#include "avfilter.h"
#include "internal.h"
+enum SilenceDetect {
+ D_PEAK,
+ D_RMS,
+};
+
+enum ThresholdMode {
+ T_ANY,
+ T_ALL,
+};
+
enum SilenceMode {
SILENCE_TRIM,
SILENCE_TRIM_FLUSH,
#define AF AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_AUDIO_PARAM
static const AVOption silenceremove_options[] = {
- { "start_periods", NULL, OFFSET(start_periods), AV_OPT_TYPE_INT, {.i64=0}, 0, 9000, AF },
- { "start_duration", NULL, OFFSET(start_duration_opt), AV_OPT_TYPE_DURATION, {.i64=0}, 0, INT32_MAX, AF },
- { "start_threshold", NULL, OFFSET(start_threshold), AV_OPT_TYPE_DOUBLE, {.dbl=0}, 0, DBL_MAX, AF },
- { "start_silence", NULL, OFFSET(start_silence_opt), AV_OPT_TYPE_DURATION, {.i64=0}, 0, INT32_MAX, AF },
- { "start_mode", NULL, OFFSET(start_mode), AV_OPT_TYPE_INT, {.i64=0}, 0, 1, AF, "mode" },
- { "any", 0, 0, AV_OPT_TYPE_CONST, {.i64=0}, 0, 0, AF, "mode" },
- { "all", 0, 0, AV_OPT_TYPE_CONST, {.i64=1}, 0, 0, AF, "mode" },
- { "stop_periods", NULL, OFFSET(stop_periods), AV_OPT_TYPE_INT, {.i64=0}, -9000, 9000, AF },
- { "stop_duration", NULL, OFFSET(stop_duration_opt), AV_OPT_TYPE_DURATION, {.i64=0}, 0, INT32_MAX, AF },
- { "stop_threshold", NULL, OFFSET(stop_threshold), AV_OPT_TYPE_DOUBLE, {.dbl=0}, 0, DBL_MAX, AF },
- { "stop_silence", NULL, OFFSET(stop_silence_opt), AV_OPT_TYPE_DURATION, {.i64=0}, 0, INT32_MAX, AF },
- { "stop_mode", NULL, OFFSET(stop_mode), AV_OPT_TYPE_INT, {.i64=0}, 0, 1, AF, "mode" },
- { "detection", NULL, OFFSET(detection), AV_OPT_TYPE_INT, {.i64=1}, 0, 1, AF, "detection" },
- { "peak", 0, 0, AV_OPT_TYPE_CONST, {.i64=0}, 0, 0, AF, "detection" },
- { "rms", 0, 0, AV_OPT_TYPE_CONST, {.i64=1}, 0, 0, AF, "detection" },
- { "window", NULL, OFFSET(window_ratio), AV_OPT_TYPE_DOUBLE, {.dbl=0.02}, 0, 10, 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", 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 },
+ { "stop_mode", "set which channel will trigger trimming from end", OFFSET(stop_mode), AV_OPT_TYPE_INT, {.i64=T_ANY}, T_ANY, T_ALL, AF, "mode" },
+ { "detection", "set how silence is detected", OFFSET(detection), AV_OPT_TYPE_INT, {.i64=D_RMS}, D_PEAK,D_RMS, AF, "detection" },
+ { "peak", "use absolute values of samples", 0, AV_OPT_TYPE_CONST, {.i64=D_PEAK},0, 0, AF, "detection" },
+ { "rms", "use squared values of samples", 0, AV_OPT_TYPE_CONST, {.i64=D_RMS}, 0, 0, AF, "detection" },
+ { "window", "set duration of window in seconds", OFFSET(window_ratio), AV_OPT_TYPE_DOUBLE, {.dbl=0.02}, 0, 10, AF },
{ NULL }
};
}
switch (s->detection) {
- case 0:
+ case D_PEAK:
s->update = update_peak;
s->compute = compute_peak;
break;
- case 1:
+ case D_RMS:
s->update = update_rms;
s->compute = compute_rms;
break;
AVFilterContext *ctx = inlink->dst;
SilenceRemoveContext *s = ctx->priv;
+ s->next_pts = AV_NOPTS_VALUE;
s->window_size = FFMAX((inlink->sample_rate * s->window_ratio), 1) * inlink->channels;
s->window = av_malloc_array(s->window_size, sizeof(*s->window));
if (!s->window)
nb_samples_read = nb_samples_written = 0;
+ if (s->next_pts == AV_NOPTS_VALUE)
+ s->next_pts = in->pts;
+
switch (s->mode) {
case SILENCE_TRIM:
silence_trim:
break;
for (i = 0; i < nbs; i++) {
- if (s->start_mode) {
+ if (s->start_mode == T_ANY) {
threshold = 0;
for (j = 0; j < outlink->channels; j++) {
threshold |= s->compute(s, ibuf[j]) > s->start_threshold;
if (s->stop_periods) {
for (i = 0; i < nbs; i++) {
- if (s->stop_mode) {
+ if (s->stop_mode == T_ANY) {
threshold = 0;
for (j = 0; j < outlink->channels; j++) {
threshold |= s->compute(s, ibuf[j]) > s->stop_threshold;
{ NULL }
};
-AVFilter ff_af_silenceremove = {
+const AVFilter ff_af_silenceremove = {
.name = "silenceremove",
.description = NULL_IF_CONFIG_SMALL("Remove silence."),
.priv_size = sizeof(SilenceRemoveContext),