1/*
2 * Interface to libtwolame for mp2 encoding
3 * Copyright (c) 2012 Paul B Mahol
4 *
5 * This file is part of FFmpeg.
6 *
7 * FFmpeg is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU Lesser General Public
9 * License as published by the Free Software Foundation; either
10 * version 2.1 of the License, or (at your option) any later version.
11 *
12 * FFmpeg is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
15 * Lesser General Public License for more details.
16 *
17 * You should have received a copy of the GNU Lesser General Public
18 * License along with FFmpeg; if not, write to the Free Software
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 */
21
22/**
23 * @file
24 * Interface to libtwolame for mp2 encoding.
25 */
26
27#include <twolame.h>
28
29#include "libavutil/channel_layout.h"
30#include "libavutil/common.h"
31#include "libavutil/opt.h"
32
33#include "avcodec.h"
34#include "codec_internal.h"
35#include "encode.h"
36#include "internal.h"
37#include "mpegaudio.h"
38
39typedef struct TWOLAMEContext {
40    AVClass *class;
41    int mode;
42    int psymodel;
43    int energy;
44    int error_protection;
45    int copyright;
46    int original;
47    int verbosity;
48
49    twolame_options *glopts;
50    int64_t next_pts;
51} TWOLAMEContext;
52
53static av_cold int twolame_encode_close(AVCodecContext *avctx)
54{
55    TWOLAMEContext *s = avctx->priv_data;
56    twolame_close(&s->glopts);
57    return 0;
58}
59
60static av_cold int twolame_encode_init(AVCodecContext *avctx)
61{
62    TWOLAMEContext *s = avctx->priv_data;
63    int ret;
64
65    avctx->frame_size = TWOLAME_SAMPLES_PER_FRAME;
66    avctx->initial_padding = 512 - 32 + 1;
67
68    s->glopts = twolame_init();
69    if (!s->glopts)
70        return AVERROR(ENOMEM);
71
72    twolame_set_verbosity(s->glopts, s->verbosity);
73    twolame_set_mode(s->glopts, s->mode);
74    twolame_set_psymodel(s->glopts, s->psymodel);
75    twolame_set_energy_levels(s->glopts, s->energy);
76    twolame_set_error_protection(s->glopts, s->error_protection);
77    twolame_set_copyright(s->glopts, s->copyright);
78    twolame_set_original(s->glopts, s->original);
79
80    twolame_set_num_channels(s->glopts, avctx->ch_layout.nb_channels);
81    twolame_set_in_samplerate(s->glopts, avctx->sample_rate);
82    twolame_set_out_samplerate(s->glopts, avctx->sample_rate);
83
84    if (!avctx->bit_rate) {
85        if ((s->mode == TWOLAME_AUTO_MODE && avctx->ch_layout.nb_channels == 1) || s->mode == TWOLAME_MONO)
86            avctx->bit_rate = avctx->sample_rate < 28000 ? 80000 : 192000;
87        else
88            avctx->bit_rate = avctx->sample_rate < 28000 ? 160000 : 384000;
89    }
90
91    if (avctx->flags & AV_CODEC_FLAG_QSCALE || !avctx->bit_rate) {
92        twolame_set_VBR(s->glopts, TRUE);
93        twolame_set_VBR_level(s->glopts,
94                              avctx->global_quality / (float) FF_QP2LAMBDA);
95        av_log(avctx, AV_LOG_WARNING,
96               "VBR in MP2 is a hack, use another codec that supports it.\n");
97    } else {
98        twolame_set_bitrate(s->glopts, avctx->bit_rate / 1000);
99    }
100
101    ret = twolame_init_params(s->glopts);
102    if (ret) {
103        twolame_encode_close(avctx);
104        return AVERROR_UNKNOWN;
105    }
106
107    return 0;
108}
109
110static int twolame_encode_frame(AVCodecContext *avctx, AVPacket *avpkt,
111                                const AVFrame *frame, int *got_packet_ptr)
112{
113    TWOLAMEContext *s = avctx->priv_data;
114    int ret;
115
116    if ((ret = ff_alloc_packet(avctx, avpkt, MPA_MAX_CODED_FRAME_SIZE)) < 0)
117        return ret;
118
119    if (frame) {
120        switch (avctx->sample_fmt) {
121        case AV_SAMPLE_FMT_FLT:
122            ret = twolame_encode_buffer_float32_interleaved(s->glopts,
123                                                            (const float *)frame->data[0],
124                                                            frame->nb_samples,
125                                                            avpkt->data,
126                                                            avpkt->size);
127            break;
128        case AV_SAMPLE_FMT_FLTP:
129            ret = twolame_encode_buffer_float32(s->glopts,
130                                                (const float *)frame->data[0],
131                                                (const float *)frame->data[1],
132                                                frame->nb_samples,
133                                                avpkt->data, avpkt->size);
134            break;
135        case AV_SAMPLE_FMT_S16:
136            ret = twolame_encode_buffer_interleaved(s->glopts,
137                                                    (const short int *)frame->data[0],
138                                                    frame->nb_samples,
139                                                    avpkt->data, avpkt->size);
140            break;
141        case AV_SAMPLE_FMT_S16P:
142            ret = twolame_encode_buffer(s->glopts,
143                                        (const short int *)frame->data[0],
144                                        (const short int *)frame->data[1],
145                                        frame->nb_samples,
146                                        avpkt->data, avpkt->size);
147            break;
148        default:
149            av_log(avctx, AV_LOG_ERROR,
150                   "Unsupported sample format %d.\n", avctx->sample_fmt);
151            return AVERROR_BUG;
152        }
153    } else {
154        ret = twolame_encode_flush(s->glopts, avpkt->data, avpkt->size);
155    }
156
157    if (!ret)     // no bytes written
158        return 0;
159    if (ret < 0)  // twolame error
160        return AVERROR_UNKNOWN;
161
162    if (frame) {
163        avpkt->duration = ff_samples_to_time_base(avctx, frame->nb_samples);
164        if (frame->pts != AV_NOPTS_VALUE)
165            avpkt->pts = frame->pts - ff_samples_to_time_base(avctx, avctx->initial_padding);
166    } else {
167        avpkt->pts = s->next_pts;
168    }
169    // this is for setting pts for flushed packet(s).
