@@ -23,7 +23,6 @@
#include <float.h> /* DBL_MAX */
-#include "libavutil/audio_fifo.h"
#include "libavutil/avassert.h"
#include "libavutil/opt.h"
#include "libavutil/timestamp.h"
@@ -94,8 +93,6 @@ typedef struct SilenceRemoveContext {
int64_t window_duration;
double sum;
- int threshold;
- int one_period;
int restart;
int64_t next_pts;
@@ -104,22 +101,20 @@ typedef struct SilenceRemoveContext {
double (*compute)(struct SilenceRemoveContext *s, AVFrame *frame, int ch, int offset);
void (*copy)(struct SilenceRemoveContext *s, AVFrame *out, AVFrame *in,
int ch, int out_offset, int in_offset);
-
- AVAudioFifo *fifo;
} SilenceRemoveContext;
#define OFFSET(x) offsetof(SilenceRemoveContext, x)
#define AF AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_AUDIO_PARAM
static const AVOption silenceremove_options[] = {
- { "start_periods", "set periods of silence parts to skip from start", OFFSET(start_periods), AV_OPT_TYPE_INT, {.i64=0}, 0, 9000, AF },
+ { "start_periods", NULL, OFFSET(start_periods), AV_OPT_TYPE_INT, {.i64=0}, 0, 9000, AF },
{ "start_duration", "set start duration of non-silence part", OFFSET(start_duration_opt), AV_OPT_TYPE_DURATION, {.i64=0}, 0, INT32_MAX, AF },
{ "start_threshold", "set threshold for start silence detection", OFFSET(start_threshold), AV_OPT_TYPE_DOUBLE, {.dbl=0}, 0, DBL_MAX, AF },
{ "start_silence", "set start duration of silence part to keep", OFFSET(start_silence_opt), AV_OPT_TYPE_DURATION, {.i64=0}, 0, INT32_MAX, AF },
{ "start_mode", "set which channel will trigger trimming from start", OFFSET(start_mode), AV_OPT_TYPE_INT, {.i64=T_ANY}, T_ANY, T_ALL, AF, "mode" },
{ "any", 0, 0, AV_OPT_TYPE_CONST, {.i64=T_ANY}, 0, 0, AF, "mode" },
{ "all", 0, 0, AV_OPT_TYPE_CONST, {.i64=T_ALL}, 0, 0, AF, "mode" },
- { "stop_periods", "set periods of silence parts to skip from end", OFFSET(stop_periods), AV_OPT_TYPE_INT, {.i64=0}, -9000, 9000, AF },
+ { "stop_periods", NULL, OFFSET(stop_periods), AV_OPT_TYPE_INT, {.i64=0}, -9000, 9000, AF },
{ "stop_duration", "set stop duration of non-silence part", OFFSET(stop_duration_opt), AV_OPT_TYPE_DURATION, {.i64=0}, 0, INT32_MAX, AF },
{ "stop_threshold", "set threshold for stop silence detection", OFFSET(stop_threshold), AV_OPT_TYPE_DOUBLE, {.dbl=0}, 0, DBL_MAX, AF },
{ "stop_silence", "set stop duration of silence part to keep", OFFSET(stop_silence_opt), AV_OPT_TYPE_DURATION, {.i64=0}, 0, INT32_MAX, AF },
@@ -435,7 +430,6 @@ static int config_input(AVFilterLink *inlink)
AVFilterContext *ctx = inlink->dst;
SilenceRemoveContext *s = ctx->priv;
- s->threshold = -1;
s->next_pts = AV_NOPTS_VALUE;
s->window_duration = av_rescale(s->window_duration_opt, inlink->sample_rate,
AV_TIME_BASE);
@@ -545,16 +539,12 @@ static int config_input(AVFilterLink *inlink)
return AVERROR_BUG;
}
- s->fifo = av_audio_fifo_alloc(inlink->format, inlink->channels, 1024);
- if (!s->fifo)
- return AVERROR(ENOMEM);
-
return 0;
}
static void flush(SilenceRemoveContext *s,
AVFrame *out, AVFilterLink *outlink,
