Recherche avancée
Autres articles (25)
-
Installation en mode ferme
4 février 2011, parLe mode ferme permet d’héberger plusieurs sites de type MediaSPIP en n’installant qu’une seule fois son noyau fonctionnel.
C’est la méthode que nous utilisons sur cette même plateforme.
L’utilisation en mode ferme nécessite de connaïtre un peu le mécanisme de SPIP contrairement à la version standalone qui ne nécessite pas réellement de connaissances spécifique puisque l’espace privé habituel de SPIP n’est plus utilisé.
Dans un premier temps, vous devez avoir installé les mêmes fichiers que l’installation (...) -
La sauvegarde automatique de canaux SPIP
1er avril 2010, parDans le cadre de la mise en place d’une plateforme ouverte, il est important pour les hébergeurs de pouvoir disposer de sauvegardes assez régulières pour parer à tout problème éventuel.
Pour réaliser cette tâche on se base sur deux plugins SPIP : Saveauto qui permet une sauvegarde régulière de la base de donnée sous la forme d’un dump mysql (utilisable dans phpmyadmin) mes_fichiers_2 qui permet de réaliser une archive au format zip des données importantes du site (les documents, les éléments (...) -
Use, discuss, criticize
13 avril 2011, parTalk to people directly involved in MediaSPIP’s development, or to people around you who could use MediaSPIP to share, enhance or develop their creative projects.
The bigger the community, the more MediaSPIP’s potential will be explored and the faster the software will evolve.
A discussion list is available for all exchanges between users.
Sur d’autres sites (3908)
-
Transcode of H.264 to VP8 using libav* has incorrect frame rate
17 avril 2014, par Kevin WatsonI’ve so far failed to get the correct output frame rate when transcoding H.264 to VP8 with the libav* libraries. I created a functioning encode of Sintel.2010.720p.mkv as WebM (VP8/Vorbis) using a modification of the transcoding.c example in the FFmpeg source. Unfortunately the resulting file is 48 fps unlike the 24 fps of the original and the output of the ffmpeg command I’m trying to mimic.
I noticed ffprobe produces a tbc of double the fps for this and other H.264 videos, while the tbc of the resulting VP8 stream produced by the ffmpeg command is the default 1000. The stock transcoding.c example copies the time base of the decoder to the encoder AVCodecContext, which is 1/48. Running the ffmpeg command through gdb it looks like the time base of the AVCodecContext is set to 1/24, but making that change alone only causes the resulting video to be slowed to twice the duration at 24 fps.
I can create a usable video, but the frame rate doubles. When the output frame rate is the correct 24 fps, the video is smooth but slowed to half speed.
Here is my modification of the example.
/*
* Copyright (c) 2010 Nicolas George
* Copyright (c) 2011 Stefano Sabatini
* Copyright (c) 2014 Andrey Utkin
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
/**
* @file
* API example for demuxing, decoding, filtering, encoding and muxing
* @example doc/examples/transcoding.c
*/
#include <libavcodec></libavcodec>avcodec.h>
#include <libavformat></libavformat>avformat.h>
#include <libavfilter></libavfilter>avfiltergraph.h>
#include <libavfilter></libavfilter>avcodec.h>
#include <libavfilter></libavfilter>buffersink.h>
#include <libavfilter></libavfilter>buffersrc.h>
#include <libavutil></libavutil>opt.h>
#include <libavutil></libavutil>pixdesc.h>
#define STATS_LOG "stats.log"
static AVFormatContext *ifmt_ctx;
static AVFormatContext *ofmt_ctx;
typedef struct FilteringContext {
AVFilterContext *buffersink_ctx;
AVFilterContext *buffersrc_ctx;
AVFilterGraph *filter_graph;
