Commit d6aa052d authored by Ramiro Polla's avatar Ramiro Polla
Browse files

mlpdec: Split filter parameters from context into their own struct.

Originally committed as revision 14716 to svn://svn.ffmpeg.org/ffmpeg/trunk
parent 0d84a8f6
......@@ -139,6 +139,20 @@ typedef struct SubStream {
} SubStream;
#define FIR 0
#define IIR 1
/** filter data */
typedef struct {
//! number of taps in filter
uint8_t order;
//! Right shift to apply to output of filter.
uint8_t shift;
int32_t coeff[MAX_FILTER_ORDER];
int32_t state[MAX_FILTER_ORDER];
} FilterParams;
typedef struct MLPDecodeContext {
AVCodecContext *avctx;
......@@ -158,18 +172,7 @@ typedef struct MLPDecodeContext {
SubStream substream[MAX_SUBSTREAMS];
//@{
/** filter data */
#define FIR 0
#define IIR 1
//! number of taps in filter
uint8_t filter_order[MAX_CHANNELS][NUM_FILTERS];
//! Right shift to apply to output of filter.
uint8_t filter_shift[MAX_CHANNELS][NUM_FILTERS];
int32_t filter_coeff[MAX_CHANNELS][NUM_FILTERS][MAX_FILTER_ORDER];
int32_t filter_state[MAX_CHANNELS][NUM_FILTERS][MAX_FILTER_ORDER];
//@}
FilterParams filter_params[MAX_CHANNELS][NUM_FILTERS];
//@{
/** sample data coding information */
......@@ -520,10 +523,10 @@ static int read_restart_header(MLPDecodeContext *m, GetBitContext *gbp,
memset(s->quant_step_size, 0, sizeof(s->quant_step_size));
for (ch = s->min_channel; ch <= s->max_channel; ch++) {
m->filter_order[ch][FIR] = 0;
m->filter_order[ch][IIR] = 0;
m->filter_shift[ch][FIR] = 0;
m->filter_shift[ch][IIR] = 0;
m->filter_params[ch][FIR].order = 0;
m->filter_params[ch][IIR].order = 0;
m->filter_params[ch][FIR].shift = 0;
m->filter_params[ch][IIR].shift = 0;
/* Default audio coding is 24-bit raw PCM. */
m->huff_offset [ch] = 0;
......@@ -544,6 +547,7 @@ static int read_restart_header(MLPDecodeContext *m, GetBitContext *gbp,
static int read_filter_params(MLPDecodeContext *m, GetBitContext *gbp,
unsigned int channel, unsigned int filter)
{
FilterParams *fp = &m->filter_params[channel][filter];
const char fchar = filter ? 'I' : 'F';
int i, order;
......@@ -557,12 +561,12 @@ static int read_filter_params(MLPDecodeContext *m, GetBitContext *gbp,
fchar, order, MAX_FILTER_ORDER);
return -1;
}
m->filter_order[channel][filter] = order;
fp->order = order;
if (order > 0) {
int coeff_bits, coeff_shift;
m->filter_shift[channel][filter] = get_bits(gbp, 4);
fp->shift = get_bits(gbp, 4);
coeff_bits = get_bits(gbp, 5);
coeff_shift = get_bits(gbp, 3);
......@@ -580,7 +584,7 @@ static int read_filter_params(MLPDecodeContext *m, GetBitContext *gbp,
}
for (i = 0; i < order; i++)
m->filter_coeff[channel][filter][i] =
fp->coeff[i] =
get_sbits(gbp, coeff_bits) << coeff_shift;
if (get_bits1(gbp)) {
......@@ -598,7 +602,7 @@ static int read_filter_params(MLPDecodeContext *m, GetBitContext *gbp,
/* TODO: Check validity of state data. */
for (i = 0; i < order; i++)
m->filter_state[channel][filter][i] =
fp->state[i] =
get_sbits(gbp, state_bits) << state_shift;
}
}
......@@ -689,6 +693,9 @@ static int read_decoding_params(MLPDecodeContext *m, GetBitContext *gbp,
for (ch = s->min_channel; ch <= s->max_channel; ch++)
if (get_bits1(gbp)) {
FilterParams *fir = &m->filter_params[ch][FIR];
FilterParams *iir = &m->filter_params[ch][IIR];
if (s->param_presence_flags & PARAM_FIR)
if (get_bits1(gbp))
if (read_filter_params(m, gbp, ch, FIR) < 0)
......@@ -699,8 +706,8 @@ static int read_decoding_params(MLPDecodeContext *m, GetBitContext *gbp,
if (read_filter_params(m, gbp, ch, IIR) < 0)
return -1;
if (m->filter_order[ch][FIR] && m->filter_order[ch][IIR] &&
m->filter_shift[ch][FIR] != m->filter_shift[ch][IIR]) {
if (fir->order && iir->order &&
fir->shift != iir->shift) {
av_log(m->avctx, AV_LOG_ERROR,
"FIR and IIR filters must use the same precision.\n");
return -1;
......@@ -710,8 +717,8 @@ static int read_decoding_params(MLPDecodeContext *m, GetBitContext *gbp,
* FIR filter is considered. If only the IIR filter is employed,
* the FIR filter precision is set to that of the IIR filter, so
* that the filtering code can use it. */
if (!m->filter_order[ch][FIR] && m->filter_order[ch][IIR])
m->filter_shift[ch][FIR] = m->filter_shift[ch][IIR];
if (!fir->order && iir->order)
fir->shift = iir->shift;
if (s->param_presence_flags & PARAM_HUFFOFFSET)
if (get_bits1(gbp))
......@@ -738,14 +745,16 @@ static void filter_channel(MLPDecodeContext *m, unsigned int substr,
{
SubStream *s = &m->substream[substr];
int32_t filter_state_buffer[NUM_FILTERS][MAX_BLOCKSIZE + MAX_FILTER_ORDER];
unsigned int filter_shift = m->filter_shift[channel][FIR];
FilterParams *fp[NUM_FILTERS] = { &m->filter_params[channel][FIR],
&m->filter_params[channel][IIR], };
unsigned int filter_shift = fp[FIR]->shift;
int32_t mask = MSB_MASK(s->quant_step_size[channel]);
int index = MAX_BLOCKSIZE;
int j, i;
for (j = 0; j < NUM_FILTERS; j++) {
memcpy(& filter_state_buffer [j][MAX_BLOCKSIZE],
&m->filter_state[channel][j][0],
&fp[j]->state[0],
MAX_FILTER_ORDER * sizeof(int32_t));
}
......@@ -758,9 +767,9 @@ static void filter_channel(MLPDecodeContext *m, unsigned int substr,
/* TODO: Move this code to DSPContext? */
for (j = 0; j < NUM_FILTERS; j++)
for (order = 0; order < m->filter_order[channel][j]; order++)
for (order = 0; order < fp[j]->order; order++)
accum += (int64_t)filter_state_buffer[j][index + order] *
m->filter_coeff[channel][j][order];
fp[j]->coeff[order];
accum = accum >> filter_shift;
result = (accum + residual) & mask;
......@@ -774,7 +783,7 @@ static void filter_channel(MLPDecodeContext *m, unsigned int substr,
}
for (j = 0; j < NUM_FILTERS; j++) {
memcpy(&m->filter_state[channel][j][0],
memcpy(&fp[j]->state[0],
& filter_state_buffer [j][index],
MAX_FILTER_ORDER * sizeof(int32_t));
}
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment