[FFmpeg-devel] [PATCH][GSoC]audio filter-use cellular automata to generate tones
Paul B Mahol
onemda at gmail.com
Wed Mar 18 16:13:16 EET 2020
On 3/18/20, Ashutosh Pradhan <ashutoshp012345 at gmail.com> wrote:
> Use cellular automata to generate tones
>
> diff --git a/Changelog b/Changelog
> index d1572553a5..5ddd2484b0 100644
> --- a/Changelog
> +++ b/Changelog
> @@ -48,6 +48,7 @@ version <next>:
> - AMQP 0-9-1 protocol (RabbitMQ)
> - Vulkan support
> - avgblur_vulkan, overlay_vulkan, scale_vulkan and chromaber_vulkan filters
> +- atone filter
>
>
> version 4.2:
> diff --git a/configure b/configure
> index 6ceb0c7af4..2ec7e377f3 100755
> --- a/configure
> +++ b/configure
> @@ -233,6 +233,7 @@ External library support:
> and libraw1394 [no]
> --enable-libfdk-aac enable AAC de/encoding via libfdk-aac [no]
> --enable-libflite enable flite (voice synthesis) support via
> libflite [no]
> + --enable-libfluidsynth enable libfluidsynth support for fluidsynth [no]
> --enable-libfontconfig enable libfontconfig, useful for drawtext filter
> [no]
> --enable-libfreetype enable libfreetype, needed for drawtext filter
> [no]
> --enable-libfribidi enable libfribidi, improves drawtext filter [no]
> @@ -1770,6 +1771,7 @@ EXTERNAL_LIBRARY_LIST="
> libdc1394
> libdrm
> libflite
> + libfluidsynth
> libfontconfig
> libfreetype
> libfribidi
> @@ -3465,6 +3467,7 @@ asr_filter_deps="pocketsphinx"
> ass_filter_deps="libass"
> atempo_filter_deps="avcodec"
> atempo_filter_select="rdft"
> +atone_filter_deps="libfluidsynth"
> avgblur_opencl_filter_deps="opencl"
> avgblur_vulkan_filter_deps="vulkan libglslang"
> azmq_filter_deps="libzmq"
> @@ -6270,6 +6273,7 @@ enabled libfdk_aac && { check_pkg_config
> libfdk_aac fdk-aac "fdk-aac/aace
> warn "using libfdk without pkg-config"; }
> }
> flite_extralibs="-lflite_cmu_time_awb -lflite_cmu_us_awb -lflite_cmu_us_kal
> -lflite_cmu_us_kal16 -lflite_cmu_us_rms -lflite_cmu_us_slt -lflite_usenglish
> -lflite_cmulex -lflite"
> enabled libflite && require libflite "flite/flite.h" flite_init
> $flite_extralibs
> +enabled libfluidsynth && require_pkg_config libfluidsynth fluidsynth
> "fluidsynth.h" fluid_log
> enabled fontconfig && enable libfontconfig
> enabled libfontconfig && require_pkg_config libfontconfig fontconfig
> "fontconfig/fontconfig.h" FcInit
> enabled libfreetype && require_pkg_config libfreetype freetype2
> "ft2build.h FT_FREETYPE_H" FT_Init_FreeType
> diff --git a/doc/filters.texi b/doc/filters.texi
> index 328e984e92..0eda8a4c6e 100644
> --- a/doc/filters.texi
> +++ b/doc/filters.texi
> @@ -6050,6 +6050,57 @@ anoisesrc=d=60:c=pink:r=44100:a=0.5
> @end example
> @end itemize
>
> + at section atone
> +
> +Generate legato tones using cellular automata.
> +To compile filter configure ffmpeg with @code{--enable-libfluidsynth}
> +
> +The filter accepts the following options:
> +
> + at table @option
> + at item sample_rate, r
> +Specify the sample rate. Default value is 44100 Hz.
> +
> + at item sfont
> +Specify the location of soundfont file. Default value is
> +"/usr/share/sounds/sf2/FluidR3_GM.sf2"(for linux).
> +
> + at item duration, d
> +Specify the duration of the generated audio stream. Not specifying this
> option
> +results in playing tones for infinite length.
> +
> + at item velocity, v
> +Specify the velocity of key press. Default value is 80.
> +
> + at item MIDI_channel
> +Specify the MIDI channel number between(0 to 16) to play tones. Default is
> 0.
> +
> + at item tone_change_interval, t
> +Specify the time interval between successive tones in seconds. Default is
> 0.2s.
> +
> + at item rule
> +Specify the rule between 0 to 255 to get the rule set. Default is 30.
> +
> + at item width
> +Specify the width of the cells array. Default is 64.