} FilteringContext;
static FilteringContext *filter_ctx;
static int open_input_file(const char *filename) {
int ret;
unsigned int i;
ifmt_ctx = NULL;
if ((ret = avformat_open_input(&ifmt_ctx, filename, NULL, NULL)) < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot open input file\n");
return ret;
}
if ((ret = avformat_find_stream_info(ifmt_ctx, NULL)) < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot find stream information\n");
return ret;
}
for (i = 0; i < ifmt_ctx->nb_streams; i++) {
AVStream *stream;
AVCodecContext *codec_ctx;
stream = ifmt_ctx->streams[i];
codec_ctx = stream->codec;
/* Reencode video & audio and remux subtitles etc. */
if (codec_ctx->codec_type == AVMEDIA_TYPE_VIDEO
|| codec_ctx->codec_type == AVMEDIA_TYPE_AUDIO) {
/* Open decoder */
ret = avcodec_open2(codec_ctx,
avcodec_find_decoder(codec_ctx->codec_id), NULL);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Failed to open decoder for stream #%u\n", i);
return ret;
}
}
}
av_dump_format(ifmt_ctx, 0, filename, 0);
return 0;
}
static int init_output_context(char* filename) {
int ret;
ofmt_ctx = NULL;
avformat_alloc_output_context2(&ofmt_ctx, NULL, NULL, filename);
if (!ofmt_ctx) {
av_log(NULL, AV_LOG_ERROR, "Could not create output context\n");
return AVERROR_UNKNOWN;
}
return 0;
}
static int init_webm_encoders(int audioBitRate, int crf, int videoMaxBitRate, int threads,
char* quality, int speed, int pass, char* stats) {
AVStream *out_stream;
AVStream *in_stream;
AVCodecContext *dec_ctx, *enc_ctx;
AVCodec *encoder;
int ret;
unsigned int i;
for (i = 0; i < ifmt_ctx->nb_streams; i++) {
in_stream = ifmt_ctx->streams[i];
dec_ctx = in_stream->codec;
if (dec_ctx->codec_type == AVMEDIA_TYPE_VIDEO || dec_ctx->codec_type == AVMEDIA_TYPE_AUDIO) {
AVDictionary *opts = NULL;
if (dec_ctx->codec_type == AVMEDIA_TYPE_VIDEO) {
encoder = avcodec_find_encoder(AV_CODEC_ID_VP8);
out_stream = avformat_new_stream(ofmt_ctx, encoder);
if (!out_stream) {
av_log(NULL, AV_LOG_ERROR, "Failed allocating output stream\n");
return AVERROR_UNKNOWN;
}
enc_ctx = out_stream->codec;
enc_ctx->height = dec_ctx->height;
enc_ctx->width = dec_ctx->width;
enc_ctx->sample_aspect_ratio = dec_ctx->sample_aspect_ratio;
/* take first format from list of supported formats */
enc_ctx->pix_fmt = encoder->pix_fmts[0];
/* video time_base can be set to whatever is handy and supported by encoder */
enc_ctx->time_base = dec_ctx->time_base;
/* enc_ctx->time_base.num = 1; */
/* enc_ctx->time_base.den = 24; */
enc_ctx->bit_rate = videoMaxBitRate;
enc_ctx->thread_count = threads;
switch (pass) {
case 1:
enc_ctx->flags |= CODEC_FLAG_PASS1;
break;
case 2:
enc_ctx->flags |= CODEC_FLAG_PASS2;
if (stats) {
enc_ctx->stats_in = stats;
}
break;
}
char crfString[3];
snprintf(crfString, 3, "%d", crf);
av_dict_set(&opts, "crf", crfString, 0);
av_dict_set(&opts, "quality", quality, 0);
char speedString[3];
snprintf(speedString, 3, "%d", speed);
av_dict_set(&opts, "speed", speedString, 0);
} else {
encoder = avcodec_find_encoder(AV_CODEC_ID_VORBIS);
out_stream = avformat_new_stream(ofmt_ctx, encoder);
if (!out_stream) {
av_log(NULL, AV_LOG_ERROR, "Failed allocating output stream\n");
return AVERROR_UNKNOWN;
}
/* in_stream = ifmt_ctx->streams[i]; */
/* dec_ctx = in_stream->codec; */
enc_ctx = out_stream->codec;
/* encoder = out_stream->codec->codec; */
enc_ctx->sample_rate = dec_ctx->sample_rate;
enc_ctx->channel_layout = dec_ctx->channel_layout;
enc_ctx->channels = av_get_channel_layout_nb_channels(enc_ctx->channel_layout);
/* take first format from list of supported formats */
enc_ctx->sample_fmt = encoder->sample_fmts[0];
enc_ctx->time_base = (AVRational){1, enc_ctx->sample_rate};
