1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
|
/*
* Copyright (c) 2011 Nicolas George <nicolas.george@normalesup.org>
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* @file
* Audio merging filter
*/
#include "libswresample/swresample.h" // only for SWR_CH_MAX
#include "avfilter.h"
#include "audio.h"
#include "bufferqueue.h"
#include "internal.h"
typedef struct {
int nb_in_ch[2]; /**< number of channels for each input */
int route[SWR_CH_MAX]; /**< channels routing, see copy_samples */
int bps;
struct amerge_input {
struct FFBufQueue queue;
int nb_samples;
int pos;
} in[2];
} AMergeContext;
static av_cold void uninit(AVFilterContext *ctx)
{
AMergeContext *am = ctx->priv;
int i;
for (i = 0; i < 2; i++)
ff_bufqueue_discard_all(&am->in[i].queue);
}
static int query_formats(AVFilterContext *ctx)
{
AMergeContext *am = ctx->priv;
int64_t inlayout[2], outlayout;
AVFilterFormats *formats;
AVFilterChannelLayouts *layouts;
int i;
for (i = 0; i < 2; i++) {
if (!ctx->inputs[i]->in_channel_layouts ||
!ctx->inputs[i]->in_channel_layouts->nb_channel_layouts) {
av_log(ctx, AV_LOG_ERROR,
"No channel layout for input %d\n", i + 1);
return AVERROR(EINVAL);
}
inlayout[i] = ctx->inputs[i]->in_channel_layouts->channel_layouts[0];
if (ctx->inputs[i]->in_channel_layouts->nb_channel_layouts > 1) {
char buf[256];
av_get_channel_layout_string(buf, sizeof(buf), 0, inlayout[i]);
av_log(ctx, AV_LOG_INFO, "Using \"%s\" for input %d\n", buf, i + 1);
}
am->nb_in_ch[i] = av_get_channel_layout_nb_channels(inlayout[i]);
}
if (am->nb_in_ch[0] + am->nb_in_ch[1] > SWR_CH_MAX) {
av_log(ctx, AV_LOG_ERROR, "Too many channels (max %d)\n", SWR_CH_MAX);
return AVERROR(EINVAL);
}
if (inlayout[0] & inlayout[1]) {
av_log(ctx, AV_LOG_WARNING,
"Inputs overlap: output layout will be meaningless\n");
for (i = 0; i < am->nb_in_ch[0] + am->nb_in_ch[1]; i++)
am->route[i] = i;
outlayout = av_get_default_channel_layout(am->nb_in_ch[0] +
am->nb_in_ch[1]);
if (!outlayout)
outlayout = ((int64_t)1 << (am->nb_in_ch[0] + am->nb_in_ch[1])) - 1;
} else {
int *route[2] = { am->route, am->route + am->nb_in_ch[0] };
int c, out_ch_number = 0;
outlayout = inlayout[0] | inlayout[1];
for (c = 0; c < 64; c++)
for (i = 0; i < 2; i++)
if ((inlayout[i] >> c) & 1)
*(route[i]++) = out_ch_number++;
}
formats = avfilter_make_format_list(ff_packed_sample_fmts);
avfilter_set_common_sample_formats(ctx, formats);
for (i = 0; i < 2; i++) {
layouts = NULL;
ff_add_channel_layout(&layouts, inlayout[i]);
ff_channel_layouts_ref(layouts, &ctx->inputs[i]->out_channel_layouts);
}
layouts = NULL;
ff_add_channel_layout(&layouts, outlayout);
ff_channel_layouts_ref(layouts, &ctx->outputs[0]->in_channel_layouts);
ff_set_common_samplerates(ctx, ff_all_samplerates());
return 0;
}
static int config_output(AVFilterLink *outlink)
{
AVFilterContext *ctx = outlink->src;
AMergeContext *am = ctx->priv;
int64_t layout;
char name[3][256];
int i;
if (ctx->inputs[0]->sample_rate != ctx->inputs[1]->sample_rate) {
av_log(ctx, AV_LOG_ERROR,
"Inputs must have the same sample rate "
"(%"PRIi64" vs %"PRIi64")\n",
ctx->inputs[0]->sample_rate, ctx->inputs[1]->sample_rate);
return AVERROR(EINVAL);
}
am->bps = av_get_bytes_per_sample(ctx->outputs[0]->format);
outlink->sample_rate = ctx->inputs[0]->sample_rate;
outlink->time_base = ctx->inputs[0]->time_base;
for (i = 0; i < 3; i++) {
layout = (i < 2 ? ctx->inputs[i] : ctx->outputs[0])->channel_layout;
av_get_channel_layout_string(name[i], sizeof(name[i]), -1, layout);
}
av_log(ctx, AV_LOG_INFO,
"in1:%s + in2:%s -> out:%s\n", name[0], name[1], name[2]);
return 0;
}
static int request_frame(AVFilterLink *outlink)
{
AVFilterContext *ctx = outlink->src;
AMergeContext *am = ctx->priv;
int i, ret;
for (i = 0; i < 2; i++)
if (!am->in[i].nb_samples)
if ((ret = avfilter_request_frame(ctx->inputs[i])) < 0)
return ret;
return 0;
}
/**
* Copy samples from two input streams to one output stream.
