af_volume.c 12.4 KB
Newer Older
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
/*
 * Copyright (c) 2011 Stefano Sabatini
 * Copyright (c) 2012 Justin Ruggles <justin.ruggles@gmail.com>
 *
 * This file is part of Libav.
 *
 * Libav 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.
 *
 * Libav 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
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with Libav; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
 */

/**
 * @file
 * audio volume filter
 */

27
#include "libavutil/channel_layout.h"
28 29 30
#include "libavutil/common.h"
#include "libavutil/eval.h"
#include "libavutil/float_dsp.h"
31
#include "libavutil/intreadwrite.h"
32
#include "libavutil/opt.h"
33 34
#include "libavutil/replaygain.h"

35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
#include "audio.h"
#include "avfilter.h"
#include "formats.h"
#include "internal.h"
#include "af_volume.h"

static const char *precision_str[] = {
    "fixed", "float", "double"
};

#define OFFSET(x) offsetof(VolumeContext, x)
#define A AV_OPT_FLAG_AUDIO_PARAM

static const AVOption options[] = {
    { "volume", "Volume adjustment.",
            OFFSET(volume), AV_OPT_TYPE_DOUBLE, { .dbl = 1.0 }, 0, 0x7fffff, A },
    { "precision", "Mathematical precision.",
            OFFSET(precision), AV_OPT_TYPE_INT, { .i64 = PRECISION_FLOAT }, PRECISION_FIXED, PRECISION_DOUBLE, A, "precision" },
        { "fixed",  "8-bit fixed-point.",     0, AV_OPT_TYPE_CONST, { .i64 = PRECISION_FIXED  }, INT_MIN, INT_MAX, A, "precision" },
        { "float",  "32-bit floating-point.", 0, AV_OPT_TYPE_CONST, { .i64 = PRECISION_FLOAT  }, INT_MIN, INT_MAX, A, "precision" },
        { "double", "64-bit floating-point.", 0, AV_OPT_TYPE_CONST, { .i64 = PRECISION_DOUBLE }, INT_MIN, INT_MAX, A, "precision" },
56 57 58 59 60 61
    { "replaygain", "Apply replaygain side data when present",
            OFFSET(replaygain), AV_OPT_TYPE_INT, { .i64 = REPLAYGAIN_DROP }, REPLAYGAIN_DROP, REPLAYGAIN_ALBUM, A, "replaygain" },
        { "drop",   "replaygain side data is dropped", 0, AV_OPT_TYPE_CONST, { .i64 = REPLAYGAIN_DROP   }, 0, 0, A, "replaygain" },
        { "ignore", "replaygain side data is ignored", 0, AV_OPT_TYPE_CONST, { .i64 = REPLAYGAIN_IGNORE }, 0, 0, A, "replaygain" },
        { "track",  "track gain is preferred",         0, AV_OPT_TYPE_CONST, { .i64 = REPLAYGAIN_TRACK  }, 0, 0, A, "replaygain" },
        { "album",  "album gain is preferred",         0, AV_OPT_TYPE_CONST, { .i64 = REPLAYGAIN_ALBUM  }, 0, 0, A, "replaygain" },
62 63
    { "replaygain_preamp", "Apply replaygain pre-amplification",
            OFFSET(replaygain_preamp), AV_OPT_TYPE_DOUBLE, { .dbl = 0.0 }, -15.0, 15.0, A },
64 65
    { "replaygain_noclip", "Apply replaygain clipping prevention",
            OFFSET(replaygain_noclip), AV_OPT_TYPE_INT, { .i64 = 1 }, 0, 1, A },
66 67 68 69 70 71 72 73 74 75
    { NULL },
};

static const AVClass volume_class = {
    .class_name = "volume filter",
    .item_name  = av_default_item_name,
    .option     = options,
    .version    = LIBAVUTIL_VERSION_INT,
};

76
static av_cold int init(AVFilterContext *ctx)
77 78 79 80 81 82 83 84 85 86 87 88 89 90
{
    VolumeContext *vol = ctx->priv;

    if (vol->precision == PRECISION_FIXED) {
        vol->volume_i = (int)(vol->volume * 256 + 0.5);
        vol->volume   = vol->volume_i / 256.0;
        av_log(ctx, AV_LOG_VERBOSE, "volume:(%d/256)(%f)(%1.2fdB) precision:fixed\n",
               vol->volume_i, vol->volume, 20.0*log(vol->volume)/M_LN10);
    } else {
        av_log(ctx, AV_LOG_VERBOSE, "volume:(%f)(%1.2fdB) precision:%s\n",
               vol->volume, 20.0*log(vol->volume)/M_LN10,
               precision_str[vol->precision]);
    }