> +
> + at item samples_per_frame
> +Set the number of samples per each output frame. Default is 1024.
> + at end table
> +
> + at subsection Examples
> +
> + at itemize
> +
> + at item
> +Generate 10 seconds of random tones, with a key velocity of 100, midi
> channel 1
> +and an tone change interval of 0.1s:
> + at example
> +atone=d=10:MIDI_channel=1:v=100:t=0.1:sfont="example.sf2":rule=193
> +atone=duration=10:MIDI_channel=1:velocity=100:tone_change_interval=0.1:sfont="example.sf2":rule=193
> + at end example
> + at end itemize
> +
> @section hilbert
>
> Generate odd-tap Hilbert transform FIR coefficients.
> diff --git a/libavfilter/Makefile b/libavfilter/Makefile
> index 750412da6b..b1f0c4be35 100644
> --- a/libavfilter/Makefile
> +++ b/libavfilter/Makefile
> @@ -151,6 +151,7 @@ OBJS-$(CONFIG_FLITE_FILTER) +=
> asrc_flite.o
> OBJS-$(CONFIG_HILBERT_FILTER) += asrc_hilbert.o
> OBJS-$(CONFIG_SINC_FILTER) += asrc_sinc.o
> OBJS-$(CONFIG_SINE_FILTER) += asrc_sine.o
> +OBJS-$(CONFIG_ATONE_FILTER) += asrc_atone.o
>
> OBJS-$(CONFIG_ANULLSINK_FILTER) += asink_anullsink.o
>
> diff --git a/libavfilter/allfilters.c b/libavfilter/allfilters.c
> index 501e5d041b..4d3efc7e15 100644
> --- a/libavfilter/allfilters.c
> +++ b/libavfilter/allfilters.c
> @@ -145,6 +145,7 @@ extern AVFilter ff_asrc_flite;
> extern AVFilter ff_asrc_hilbert;
> extern AVFilter ff_asrc_sinc;
> extern AVFilter ff_asrc_sine;
> +extern AVFilter ff_asrc_atone;
>
> extern AVFilter ff_asink_anullsink;
>
> diff --git a/libavfilter/asrc_atone.c b/libavfilter/asrc_atone.c
> new file mode 100644
> index 0000000000..04465ec4e0
> --- /dev/null
> +++ b/libavfilter/asrc_atone.c
> @@ -0,0 +1,275 @@
> +/*
> + * 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 Lesser 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
> + */
> +#include <float.h>
> +#include <stdio.h>
> +#include <string.h>
> +#include <fluidsynth.h>
> +#include <stdlib.h>
> +#include <unistd.h>
> +
> +#include "libavutil/avassert.h"
> +#include "libavutil/channel_layout.h"
> +#include "libavutil/eval.h"
> +#include "libavutil/opt.h"
> +#include "libavutil/lfg.h"
> +#include "libavutil/random_seed.h"
> +#include "audio.h"
> +#include "avfilter.h"
> +#include "internal.h"
> +
> +typedef struct AtoneContext
> +{
> + const AVClass* class;
> + int64_t duration;
> + int nb_samples;
> + int sample_rate;
> + int64_t pts;
> + int infinite;
> +
> + fluid_settings_t* settings;
> + fluid_synth_t* synth;
> + char* sfont; ///< soundfont file
> + int sfont_id;
> + int midi_chan; ///< midi channel number
> + int velocity; ///< velocity of key
> + double changerate; ///< get the time interval of changing
> tones
> + int key[7]; ///< play tones within one octave
> + int *cells;
> + int *nextgen;
> + int rule; ///< Rule for changing value in
> cell(taking 3 cells at time hence 2^8 rules possible)
> + int width; ///< width of the cell array
> + int ruleset[8]; ///< taking 3 cells at a time
> +}AtoneContext;
> +
> +#define CONTEXT AtoneContext
> +#define FLAGS AV_OPT_FLAG_AUDIO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
> +
> +#define OPT_GENERIC(name, field, def, min, max, descr, type, deffield, ...)