170    if (avpkt->pts != AV_NOPTS_VALUE)
171        s->next_pts = avpkt->pts + avpkt->duration;
172
173    av_shrink_packet(avpkt, ret);
174    *got_packet_ptr = 1;
175    return 0;
176}
177
178#define OFFSET(x) offsetof(TWOLAMEContext, x)
179#define AE AV_OPT_FLAG_AUDIO_PARAM | AV_OPT_FLAG_ENCODING_PARAM
180static const AVOption options[] = {
181    { "mode", "Mpeg Mode", OFFSET(mode), AV_OPT_TYPE_INT, { .i64 = TWOLAME_AUTO_MODE }, TWOLAME_AUTO_MODE, TWOLAME_MONO, AE, "mode"},
182        { "auto", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = TWOLAME_AUTO_MODE }, 0, 0, AE, "mode" },
183        { "stereo", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = TWOLAME_STEREO }, 0, 0, AE, "mode" },
184        { "joint_stereo", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = TWOLAME_JOINT_STEREO }, 0, 0, AE, "mode" },
185        { "dual_channel", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = TWOLAME_DUAL_CHANNEL }, 0, 0, AE, "mode" },
186        { "mono", NULL, 0, AV_OPT_TYPE_CONST, { .i64 = TWOLAME_MONO }, 0, 0, AE, "mode" },
187    { "psymodel", "Psychoacoustic Model", OFFSET(psymodel), AV_OPT_TYPE_INT, { .i64 = 3 }, -1, 4, AE},
188    { "energy_levels","enable energy levels", OFFSET(energy), AV_OPT_TYPE_INT, { .i64 = 0 },  0, 1, AE},
189    { "error_protection","enable CRC error protection", OFFSET(error_protection), AV_OPT_TYPE_INT, { .i64 = 0 },  0, 1, AE},
190    { "copyright", "set MPEG Audio Copyright flag", OFFSET(copyright), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, 1, AE},
191    { "original", "set MPEG Audio Original flag", OFFSET(original), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, 1, AE},
192    { "verbosity", "set library optput level (0-10)", OFFSET(verbosity), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, 10, AE},
193    { NULL },
194};
195
196static const AVClass twolame_class = {
197    .class_name = "libtwolame encoder",
198    .item_name  = av_default_item_name,
199    .option     = options,
200    .version    = LIBAVUTIL_VERSION_INT,
201};
202
203static const FFCodecDefault twolame_defaults[] = {
204    { "b", "0" },
205    { NULL },
206};
207
208static const int twolame_samplerates[] = {
209    16000, 22050, 24000, 32000, 44100, 48000, 0
210};
211
212const FFCodec ff_libtwolame_encoder = {
213    .p.name         = "libtwolame",
214    .p.long_name    = NULL_IF_CONFIG_SMALL("libtwolame MP2 (MPEG audio layer 2)"),
215    .p.type         = AVMEDIA_TYPE_AUDIO,
216    .p.id           = AV_CODEC_ID_MP2,
217    .priv_data_size = sizeof(TWOLAMEContext),
218    .init           = twolame_encode_init,
219    FF_CODEC_ENCODE_CB(twolame_encode_frame),
220    .close          = twolame_encode_close,
221    .p.capabilities = AV_CODEC_CAP_DELAY,
222    .defaults       = twolame_defaults,
223    .p.priv_class   = &twolame_class,
224    .p.sample_fmts  = (const enum AVSampleFormat[]) {
225        AV_SAMPLE_FMT_FLT,
226        AV_SAMPLE_FMT_FLTP,
227        AV_SAMPLE_FMT_S16,
228        AV_SAMPLE_FMT_S16P,
229        AV_SAMPLE_FMT_NONE
230    },
231#if FF_API_OLD_CHANNEL_LAYOUT
232    .p.channel_layouts = (const uint64_t[]) {
233        AV_CH_LAYOUT_MONO,
234        AV_CH_LAYOUT_STEREO,
235        0 },
236#endif
237    .p.ch_layouts    = (const AVChannelLayout[]) {
238        AV_CHANNEL_LAYOUT_MONO,
239        AV_CHANNEL_LAYOUT_STEREO,
240        { 0 },
241    },
242    .p.supported_samplerates = twolame_samplerates,
243    .p.wrapper_name = "libtwolame",
244};
245