enc_ctx->bit_rate = audioBitRate;
}
/* Open codec with the set options */
ret = avcodec_open2(enc_ctx, encoder, &opts);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot open video encoder for stream #%u\n", i);
return ret;
}
int unused = av_dict_count(opts);
if (unused > 0) {
av_log(NULL, AV_LOG_WARNING, "%d unused options\n", unused);
}
/* } else if (dec_ctx->codec_type == AVMEDIA_TYPE_UNKNOWN) { */
} else {
av_log(NULL, AV_LOG_FATAL, "Elementary stream #%d is of unknown type, cannot proceed\n", i);
return AVERROR_INVALIDDATA;
} /* else { */
/* /\* if this stream must be remuxed *\/ */
/* ret = avcodec_copy_context(ofmt_ctx->streams[i]->codec, */
/* ifmt_ctx->streams[i]->codec); */
/* if (ret < 0) { */
/* av_log(NULL, AV_LOG_ERROR, "Copying stream context failed\n"); */
/* return ret; */
/* } */
/* } */
if (ofmt_ctx->oformat->flags & AVFMT_GLOBALHEADER)
enc_ctx->flags |= CODEC_FLAG_GLOBAL_HEADER;
}
return 0;
}
static int open_output_file(const char *filename) {
int ret;
av_dump_format(ofmt_ctx, 0, filename, 1);
if (!(ofmt_ctx->oformat->flags & AVFMT_NOFILE)) {
ret = avio_open(&ofmt_ctx->pb, filename, AVIO_FLAG_WRITE);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Could not open output file '%s'", filename);
return ret;
}
}
/* init muxer, write output file header */
ret = avformat_write_header(ofmt_ctx, NULL);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Error occurred when opening output file\n");
return ret;
}
return 0;
}
static int init_filter(FilteringContext* fctx, AVCodecContext *dec_ctx,
AVCodecContext *enc_ctx, const char *filter_spec) {
char args[512];
int ret = 0;
AVFilter *buffersrc = NULL;
AVFilter *buffersink = NULL;
AVFilterContext *buffersrc_ctx = NULL;
AVFilterContext *buffersink_ctx = NULL;
AVFilterInOut *outputs = avfilter_inout_alloc();
AVFilterInOut *inputs = avfilter_inout_alloc();
AVFilterGraph *filter_graph = avfilter_graph_alloc();
if (!outputs || !inputs || !filter_graph) {
ret = AVERROR(ENOMEM);
goto end;
}
if (dec_ctx->codec_type == AVMEDIA_TYPE_VIDEO) {
buffersrc = avfilter_get_by_name("buffer");
buffersink = avfilter_get_by_name("buffersink");
if (!buffersrc || !buffersink) {
av_log(NULL, AV_LOG_ERROR, "filtering source or sink element not found\n");
ret = AVERROR_UNKNOWN;
goto end;
}
snprintf(args, sizeof(args),
"video_size=%dx%d:pix_fmt=%d:time_base=%d/%d:pixel_aspect=%d/%d",
dec_ctx->width, dec_ctx->height, dec_ctx->pix_fmt,
dec_ctx->time_base.num, dec_ctx->time_base.den,
dec_ctx->sample_aspect_ratio.num,
dec_ctx->sample_aspect_ratio.den);
ret = avfilter_graph_create_filter(&buffersrc_ctx, buffersrc, "in",
args, NULL, filter_graph);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot create buffer source\n");
goto end;
}
ret = avfilter_graph_create_filter(&buffersink_ctx, buffersink, "out",
NULL, NULL, filter_graph);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot create buffer sink\n");
goto end;
}
ret = av_opt_set_bin(buffersink_ctx, "pix_fmts",
(uint8_t*)&enc_ctx->pix_fmt, sizeof(enc_ctx->pix_fmt),
AV_OPT_SEARCH_CHILDREN);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot set output pixel format\n");
goto end;
}
} else if (dec_ctx->codec_type == AVMEDIA_TYPE_AUDIO) {
buffersrc = avfilter_get_by_name("abuffer");
buffersink = avfilter_get_by_name("abuffersink");
if (!buffersrc || !buffersink) {
av_log(NULL, AV_LOG_ERROR, "filtering source or sink element not found\n");
ret = AVERROR_UNKNOWN;
goto end;
}
if (!dec_ctx->channel_layout)
dec_ctx->channel_layout =
av_get_default_channel_layout(dec_ctx->channels);
snprintf(args, sizeof(args),
"time_base=%d/%d:sample_rate=%d:sample_fmt=%s:channel_layout=0x%"PRIx64,