> \
> + { name, descr, offsetof(CONTEXT, field), AV_OPT_TYPE_ ## type,
> \
> + { .deffield = def }, min, max, FLAGS, __VA_ARGS__ }
> +
> +#define OPT_INT(name, field, def, min, max, descr, ...) \
> + OPT_GENERIC(name, field, def, min, max, descr, INT, i64, __VA_ARGS__)
> +
> +#define OPT_DBL(name, field, def, min, max, descr, ...) \
> + OPT_GENERIC(name, field, def, min, max, descr, DOUBLE, dbl,
> __VA_ARGS__)
> +
> +#define OPT_DUR(name, field, def, min, max, descr, ...) \
> + OPT_GENERIC(name, field, def, min, max, descr, DURATION, str,
> __VA_ARGS__)
> +
> +#define OPT_STR(name, field, def, min, max, descr, ...) \
> + OPT_GENERIC(name, field, def, min, max, descr, STRING, str,
> __VA_ARGS__)
> +
> +static const AVOption atone_options[] = {
> + OPT_INT("velocity", velocity, 80,
> 0, INT_MAX, "set the velocity of key press",),
> + OPT_INT("v", velocity, 80,
> 0, INT_MAX, "set the velocity of key press",),
> + OPT_INT("sample_rate", sample_rate, 44100,
> 1, INT_MAX, "set the sample rate",),
> + OPT_INT("r", sample_rate, 44100,
> 1, INT_MAX, "set the sample rate",),
> + OPT_DUR("duration", duration, 0,
> 0, INT64_MAX, "set the audio duration",),
> + OPT_DUR("d", duration, 0,
> 0, INT64_MAX, "set the audio duration",),
> + OPT_STR("sfont", sfont,
> "/usr/share/sounds/sf2/FluidR3_GM.sf2", 0, 0, "set
> the soundfont file",),
> + OPT_INT("samples_per_frame", nb_samples, 1024,
> 0, INT_MAX, "set the number of samples per
> frame",),
> + OPT_INT("MIDI_channel", midi_chan, 0,
> 0, 16, "set the MIDI Channel",),
> + OPT_DBL("tone_change_interval",changerate, 0.2,
> 0, DBL_MAX, "set the random tone change time
> in seconds",),
> + OPT_DBL("t", changerate, 0.2,
> 0, DBL_MAX, "set the random tone change time
> in seconds",),
> + OPT_INT("rule", rule, 30,
> 0, 256, "set the rule for changing value
> in cell",),
> + OPT_INT("width", width, 64,
> 0, INT_MAX, "set the width of the cell
> array",),
> + {NULL}
> +};
> +
> +AVFILTER_DEFINE_CLASS(atone);
> +
> +static av_cold int init(AVFilterContext *ctx)
> +{
> + AtoneContext *s = ctx->priv;
> + int val;
> + /*Initialise the fluidsynth settings object followed by synthesizer*/
> + s->settings = new_fluid_settings();
> + if (s->settings== NULL){
> + av_log(s, AV_LOG_ERROR, "Failed to create the fluidsynth
> settings");
> + return AVERROR_EXTERNAL;
> + }
> +
> + s->synth = new_fluid_synth(s->settings);
> + if (s->synth== NULL){
> + av_log(s, AV_LOG_ERROR, "Failed to create the fluidsynth synth");
> + return AVERROR_EXTERNAL;
> + }
> +
> + s->sfont_id= fluid_synth_sfload(s->synth, s->sfont, 1);
> + if(s->sfont_id== FLUID_FAILED){
> + av_log(s, AV_LOG_ERROR, "Loading the Soundfont Failed");
> + return AVERROR_EXTERNAL;
> + }
> + if (!(s->cells = av_malloc(sizeof(int)*(s->width))))
> + return AVERROR(ENOMEM);
> +
> + if (!(s->nextgen = av_malloc(sizeof(int)*(s->width))))
> + return AVERROR(ENOMEM);
> +
> + for (int i=0; i<7; i++)
> + s->key[i]=0;
> +
> + s->changerate= s->changerate*s->sample_rate/s->nb_samples;
> + if (s->changerate<1.0)
> + s->changerate= 1.0;
> +
> + for (int i=0; i< s->width; i++)
> + s->cells[i]=0;
> + s->cells[s->width/2]=1;
> +
> + val= s->rule;
> + for (int i=0; i<8; i++){
> + s->ruleset[i]= val%2;
> + val= val>>1;
> + }
> + return 0;
> +}
> +
> +static av_cold void uninit(AVFilterContext *ctx)
> +{
> + AtoneContext *s = ctx->priv;
> +
> + delete_fluid_synth(s->synth);
> + delete_fluid_settings(s->settings);
> + av_freep(&s->cells);
> + av_freep(&s->nextgen);
> +}
> +
> +static av_cold int config_props(AVFilterLink *outlink)
> +{
> + AtoneContext *s = outlink->src->priv;
> +
> + if (s->duration==0)
> + s->infinite=1;
> + s->duration = av_rescale(s->duration, s->sample_rate, AV_TIME_BASE);
> + return 0;
> +}
> +/*
> +Generate keys using cellular automata. Reference taken from
> http://tones.wolfram.com/about/how-it-works
> +to generate the cells. Rule 30(default) gives random numbers. Here middle
> part(+-3) of the cell generation
> +is taken to get the keys. All the keys generated in the octave are played
> at the same time.