* @param nb_in_ch number of channels in each input stream
* @param route routing values;
* input channel i goes to output channel route[i];
* i < nb_in_ch[0] are the channels from the first output;
* i >= nb_in_ch[0] are the channels from the second output
* @param ins pointer to the samples of each inputs, in packed format;
* will be left at the end of the copied samples
* @param outs pointer to the samples of the output, in packet format;
* must point to a buffer big enough;
* will be left at the end of the copied samples
* @param ns number of samples to copy
* @param bps bytes per sample
*/
static inline void copy_samples(int nb_in_ch[2], int *route, uint8_t *ins[2],
uint8_t **outs, int ns, int bps)
{
int *route_cur;
int i, c;
while (ns--) {
route_cur = route;
for (i = 0; i < 2; i++) {
for (c = 0; c < nb_in_ch[i]; c++) {
memcpy((*outs) + bps * *(route_cur++), ins[i], bps);
ins[i] += bps;
}
}
*outs += (nb_in_ch[0] + nb_in_ch[1]) * bps;
}
}
static void filter_samples(AVFilterLink *inlink, AVFilterBufferRef *insamples)
{
AVFilterContext *ctx = inlink->dst;
AMergeContext *am = ctx->priv;
AVFilterLink *const outlink = ctx->outputs[0];
int input_number = inlink == ctx->inputs[1];
int nb_samples, ns, i;
AVFilterBufferRef *outbuf, *inbuf[2];
uint8_t *ins[2], *outs;
ff_bufqueue_add(ctx, &am->in[input_number].queue, insamples);
am->in[input_number].nb_samples += insamples->audio->nb_samples;
if (!am->in[!input_number].nb_samples)
return;
nb_samples = FFMIN(am->in[0].nb_samples,
am->in[1].nb_samples);
outbuf = ff_get_audio_buffer(ctx->outputs[0], AV_PERM_WRITE, nb_samples);
outs = outbuf->data[0];
for (i = 0; i < 2; i++) {
inbuf[i] = ff_bufqueue_peek(&am->in[i].queue, 0);
ins[i] = inbuf[i]->data[0] +
am->in[i].pos * am->nb_in_ch[i] * am->bps;
}
outbuf->pts = inbuf[0]->pts == AV_NOPTS_VALUE ? AV_NOPTS_VALUE :
inbuf[0]->pts +
av_rescale_q(am->in[0].pos,
(AVRational){ 1, ctx->inputs[0]->sample_rate },
ctx->outputs[0]->time_base);
avfilter_copy_buffer_ref_props(outbuf, inbuf[0]);
outbuf->audio->nb_samples = nb_samples;
outbuf->audio->channel_layout = outlink->channel_layout;
while (nb_samples) {
ns = nb_samples;
for (i = 0; i < 2; i++)
ns = FFMIN(ns, inbuf[i]->audio->nb_samples - am->in[i].pos);
/* Unroll the most common sample formats: speed +~350% for the loop,
+~13% overall (including two common decoders) */
switch (am->bps) {
case 1:
copy_samples(am->nb_in_ch, am->route, ins, &outs, ns, 1);
break;
case 2:
copy_samples(am->nb_in_ch, am->route, ins, &outs, ns, 2);
break;
case 4:
copy_samples(am->nb_in_ch, am->route, ins, &outs, ns, 4);
break;
default:
copy_samples(am->nb_in_ch, am->route, ins, &outs, ns, am->bps);
break;
}
nb_samples -= ns;
for (i = 0; i < 2; i++) {
am->in[i].nb_samples -= ns;
am->in[i].pos += ns;
if (am->in[i].pos == inbuf[i]->audio->nb_samples) {
am->in[i].pos = 0;
avfilter_unref_buffer(inbuf[i]);
ff_bufqueue_get(&am->in[i].queue);
inbuf[i] = ff_bufqueue_peek(&am->in[i].queue, 0);
ins[i] = inbuf[i] ? inbuf[i]->data[0] : NULL;
}
}
}
ff_filter_samples(ctx->outputs[0], outbuf);
}
AVFilter avfilter_af_amerge = {
.name = "amerge",
.description = NULL_IF_CONFIG_SMALL("Merge two audio streams into "
"a single multi-channel stream."),
.priv_size = sizeof(AMergeContext),
.uninit = uninit,
.query_formats = query_formats,
.inputs = (const AVFilterPad[]) {
{ .name = "in1",
.type = AVMEDIA_TYPE_AUDIO,
.filter_samples = filter_samples,
.min_perms = AV_PERM_READ | AV_PERM_PRESERVE, },
{ .name = "in2",
.type = AVMEDIA_TYPE_AUDIO,
.filter_samples = filter_samples,
.min_perms = AV_PERM_READ | AV_PERM_PRESERVE, },
{ .name = NULL }
},
.outputs = (const AVFilterPad[]) {
{ .name = "default",
.type = AVMEDIA_TYPE_AUDIO,
.config_props = config_output,
.request_frame = request_frame, },
{ .name = NULL }
},
};
|