dec_ctx->time_base.num, dec_ctx->time_base.den, dec_ctx->sample_rate,
av_get_sample_fmt_name(dec_ctx->sample_fmt),
dec_ctx->channel_layout);
ret = avfilter_graph_create_filter(&buffersrc_ctx, buffersrc, "in",
args, NULL, filter_graph);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot create audio buffer source\n");
goto end;
}
ret = avfilter_graph_create_filter(&buffersink_ctx, buffersink, "out",
NULL, NULL, filter_graph);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot create audio buffer sink\n");
goto end;
}
ret = av_opt_set_bin(buffersink_ctx, "sample_fmts",
(uint8_t*)&enc_ctx->sample_fmt, sizeof(enc_ctx->sample_fmt),
AV_OPT_SEARCH_CHILDREN);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot set output sample format\n");
goto end;
}
ret = av_opt_set_bin(buffersink_ctx, "channel_layouts",
(uint8_t*)&enc_ctx->channel_layout,
sizeof(enc_ctx->channel_layout), AV_OPT_SEARCH_CHILDREN);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot set output channel layout\n");
goto end;
}
ret = av_opt_set_bin(buffersink_ctx, "sample_rates",
(uint8_t*)&enc_ctx->sample_rate, sizeof(enc_ctx->sample_rate),
AV_OPT_SEARCH_CHILDREN);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Cannot set output sample rate\n");
goto end;
}
} else {
ret = AVERROR_UNKNOWN;
goto end;
}
/* Endpoints for the filter graph. */
outputs->name = av_strdup("in");
outputs->filter_ctx = buffersrc_ctx;
outputs->pad_idx = 0;
outputs->next = NULL;
inputs->name = av_strdup("out");
inputs->filter_ctx = buffersink_ctx;
inputs->pad_idx = 0;
inputs->next = NULL;
if (!outputs->name || !inputs->name) {
ret = AVERROR(ENOMEM);
goto end;
}
if ((ret = avfilter_graph_parse_ptr(filter_graph, filter_spec,
&inputs, &outputs, NULL)) < 0)
goto end;
if ((ret = avfilter_graph_config(filter_graph, NULL)) < 0)
goto end;
/* Fill FilteringContext */
fctx->buffersrc_ctx = buffersrc_ctx;
fctx->buffersink_ctx = buffersink_ctx;
fctx->filter_graph = filter_graph;
end:
avfilter_inout_free(&inputs);
avfilter_inout_free(&outputs);
return ret;
}
static int init_filters(enum AVCodecID audioCodec) {
const char *filter_spec;
unsigned int i;
int ret;
filter_ctx = av_malloc_array(ifmt_ctx->nb_streams, sizeof(*filter_ctx));
if (!filter_ctx)
return AVERROR(ENOMEM);
for (i = 0; i < ifmt_ctx->nb_streams; i++) {
filter_ctx[i].buffersrc_ctx = NULL;
filter_ctx[i].buffersink_ctx = NULL;
filter_ctx[i].filter_graph = NULL;
/* Skip streams that are neither audio nor video */
if (!(ifmt_ctx->streams[i]->codec->codec_type == AVMEDIA_TYPE_AUDIO
|| ifmt_ctx->streams[i]->codec->codec_type == AVMEDIA_TYPE_VIDEO))
continue;
if (ifmt_ctx->streams[i]->codec->codec_type == AVMEDIA_TYPE_VIDEO)
filter_spec = "null"; /* passthrough (dummy) filter for video */
else
/* TODO: make this more general */
if (audioCodec == AV_CODEC_ID_VORBIS) {
filter_spec = "asetnsamples=n=64";
} else {
/* filter_spec = "null"; /\* passthrough (dummy) filter for audio *\/ */
filter_spec = "fps=24";
/* filter_spec = "settb=expr=1/24"; */
}
ret = init_filter(&filter_ctx[i], ifmt_ctx->streams[i]->codec,
ofmt_ctx->streams[i]->codec, filter_spec);
if (ret)
return ret;
}
return 0;
}
static int encode_write_frame(AVFrame *filt_frame, unsigned int stream_index, int *got_frame) {
int ret;
int got_frame_local;
AVPacket enc_pkt;
int (*enc_func)(AVCodecContext *, AVPacket *, const AVFrame *, int *) =
(ifmt_ctx->streams[stream_index]->codec->codec_type ==
AVMEDIA_TYPE_VIDEO) ? avcodec_encode_video2 : avcodec_encode_audio2;
if (!got_frame)
got_frame = &got_frame_local;
/* av_log(NULL, AV_LOG_INFO, "Encoding frame\n"); */
/* encode filtered frame */
enc_pkt.data = NULL;
enc_pkt.size = 0;
av_init_packet(&enc_pkt);