> +*/
> +static void generate(AtoneContext* s){
> + for (int i=0; i< s->width; i++){
> + int c=0;
> + for (int j=0; j<3; j++){
> + c+= s->cells[(i+j-1)%s->width]<<(j);
> + }
> + s->nextgen[i]= s->ruleset[c];
> + }
> + memcpy(s->cells, s->nextgen, s->width*sizeof(int));
> + memcpy(s->key, &s->cells[s->width/2-3], 7*sizeof(int));
> +}
> +
> +static void gen_random_tone_samples(AtoneContext* s, int buf_size, float*
> data){
> +
> + for(int i=0; i<7; i++)
> + if (s->key[i]==1)
> + fluid_synth_noteon(s->synth, s->midi_chan, (64+i),
> s->velocity);//pp range octave
> +
> + fluid_synth_write_float(s->synth, buf_size, data, 0, 2, data, 1, 2);
> + fluid_synth_all_notes_off(s->synth, s->midi_chan);
> +}
> +
> +static int request_frame(AVFilterLink *outlink)
> +{
> + AVFilterContext *ctx = outlink->src;
> + AtoneContext *s = ctx->priv;
> + AVFrame *frame;
> + int nb_samples;
> + if (!s->infinite && s->duration <= 0) {
> + return AVERROR_EOF;
> + } else if (!s->infinite && s->duration < s->nb_samples) {
> + nb_samples = s->duration;
> + } else {
> + nb_samples = s->nb_samples;
> + }
> +
> + if (!(frame = ff_get_audio_buffer(outlink, nb_samples)))
> + return AVERROR(ENOMEM);
> +
> + if (s->pts % ((int)(s->changerate)*nb_samples) == 0)
> + generate(s);
Why this approach?
> +
> + gen_random_tone_samples(s, nb_samples, (float*)frame->data[0]);
> +
> + if (!s->infinite)
> + s->duration -= nb_samples;
> +
> + frame->pts = s->pts;
> + s->pts += nb_samples;
> + return ff_filter_frame(outlink, frame);
> +}
> +
> +static av_cold int query_formats(AVFilterContext *ctx)
> +{
> + AtoneContext *s = ctx->priv;
> + static const int64_t chlayouts[] = { AV_CH_LAYOUT_STEREO, -1 };
> + int sample_rates[] = { s->sample_rate, -1 };
> + static const enum AVSampleFormat sample_fmts[] = { AV_SAMPLE_FMT_FLT,
> AV_SAMPLE_FMT_NONE};
> + AVFilterFormats *formats;
> + AVFilterChannelLayouts *layouts;
> + int ret;
> +
> + formats = ff_make_format_list(sample_fmts);
> + if (!formats)
> + return AVERROR(ENOMEM);
> + ret = ff_set_common_formats (ctx, formats);
> + if (ret < 0)
> + return ret;
> +
> + layouts = avfilter_make_format64_list(chlayouts);
> + if (!layouts)
> + return AVERROR(ENOMEM);
> + ret = ff_set_common_channel_layouts(ctx, layouts);
> + if (ret < 0)
> + return ret;
> +
> + formats = ff_make_format_list(sample_rates);
> + if (!formats)
> + return AVERROR(ENOMEM);
> + return ff_set_common_samplerates(ctx, formats);
> +}
> +
> +
> +static const AVFilterPad atone_outputs[] = {
> + {
> + .name = "default",
> + .type = AVMEDIA_TYPE_AUDIO,
> + .request_frame = request_frame,
> + .config_props = config_props,
> + },
> + { NULL }
> +};
> +
> +AVFilter ff_asrc_atone = {
> + .name = "atone",
> + .description = NULL_IF_CONFIG_SMALL("Generate tones using cellular
> automata."),
> + .query_formats = query_formats,
> + .init = init,
> + .uninit = uninit,
> + .priv_size = sizeof(AtoneContext),
> + .inputs = NULL,
> + .outputs = atone_outputs,
> + .priv_class = &atone_class,
> +};
> +
> +
> +
> +
> diff --git a/libavfilter/version.h b/libavfilter/version.h
> index 7b41018be7..4c4e8afe2d 100644
> --- a/libavfilter/version.h
> +++ b/libavfilter/version.h
> @@ -30,7 +30,7 @@
> #include "libavutil/version.h"
>
> #define LIBAVFILTER_VERSION_MAJOR 7
> -#define LIBAVFILTER_VERSION_MINOR 77
> +#define LIBAVFILTER_VERSION_MINOR 78
> #define LIBAVFILTER_VERSION_MICRO 100
>
>
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".
More information about the ffmpeg-devel
mailing list