_frame_ptr);
} else if (s->num_saved_bits - s->frame_offset) {
av_dlog(avctx, "ignoring %x previously saved bits\n",
s->num_saved_bits - s->frame_offset);
(frame_size = show_bits(gb, s->log2_frame_size)) &&
frame_size <= remaining_bits(s, gb)) {
save_bits(s, gb, frame_size, 0);
- s->packet_done = !decode_frame(s);
+ s->packet_done = !decode_frame(s, data, got_frame_ptr);
} else if (!s->len_prefix
&& s->num_saved_bits > get_bits_count(&s->gb)) {
/** when the frames do not have a length prefix, we don't know
therefore we save the incoming packet first, then we append
the "previous frame" data from the next packet so that
we get a buffer that only contains full frames */
- s->packet_done = !decode_frame(s);
+ s->packet_done = !decode_frame(s, data, got_frame_ptr);
} else
s->packet_done = 1;
}
save_bits(s, gb, remaining_bits(s, gb), 0);
}
- *data_size = (int8_t *)s->samples - (int8_t *)data;
s->packet_offset = get_bits_count(gb) & 7;
+ if (s->packet_loss)
+ return AVERROR_INVALIDDATA;
- return (s->packet_loss) ? AVERROR_INVALIDDATA : get_bits_count(gb) >> 3;
+ return get_bits_count(gb) >> 3;
}
/**
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++)
+ for (i = 0; i < avctx->channels; i++)
memset(s->channel[i].out, 0, s->samples_per_frame *
sizeof(*s->channel[i].out));
s->packet_loss = 1;
AVCodec ff_wmapro_decoder = {
.name = "wmapro",
.type = AVMEDIA_TYPE_AUDIO,
- .id = CODEC_ID_WMAPRO,
+ .id = AV_CODEC_ID_WMAPRO,
.priv_data_size = sizeof(WMAProDecodeCtx),
.init = decode_init,
.close = decode_end,
.decode = decode_packet,
- .capabilities = CODEC_CAP_SUBFRAMES,
- .flush= flush,
- .long_name = NULL_IF_CONFIG_SMALL("Windows Media Audio 9 Professional"),
+ .capabilities = CODEC_CAP_SUBFRAMES | CODEC_CAP_DR1,
+ .flush = flush,
+ .long_name = NULL_IF_CONFIG_SMALL("Windows Media Audio 9 Professional"),
+ .sample_fmts = (const enum AVSampleFormat[]) { AV_SAMPLE_FMT_FLTP,
+ AV_SAMPLE_FMT_NONE },
};