ret = enc_func(ofmt_ctx->streams[stream_index]->codec, &enc_pkt,
filt_frame, got_frame);
av_frame_free(&filt_frame);
if (ret < 0)
return ret;
if (!(*got_frame))
return 0;
/* prepare packet for muxing */
enc_pkt.stream_index = stream_index;
enc_pkt.dts = av_rescale_q_rnd(enc_pkt.dts,
ofmt_ctx->streams[stream_index]->codec->time_base,
ofmt_ctx->streams[stream_index]->time_base,
AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
enc_pkt.pts = av_rescale_q_rnd(enc_pkt.pts,
ofmt_ctx->streams[stream_index]->codec->time_base,
ofmt_ctx->streams[stream_index]->time_base,
AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
enc_pkt.duration = av_rescale_q(enc_pkt.duration,
ofmt_ctx->streams[stream_index]->codec->time_base,
ofmt_ctx->streams[stream_index]->time_base);
/* av_log(NULL, AV_LOG_DEBUG, "Muxing frame\n"); */
/* mux encoded frame */
ret = av_interleaved_write_frame(ofmt_ctx, &enc_pkt);
return ret;
}
static int filter_encode_write_frame(AVFrame *frame, unsigned int stream_index) {
int ret;
AVFrame *filt_frame;
/* av_log(NULL, AV_LOG_INFO, "Pushing decoded frame to filters\n"); */
/* push the decoded frame into the filtergraph */
ret = av_buffersrc_add_frame_flags(filter_ctx[stream_index].buffersrc_ctx,
frame, 0);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Error while feeding the filtergraph\n");
return ret;
}
/* pull filtered frames from the filtergraph */
while (1) {
filt_frame = av_frame_alloc();
if (!filt_frame) {
ret = AVERROR(ENOMEM);
break;
}
/* av_log(NULL, AV_LOG_INFO, "Pulling filtered frame from filters\n"); */
ret = av_buffersink_get_frame(filter_ctx[stream_index].buffersink_ctx,
filt_frame);
if (ret < 0) {
/* if no more frames for output - returns AVERROR(EAGAIN)
* if flushed and no more frames for output - returns AVERROR_EOF
* rewrite retcode to 0 to show it as normal procedure completion
*/
if (ret == AVERROR(EAGAIN) || ret == AVERROR_EOF)
ret = 0;
av_frame_free(&filt_frame);
break;
}
filt_frame->pict_type = AV_PICTURE_TYPE_NONE;
ret = encode_write_frame(filt_frame, stream_index, NULL);
if (ret < 0)
break;
}
return ret;
}
static int flush_encoder(unsigned int stream_index) {
int ret;
int got_frame;
if (!(ofmt_ctx->streams[stream_index]->codec->codec->capabilities &
CODEC_CAP_DELAY))
return 0;
while (1) {
av_log(NULL, AV_LOG_INFO, "Flushing stream #%u encoder\n", stream_index);
ret = encode_write_frame(NULL, stream_index, &got_frame);
if (ret < 0)
break;
if (!got_frame)
return 0;
}
return ret;
}
static int transcode() {
int ret;
AVPacket packet = { .data = NULL, .size = 0 };
AVFrame *frame = NULL;
enum AVMediaType type;
unsigned int stream_index;
unsigned int i;
int got_frame;
int (*dec_func)(AVCodecContext *, AVFrame *, int *, const AVPacket *);
/* read all packets */
while (1) {
if ((ret = av_read_frame(ifmt_ctx, &packet)) < 0)
break;
stream_index = packet.stream_index;
type = ifmt_ctx->streams[packet.stream_index]->codec->codec_type;
av_log(NULL, AV_LOG_DEBUG, "Demuxer gave frame of stream_index %u\n",
stream_index);
if (filter_ctx[stream_index].filter_graph) {
av_log(NULL, AV_LOG_DEBUG, "Going to reencode&filter the frame\n");
frame = av_frame_alloc();
if (!frame) {
ret = AVERROR(ENOMEM);
break;
}
packet.dts = av_rescale_q_rnd(packet.dts,
ifmt_ctx->streams[stream_index]->time_base,
ifmt_ctx->streams[stream_index]->codec->time_base,
AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
packet.pts = av_rescale_q_rnd(packet.pts,
ifmt_ctx->streams[stream_index]->time_base,
ifmt_ctx->streams[stream_index]->codec->time_base,
AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
dec_func = (type == AVMEDIA_TYPE_VIDEO) ? avcodec_decode_video2 :
avcodec_decode_audio4;
