blob: 0a2d4206a9ca76305cbd296c00425cfb480f416d
1 | /* |
2 | * Copyright (c) 2012 Michael Niedermayer |
3 | * |
4 | * This file is part of FFmpeg. |
5 | * |
6 | * FFmpeg is free software; you can redistribute it and/or |
7 | * modify it under the terms of the GNU Lesser General Public |
8 | * License as published by the Free Software Foundation; either |
9 | * version 2.1 of the License, or (at your option) any later version. |
10 | * |
11 | * FFmpeg is distributed in the hope that it will be useful, |
12 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
13 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
14 | * Lesser General Public License for more details. |
15 | * |
16 | * You should have received a copy of the GNU Lesser General Public |
17 | * License along with FFmpeg; if not, write to the Free Software |
18 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
19 | */ |
20 | |
21 | /** |
22 | * @file |
23 | * audio pad filter. |
24 | * |
25 | * Based on af_aresample.c |
26 | */ |
27 | |
28 | #include "libavutil/avstring.h" |
29 | #include "libavutil/channel_layout.h" |
30 | #include "libavutil/opt.h" |
31 | #include "libavutil/samplefmt.h" |
32 | #include "libavutil/avassert.h" |
33 | #include "avfilter.h" |
34 | #include "audio.h" |
35 | #include "internal.h" |
36 | |
37 | typedef struct { |
38 | const AVClass *class; |
39 | int64_t next_pts; |
40 | |
41 | int packet_size; |
42 | int64_t pad_len, pad_len_left; |
43 | int64_t whole_len, whole_len_left; |
44 | } APadContext; |
45 | |
46 | #define OFFSET(x) offsetof(APadContext, x) |
47 | #define A AV_OPT_FLAG_AUDIO_PARAM|AV_OPT_FLAG_FILTERING_PARAM |
48 | |
49 | static const AVOption apad_options[] = { |
50 | { "packet_size", "set silence packet size", OFFSET(packet_size), AV_OPT_TYPE_INT, { .i64 = 4096 }, 0, INT_MAX, A }, |
51 | { "pad_len", "set number of samples of silence to add", OFFSET(pad_len), AV_OPT_TYPE_INT64, { .i64 = -1 }, -1, INT64_MAX, A }, |
52 | { "whole_len", "set minimum target number of samples in the audio stream", OFFSET(whole_len), AV_OPT_TYPE_INT64, { .i64 = -1 }, -1, INT64_MAX, A }, |
53 | { NULL } |
54 | }; |
55 | |
56 | AVFILTER_DEFINE_CLASS(apad); |
57 | |
58 | static av_cold int init(AVFilterContext *ctx) |
59 | { |
60 | APadContext *s = ctx->priv; |
61 | |
62 | s->next_pts = AV_NOPTS_VALUE; |
63 | if (s->whole_len >= 0 && s->pad_len >= 0) { |
64 | av_log(ctx, AV_LOG_ERROR, "Both whole and pad length are set, this is not possible\n"); |
65 | return AVERROR(EINVAL); |
66 | } |
67 | s->pad_len_left = s->pad_len; |
68 | s->whole_len_left = s->whole_len; |
69 | |
70 | return 0; |
71 | } |
72 | |
73 | static int filter_frame(AVFilterLink *inlink, AVFrame *frame) |
74 | { |
75 | AVFilterContext *ctx = inlink->dst; |
76 | APadContext *s = ctx->priv; |
77 | |
78 | if (s->whole_len >= 0) { |
79 | s->whole_len_left = FFMAX(s->whole_len_left - frame->nb_samples, 0); |
80 | av_log(ctx, AV_LOG_DEBUG, |
81 | "n_out:%d whole_len_left:%"PRId64"\n", frame->nb_samples, s->whole_len_left); |
82 | } |
83 | |
84 | s->next_pts = frame->pts + av_rescale_q(frame->nb_samples, (AVRational){1, inlink->sample_rate}, inlink->time_base); |
85 | return ff_filter_frame(ctx->outputs[0], frame); |
86 | } |
87 | |
88 | static int request_frame(AVFilterLink *outlink) |
89 | { |
90 | AVFilterContext *ctx = outlink->src; |
91 | APadContext *s = ctx->priv; |
92 | int ret; |
93 | |
94 | ret = ff_request_frame(ctx->inputs[0]); |
95 | |
96 | if (ret == AVERROR_EOF && !ctx->is_disabled) { |
97 | int n_out = s->packet_size; |
98 | AVFrame *outsamplesref; |
99 | |
100 | if (s->whole_len >= 0 && s->pad_len < 0) { |
101 | s->pad_len = s->pad_len_left = s->whole_len_left; |
102 | } |
103 | if (s->pad_len >=0 || s->whole_len >= 0) { |
104 | n_out = FFMIN(n_out, s->pad_len_left); |
105 | s->pad_len_left -= n_out; |
106 | av_log(ctx, AV_LOG_DEBUG, |
107 | "padding n_out:%d pad_len_left:%"PRId64"\n", n_out, s->pad_len_left); |
108 | } |
109 | |
110 | if (!n_out) |
111 | return AVERROR_EOF; |
112 | |
113 | outsamplesref = ff_get_audio_buffer(outlink, n_out); |
114 | if (!outsamplesref) |
115 | return AVERROR(ENOMEM); |
116 | |
117 | av_assert0(outsamplesref->sample_rate == outlink->sample_rate); |
118 | av_assert0(outsamplesref->nb_samples == n_out); |
119 | |
120 | av_samples_set_silence(outsamplesref->extended_data, 0, |
121 | n_out, |
122 | av_frame_get_channels(outsamplesref), |
123 | outsamplesref->format); |
124 | |
125 | outsamplesref->pts = s->next_pts; |
126 | if (s->next_pts != AV_NOPTS_VALUE) |
127 | s->next_pts += av_rescale_q(n_out, (AVRational){1, outlink->sample_rate}, outlink->time_base); |
128 | |
129 | return ff_filter_frame(outlink, outsamplesref); |
130 | } |
131 | return ret; |
132 | } |
133 | |
134 | static const AVFilterPad apad_inputs[] = { |
135 | { |
136 | .name = "default", |
137 | .type = AVMEDIA_TYPE_AUDIO, |
138 | .filter_frame = filter_frame, |
139 | }, |
140 | { NULL } |
141 | }; |
142 | |
143 | static const AVFilterPad apad_outputs[] = { |
144 | { |
145 | .name = "default", |
146 | .request_frame = request_frame, |
147 | .type = AVMEDIA_TYPE_AUDIO, |
148 | }, |
149 | { NULL } |
150 | }; |
151 | |
152 | AVFilter ff_af_apad = { |
153 | .name = "apad", |
154 | .description = NULL_IF_CONFIG_SMALL("Pad audio with silence."), |
155 | .init = init, |
156 | .priv_size = sizeof(APadContext), |
157 | .inputs = apad_inputs, |
158 | .outputs = apad_outputs, |
159 | .priv_class = &apad_class, |
160 | .flags = AVFILTER_FLAG_SUPPORT_TIMELINE_INTERNAL, |
161 | }; |
162 |