91
    return 0;
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
}

static int query_formats(AVFilterContext *ctx)
{
    VolumeContext *vol = ctx->priv;
    AVFilterFormats *formats = NULL;
    AVFilterChannelLayouts *layouts;
    static const enum AVSampleFormat sample_fmts[][7] = {
        /* PRECISION_FIXED */
        {
            AV_SAMPLE_FMT_U8,
            AV_SAMPLE_FMT_U8P,
            AV_SAMPLE_FMT_S16,
            AV_SAMPLE_FMT_S16P,
            AV_SAMPLE_FMT_S32,
            AV_SAMPLE_FMT_S32P,
            AV_SAMPLE_FMT_NONE
        },
        /* PRECISION_FLOAT */
        {
            AV_SAMPLE_FMT_FLT,
            AV_SAMPLE_FMT_FLTP,
            AV_SAMPLE_FMT_NONE
        },
        /* PRECISION_DOUBLE */
        {
            AV_SAMPLE_FMT_DBL,
            AV_SAMPLE_FMT_DBLP,
            AV_SAMPLE_FMT_NONE
        }
    };

    layouts = ff_all_channel_layouts();
    if (!layouts)
        return AVERROR(ENOMEM);
    ff_set_common_channel_layouts(ctx, layouts);

    formats = ff_make_format_list(sample_fmts[vol->precision]);
    if (!formats)
        return AVERROR(ENOMEM);
    ff_set_common_formats(ctx, formats);

    formats = ff_all_samplerates();
    if (!formats)
        return AVERROR(ENOMEM);
    ff_set_common_samplerates(ctx, formats);

    return 0;
}

static inline void scale_samples_u8(uint8_t *dst, const uint8_t *src,
                                    int nb_samples, int volume)
{
    int i;
    for (i = 0; i < nb_samples; i++)
        dst[i] = av_clip_uint8(((((int64_t)src[i] - 128) * volume + 128) >> 8) + 128);
}

static inline void scale_samples_u8_small(uint8_t *dst, const uint8_t *src,
                                          int nb_samples, int volume)
{
    int i;
    for (i = 0; i < nb_samples; i++)
        dst[i] = av_clip_uint8((((src[i] - 128) * volume + 128) >> 8) + 128);
}

static inline void scale_samples_s16(uint8_t *dst, const uint8_t *src,
                                     int nb_samples, int volume)
{
    int i;
    int16_t *smp_dst       = (int16_t *)dst;
    const int16_t *smp_src = (const int16_t *)src;
    for (i = 0; i < nb_samples; i++)
        smp_dst[i] = av_clip_int16(((int64_t)smp_src[i] * volume + 128) >> 8);
}

static inline void scale_samples_s16_small(uint8_t *dst, const uint8_t *src,
                                           int nb_samples, int volume)
{
    int i;
    int16_t *smp_dst       = (int16_t *)dst;
    const int16_t *smp_src = (const int16_t *)src;
    for (i = 0; i < nb_samples; i++)
        smp_dst[i] = av_clip_int16((smp_src[i] * volume + 128) >> 8);
}

static inline void scale_samples_s32(uint8_t *dst, const uint8_t *src,
                                     int nb_samples, int volume)
{
    int i;
    int32_t *smp_dst       = (int32_t *)dst;
    const int32_t *smp_src = (const int32_t *)src;
    for (i = 0; i < nb_samples; i++)
        smp_dst[i] = av_clipl_int32((((int64_t)smp_src[i] * volume + 128) >> 8));
}



190
static av_cold void volume_init(VolumeContext *vol)
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
{
    vol->samples_align = 1;

    switch (av_get_packed_sample_fmt(vol->sample_fmt)) {
    case AV_SAMPLE_FMT_U8:
        if (vol->volume_i < 0x1000000)
            vol->scale_samples = scale_samples_u8_small;
        else
            vol->scale_samples = scale_samples_u8;
        break;
    case AV_SAMPLE_FMT_S16:
        if (vol->volume_i < 0x10000)
            vol->scale_samples = scale_samples_s16_small;
        else
            vol->scale_samples = scale_samples_s16;
        break;
    case AV_SAMPLE_FMT_S32:
        vol->scale_samples = scale_samples_s32;
        break;
    case AV_SAMPLE_FMT_FLT:
        avpriv_float_dsp_init(&vol->fdsp, 0);
        vol->samples_align = 4;
        break;
    case AV_SAMPLE_FMT_DBL:
        avpriv_float_dsp_init(&vol->fdsp, 0);
        vol->samples_align = 8;
        break;
    }
219 220 221

    if (ARCH_X86)
        ff_volume_init_x86(vol);
222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238
}

static int config_output(AVFilterLink *outlink)
{
    AVFilterContext *ctx = outlink->src;
    VolumeContext *vol   = ctx->priv;
    AVFilterLink *inlink = ctx->inputs[0];

    vol->sample_fmt = inlink->format;
    vol->channels   = av_get_channel_layout_nb_channels(inlink->channel_layout);
    vol->planes     = av_sample_fmt_is_planar(inlink->format) ? vol->channels : 1;

    volume_init(vol);

    return 0;
}

239
static int filter_frame(AVFilterLink *inlink, AVFrame *buf)
240 241 242
{
    VolumeContext *vol    = inlink->dst->priv;
    AVFilterLink *outlink = inlink->dst->outputs[0];
243 244
    int nb_samples        = buf->nb_samples;
    AVFrame *out_buf;
245
    AVFrameSideData *sd = av_frame_get_side_data(buf, AV_FRAME_DATA_REPLAYGAIN);
246
    int ret;
247