ret = dec_func(ifmt_ctx->streams[stream_index]->codec, frame,
&got_frame, &packet);
if (ret < 0) {
av_frame_free(&frame);
av_log(NULL, AV_LOG_ERROR, "Decoding failed\n");
break;
}
if (got_frame) {
frame->pts = av_frame_get_best_effort_timestamp(frame);
ret = filter_encode_write_frame(frame, stream_index);
av_frame_free(&frame);
if (ret < 0)
goto end;
} else {
av_frame_free(&frame);
}
} else {
/* remux this frame without reencoding */
packet.dts = av_rescale_q_rnd(packet.dts,
ifmt_ctx->streams[stream_index]->time_base,
ofmt_ctx->streams[stream_index]->time_base,
AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
packet.pts = av_rescale_q_rnd(packet.pts,
ifmt_ctx->streams[stream_index]->time_base,
ofmt_ctx->streams[stream_index]->time_base,
AV_ROUND_NEAR_INF|AV_ROUND_PASS_MINMAX);
ret = av_interleaved_write_frame(ofmt_ctx, &packet);
if (ret < 0)
goto end;
}
av_free_packet(&packet);
}
/* flush filters and encoders */
for (i = 0; i < ifmt_ctx->nb_streams; i++) {
/* flush filter */
if (!filter_ctx[i].filter_graph)
continue;
ret = filter_encode_write_frame(NULL, i);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Flushing filter failed\n");
goto end;
}
/* flush encoder */
ret = flush_encoder(i);
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Flushing encoder failed\n");
goto end;
}
}
av_write_trailer(ofmt_ctx);
// Retrieve and store the first instance of codec statistics
// TODO: less naive, deal with multiple instances of statistics
for (i = 0; i < ofmt_ctx->nb_streams; i++) {
AVCodecContext* codec = ofmt_ctx->streams[i]->codec;
if ((codec->flags & CODEC_FLAG_PASS1) && (codec->stats_out)){
FILE* logfile = fopen(STATS_LOG, "wb");
fprintf(logfile, "%s", codec->stats_out);
if (fclose(logfile) < 0) {
av_log(NULL, AV_LOG_ERROR, "Error closing log file.\n");
}
break;
}
}
av_log(NULL, AV_LOG_INFO, "output duration = %" PRId64 "\n", ofmt_ctx->duration);
end:
av_free_packet(&packet);
av_frame_free(&frame);
for (i = 0; i < ifmt_ctx->nb_streams; i++) {
avcodec_close(ifmt_ctx->streams[i]->codec);
if (ofmt_ctx && ofmt_ctx->nb_streams > i && ofmt_ctx->streams[i] && ofmt_ctx->streams[i]->codec)
avcodec_close(ofmt_ctx->streams[i]->codec);
if (filter_ctx && filter_ctx[i].filter_graph)
avfilter_graph_free(&filter_ctx[i].filter_graph);
}
av_free(filter_ctx);
avformat_close_input(&ifmt_ctx);
if (ofmt_ctx && !(ofmt_ctx->oformat->flags & AVFMT_NOFILE))
avio_close(ofmt_ctx->pb);
avformat_free_context(ofmt_ctx);
if (ret < 0)
av_log(NULL, AV_LOG_ERROR, "Error occurred: %s\n", av_err2str(ret));
return ret ? 1 : 0;
}
int TranscodeToWebM(char* inputPath, char* outputPath, int audioBitRate, int crf, int videoMaxBitRate, int threads,
char* quality, int speed) {
int ret;
unsigned int pass;
char* stats = NULL;
av_register_all();
avfilter_register_all();
for (pass = 1; pass <= 2; pass++) {
if ((ret = open_input_file(inputPath)) < 0)
goto end;
if ((ret = init_output_context(outputPath)) < 0)
goto end;
if (pass == 2) {
size_t stats_length;
if (cmdutils_read_file(STATS_LOG, &stats, &stats_length) < 0) {
av_log(NULL, AV_LOG_ERROR, "Error reading stats file.\n");
break;
}
}
if ((ret = init_webm_encoders(audioBitRate, crf, videoMaxBitRate, threads, quality, speed, pass, stats)) < 0)
goto end;
if ((ret = open_output_file(outputPath)) < 0)
goto end;
if ((ret = init_filters(AV_CODEC_ID_VORBIS)) < 0)
goto end;
if ((ret = transcode()) < 0)
goto end;
}
if (remove(STATS_LOG) != 0) {
av_log(NULL, AV_LOG_ERROR, "Failed to remove %s\n", STATS_LOG);
}
end:
if (ret < 0) {
av_log(NULL, AV_LOG_ERROR, "Error occurred: %s\n", av_err2str(ret));
return ret;
}
return 0;
}Here is the output from the ffmpeg command I am trying to mimic.