- int *nb_samples_written, int flush_silence)
+ int *nb_samples_written, int *ret, int flush_silence)
{
AVFrame *silence;
@@ -566,18 +556,22 @@ static void flush(SilenceRemoveContext *s,
(AVRational){1, outlink->sample_rate},
outlink->time_base);
- av_audio_fifo_write(s->fifo, (void **)out->extended_data, out->nb_samples);
+ *ret = ff_filter_frame(outlink, out);
+ if (*ret < 0)
+ return;
*nb_samples_written = 0;
+ } else {
+ av_frame_free(&out);
}
- av_frame_free(&out);
-
if (s->stop_silence_end <= 0 || !flush_silence)
return;
silence = ff_get_audio_buffer(outlink, s->stop_silence_end);
- if (!silence)
+ if (!silence) {
+ *ret = AVERROR(ENOMEM);
return;
+ }
if (s->stop_silence_offset < s->stop_silence_end) {
av_samples_copy(silence->extended_data, s->stop_silence_hold->extended_data, 0,
@@ -601,8 +595,7 @@ static void flush(SilenceRemoveContext *s,
(AVRational){1, outlink->sample_rate},
outlink->time_base);
- av_audio_fifo_write(s->fifo, (void **)silence->extended_data, silence->nb_samples);
- av_frame_free(&silence);
+ *ret = ff_filter_frame(outlink, silence);
}
static int filter_frame(AVFilterLink *inlink, AVFrame *in)
@@ -610,8 +603,8 @@ static int filter_frame(AVFilterLink *inlink, AVFrame *in)
AVFilterContext *ctx = inlink->dst;
AVFilterLink *outlink = ctx->outputs[0];
SilenceRemoveContext *s = ctx->priv;
- int nbs, nb_samples_read, nb_samples_written;
int i, j, threshold, ret = 0;
+ int nbs, nb_samples_read, nb_samples_written;
AVFrame *out;
nb_samples_read = nb_samples_written = 0;
@@ -639,10 +632,6 @@ silence_trim:
}
}
- if (s->threshold >= 0)
- s->one_period = s->threshold != threshold;
- s->threshold = threshold;
-
if (threshold) {
for (j = 0; j < outlink->channels; j++) {
s->update(s, in, j, nb_samples_read);
@@ -656,8 +645,7 @@ silence_trim:
nb_samples_read++;
if (s->start_holdoff_end >= s->start_duration) {
- s->start_found_periods += s->one_period;
- if (s->start_found_periods >= s->start_periods) {
+ if (++s->start_found_periods >= s->start_periods) {
s->mode = SILENCE_TRIM_FLUSH;
goto silence_trim_flush;
}
@@ -731,8 +719,7 @@ silence_trim_flush:
s->start_holdoff_offset += nbs;
- av_audio_fifo_write(s->fifo, (void **)out->extended_data, out->nb_samples);
- av_frame_free(&out);
+ ret = ff_filter_frame(outlink, out);
if (s->start_holdoff_offset == s->start_holdoff_end) {
s->start_holdoff_offset = 0;
@@ -770,13 +757,9 @@ silence_copy:
}
}
- if (s->threshold >= 0)
- s->one_period = s->threshold != threshold;
- s->threshold = threshold;
-
if (threshold && s->stop_holdoff_end && !s->stop_silence) {
s->mode = SILENCE_COPY_FLUSH;
- flush(s, out, outlink, &nb_samples_written, 0);
+ flush(s, out, outlink, &nb_samples_written, &ret, 0);
goto silence_copy_flush;
} else if (threshold) {
for (j = 0; j < outlink->channels; j++) {
@@ -813,14 +796,13 @@ silence_copy:
s->stop_holdoff_end++;
if (s->stop_holdoff_end >= s->stop_duration) {
- s->stop_found_periods += s->one_period;
- if (s->stop_found_periods >= s->stop_periods) {