248 249 250
    if (sd && vol->replaygain != REPLAYGAIN_IGNORE) {
        if (vol->replaygain != REPLAYGAIN_DROP) {
            AVReplayGain *replaygain = (AVReplayGain*)sd->data;
251 252 253
            int32_t gain  = 100000;
            uint32_t peak = 100000;
            float g, p;
254 255

            if (vol->replaygain == REPLAYGAIN_TRACK &&
256
                replaygain->track_gain != INT32_MIN) {
257
                gain = replaygain->track_gain;
258 259 260 261

                if (replaygain->track_peak != 0)
                    peak = replaygain->track_peak;
            } else if (replaygain->album_gain != INT32_MIN) {
262
                gain = replaygain->album_gain;
263 264 265 266

                if (replaygain->album_peak != 0)
                    peak = replaygain->album_peak;
            } else {
267 268 269 270
                av_log(inlink->dst, AV_LOG_WARNING, "Both ReplayGain gain "
                       "values are unknown.\n");
            }
            g = gain / 100000.0f;
271
            p = peak / 100000.0f;
272 273 274 275

            av_log(inlink->dst, AV_LOG_VERBOSE,
                   "Using gain %f dB from replaygain side data.\n", g);

276
            vol->volume   = pow(10, (g + vol->replaygain_preamp) / 20);
277 278
            if (vol->replaygain_noclip)
                vol->volume = FFMIN(vol->volume, 1.0 / p);
279 280 281 282 283 284 285
            vol->volume_i = (int)(vol->volume * 256 + 0.5);

            volume_init(vol);
        }
        av_frame_remove_side_data(buf, AV_FRAME_DATA_REPLAYGAIN);
    }

286 287 288 289
    if (vol->volume == 1.0 || vol->volume_i == 256)
        return ff_filter_frame(outlink, buf);

    /* do volume scaling in-place if input buffer is writable */
290
    if (av_frame_is_writable(buf)) {
291 292
        out_buf = buf;
    } else {
293
        out_buf = ff_get_audio_buffer(inlink, nb_samples);
294 295
        if (!out_buf)
            return AVERROR(ENOMEM);
296 297 298 299 300 301
        ret = av_frame_copy_props(out_buf, buf);
        if (ret < 0) {
            av_frame_free(&out_buf);
            av_frame_free(&buf);
            return ret;
        }
302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332
    }

    if (vol->precision != PRECISION_FIXED || vol->volume_i > 0) {
        int p, plane_samples;

        if (av_sample_fmt_is_planar(buf->format))
            plane_samples = FFALIGN(nb_samples, vol->samples_align);
        else
            plane_samples = FFALIGN(nb_samples * vol->channels, vol->samples_align);

        if (vol->precision == PRECISION_FIXED) {
            for (p = 0; p < vol->planes; p++) {
                vol->scale_samples(out_buf->extended_data[p],
                                   buf->extended_data[p], plane_samples,
                                   vol->volume_i);
            }
        } else if (av_get_packed_sample_fmt(vol->sample_fmt) == AV_SAMPLE_FMT_FLT) {
            for (p = 0; p < vol->planes; p++) {
                vol->fdsp.vector_fmul_scalar((float *)out_buf->extended_data[p],
                                             (const float *)buf->extended_data[p],
                                             vol->volume, plane_samples);
            }
        } else {
            for (p = 0; p < vol->planes; p++) {
                vol->fdsp.vector_dmul_scalar((double *)out_buf->extended_data[p],
                                             (const double *)buf->extended_data[p],
                                             vol->volume, plane_samples);
            }
        }
    }

333 334
    emms_c();

335
    if (buf != out_buf)
336
        av_frame_free(&buf);
337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358

    return ff_filter_frame(outlink, out_buf);
}

static const AVFilterPad avfilter_af_volume_inputs[] = {
    {
        .name           = "default",
        .type           = AVMEDIA_TYPE_AUDIO,
        .filter_frame   = filter_frame,
    },
    { NULL }
};

static const AVFilterPad avfilter_af_volume_outputs[] = {
    {
        .name         = "default",
        .type         = AVMEDIA_TYPE_AUDIO,
        .config_props = config_output,
    },
    { NULL }
};

359
AVFilter ff_af_volume = {
360 361 362 363
    .name           = "volume",
    .description    = NULL_IF_CONFIG_SMALL("Change input volume."),
    .query_formats  = query_formats,
    .priv_size      = sizeof(VolumeContext),
364
    .priv_class     = &volume_class,
365 366 367 368
    .init           = init,
    .inputs         = avfilter_af_volume_inputs,
    .outputs        = avfilter_af_volume_outputs,
};