ffmpeg version N-62301-g59a5384 Copyright (c) 2000-2014 the FFmpeg developers
built on Apr 9 2014 09:58:44 with gcc 4.8.2 (GCC) 20140206 (prerelease)
configuration: --prefix=/opt/ffmpeg --extra-cflags=-I/opt/x264/include --extra-ldflags=-L/opt/x264/lib --extra-libs=-ldl --enable-gpl --enable-nonfree --enable-libfdk-aac --enable-libopus --enable-libvorbis --enable-libvpx --enable-libx264
libavutil 52. 75.100 / 52. 75.100
libavcodec 55. 58.103 / 55. 58.103
libavformat 55. 36.102 / 55. 36.102
libavdevice 55. 11.100 / 55. 11.100
libavfilter 4. 3.100 / 4. 3.100
libswscale 2. 6.100 / 2. 6.100
libswresample 0. 18.100 / 0. 18.100
libpostproc 52. 3.100 / 52. 3.100
Input #0, matroska,webm, from '/mnt/scratch/test_source/Sintel.2010.720p.mkv':
Metadata:
encoder : libebml v1.0.0 + libmatroska v1.0.0
creation_time : 2011-04-24 17:20:33
Duration: 00:14:48.03, start: 0.000000, bitrate: 6071 kb/s
Chapter #0.0: start 0.000000, end 103.125000
Metadata:
title : Chapter 01
Chapter #0.1: start 103.125000, end 148.667000
Metadata:
title : Chapter 02
Chapter #0.2: start 148.667000, end 349.792000
Metadata:
title : Chapter 03
Chapter #0.3: start 349.792000, end 437.208000
Metadata:
title : Chapter 04
Chapter #0.4: start 437.208000, end 472.075000
Metadata:
title : Chapter 05
Chapter #0.5: start 472.075000, end 678.833000
Metadata:
title : Chapter 06
Chapter #0.6: start 678.833000, end 744.083000
Metadata:
title : Chapter 07
Chapter #0.7: start 744.083000, end 888.032000
Metadata:
title : Chapter 08
Stream #0:0(eng): Video: h264 (High), yuv420p(tv, bt709), 1280x544, SAR 1:1 DAR 40:17, 24 fps, 24 tbr, 1k tbn, 48 tbc
Stream #0:1(eng): Audio: ac3, 48000 Hz, 5.1(side), fltp, 640 kb/s
Metadata:
title : AC3 5.1 @ 640 Kbps
Stream #0:2(ger): Subtitle: subrip
Stream #0:3(eng): Subtitle: subrip
Stream #0:4(spa): Subtitle: subrip
Stream #0:5(fre): Subtitle: subrip
Stream #0:6(ita): Subtitle: subrip
Stream #0:7(dut): Subtitle: subrip
Stream #0:8(pol): Subtitle: subrip
Stream #0:9(por): Subtitle: subrip
Stream #0:10(rus): Subtitle: subrip
Stream #0:11(vie): Subtitle: subrip
[libvpx @ 0x24b74c0] v1.3.0
Output #0, webm, to '/mnt/scratch/test_out/Sintel.2010.720p.script.webm':
Metadata:
encoder : Lavf55.36.102
Chapter #0.0: start 0.000000, end 103.125000
Metadata:
title : Chapter 01
Chapter #0.1: start 103.125000, end 148.667000
Metadata:
title : Chapter 02
Chapter #0.2: start 148.667000, end 349.792000
Metadata:
title : Chapter 03
Chapter #0.3: start 349.792000, end 437.208000
Metadata:
title : Chapter 04
Chapter #0.4: start 437.208000, end 472.075000
Metadata:
title : Chapter 05
Chapter #0.5: start 472.075000, end 678.833000
Metadata:
title : Chapter 06
Chapter #0.6: start 678.833000, end 744.083000
Metadata:
title : Chapter 07
Chapter #0.7: start 744.083000, end 888.032000
Metadata:
title : Chapter 08
Stream #0:0(eng): Video: vp8 (libvpx), yuv420p, 1280x544 [SAR 1:1 DAR 40:17], q=-1--1, pass 2, 60000 kb/s, 1k tbn, 24 tbc
Stream #0:1(eng): Audio: vorbis (libvorbis), 48000 Hz, 5.1(side), fltp, 384 kb/s
Metadata:
title : AC3 5.1 @ 640 Kbps
Stream mapping:
Stream #0:0 -> #0:0 (h264 -> libvpx)
Stream #0:1 -> #0:1 (ac3 -> libvorbis)
Press [q] to stop, [?] for help
frame=21312 fps= 11 q=0.0 Lsize= 567191kB time=00:14:48.01 bitrate=5232.4kbits/s
video:537377kB audio:29266kB subtitle:0kB other streams:0kB global headers:7kB muxing overhead: 0.096885% -
A Guide to GDPR Sensitive Personal Data
13 mai 2024, par ErinThe General Data Protection Regulation (GDPR) is one of the world’s most stringent data protection laws. It provides a legal framework for collection and processing of the personal data of EU individuals.