+ if (++s->stop_found_periods >= s->stop_periods) {
s->stop_holdoff_offset = 0;
s->stop_holdoff_end = 0;
if (!s->restart) {
s->mode = SILENCE_STOP;
- flush(s, out, outlink, &nb_samples_written, 1);
+ flush(s, out, outlink, &nb_samples_written, &ret, 1);
goto silence_stop;
} else {
s->stop_found_periods = 0;
@@ -831,17 +813,17 @@ silence_copy:
s->start_silence_end = 0;
clear_window(s);
s->mode = SILENCE_TRIM;
- flush(s, out, outlink, &nb_samples_written, 1);
+ flush(s, out, outlink, &nb_samples_written, &ret, 1);
goto silence_trim;
}
}
s->mode = SILENCE_COPY_FLUSH;
- flush(s, out, outlink, &nb_samples_written, 0);
+ flush(s, out, outlink, &nb_samples_written, &ret, 0);
goto silence_copy_flush;
}
}
}
- flush(s, out, outlink, &nb_samples_written, 0);
+ flush(s, out, outlink, &nb_samples_written, &ret, 0);
} else {
av_samples_copy(out->extended_data, in->extended_data,
nb_samples_written,
@@ -853,8 +835,7 @@ silence_copy:
(AVRational){1, outlink->sample_rate},
outlink->time_base);
- av_audio_fifo_write(s->fifo, (void **)out->extended_data, out->nb_samples);
- av_frame_free(&out);
+ ret = ff_filter_frame(outlink, out);
}
break;
@@ -881,8 +862,7 @@ silence_copy_flush:
(AVRational){1, outlink->sample_rate},
outlink->time_base);
- av_audio_fifo_write(s->fifo, (void **)out->extended_data, out->nb_samples);
- av_frame_free(&out);
+ ret = ff_filter_frame(outlink, out);
if (s->stop_holdoff_offset == s->stop_holdoff_end) {
s->stop_holdoff_offset = 0;
@@ -902,16 +882,6 @@ silence_stop:
av_frame_free(&in);
- if (av_audio_fifo_size(s->fifo) > 0) {
- out = ff_get_audio_buffer(outlink, av_audio_fifo_size(s->fifo));
- if (!out)
- return AVERROR(ENOMEM);
-
- av_audio_fifo_read(s->fifo, (void **)out->extended_data, out->nb_samples);
- out->pts = s->next_pts;
- ret = ff_filter_frame(outlink, out);
- }
-
return ret;
}
@@ -975,9 +945,6 @@ static av_cold void uninit(AVFilterContext *ctx)
av_frame_free(&s->stop_holdoff);
av_frame_free(&s->stop_silence_hold);
av_frame_free(&s->window);
-
- av_audio_fifo_free(s->fifo);
- s->fifo = NULL;
}
static const AVFilterPad silenceremove_inputs[] = {
This reverts commits 3b331468dae2e88ee6c87c257ac159ad662efcac and 8a42ee6697317d0a30438df5905dfc0247cd28e7. They broke the filter-silenceremove FATE-test, as they were pushed without updating the reference file. Worse, the new output is not bitexact across all arches, as the diff on aarch64 is different; see e.g. http://fate.ffmpeg.org/report.cgi?slot=aarch64-linux-clang-10&time=20210909013452 and http://fate.ffmpeg.org/report.cgi?time=20210909095918&slot=x86_64-archlinux-gcc-threads-16 So revert the commits for now until this can be properly addressed. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com> --- I hope that this patch needn't be applied and include it here mainly so that patchwork won't be red for the other patches of this patchset. libavfilter/af_silenceremove.c | 79 ++++++++++------------------------ 1 file changed, 23 insertions(+), 56 deletions(-)