The GDPR distinguishes between “special categories of personal data” (also referred to as “sensitive”) and other personal data and imposes stricter requirements on collection and processing of sensitive data. Understanding these differences will help your company comply with the requirements and avoid heavy penalties.
In this article, we’ll explain what personal data is considered “sensitive” according to the GDPR. We’ll also examine how a web analytics solution like Matomo can help you maintain compliance.
What is sensitive personal data ?
The following categories of data are treated as sensitive :
-
- Personal data revealing :
- Racial or ethnic origin ;
- Political opinions ;
- Religious or philosophical beliefs ;
- Trade union membership ;
- Genetic and biometric data ;
- Data concerning a person’s :
- Health ; or
- Sex life or sexual orientation.
- Personal data revealing :
Sensitive vs. non-sensitive personal data : What’s the difference ?
While both categories include information about an individual, sensitive data is seen as more private, or requiring a greater protection.
Sensitive data often carries a higher degree of risk and harm to the data subject, if the data is exposed. For example, a data breach exposing health records could lead to discrimination for the individuals involved. An insurance company could use the information to increase premiums or deny coverage.
In contrast, personal data like name or gender is considered less sensitive because it doesn’t carry the same degree of harm as sensitive data.
Unauthorised access to someone’s name alone is less likely to harm them or infringe on their fundamental rights and freedoms than an unauthorised access to their health records or biometric data. Note that financial information (e.g. credit card details) does not fall into the special categories of data.
Legality of processing
Under the GDPR, both sensitive and nonsensitive personal data are protected. However, the rules and conditions for processing sensitive data are more stringent.
Article 6 deals with processing of non-sensitive data and it states that processing is lawful if one of the six lawful bases for processing applies.
In contrast, Art. 9 of the GDPR states that processing of sensitive data is prohibited as a rule, but provides ten exceptions.
It is important to note that the lawful bases in Art. 6 are not the same as exceptions in Art. 9. For example, while performance of a contract or legitimate interest of the controller are a lawful basis for processing non-sensitive personal data, they are not included as an exception in Art. 9. What follows is that controllers are not permitted to process sensitive data on the basis of contract or legitimate interest.
The exceptions where processing of sensitive personal data is permitted (subject to additional requirements) are :
- Explicit consent : The individual has given explicit consent to processing their sensitive personal data for specified purpose(s), except where an EU member state prohibits such consent. See below for more information about explicit consent.
- Employment, social security or social protection : Processing sensitive data is necessary to perform tasks under employment, social security or social protection law.
- Vital interests : Processing sensitive data is necessary to protect the interests of a data subject or if the individual is physically or legally incapable of consenting.
- Non-for-profit bodies : Foundations, associations or nonprofits with a political, philosophical, religious or trade union aim may process the sensitive data of their members or those they are in regular contact with, in connection with their purposes (and no disclosure of the data is permitted outside the organisation, without the data subject’s consent).
- Made public : In some cases, it may be permissible to process the sensitive data of a data subject if the individual has already made it public and accessible.
- Legal claims : Processing sensitive data is necessary to establish, exercise or defend legal claims, including legal or in court proceedings.
- Public interest : Processing is necessary for reasons of substantial public interest, like preventing unlawful acts or protecting the public.
- Health or social care : Processing special category data is necessary for : preventative or occupational medicine, providing health and social care, medical diagnosis or managing healthcare systems.
- Public health : It is permissible to process sensitive data for public health reasons, like protecting against cross-border threats to health or ensuring the safety of medicinal products or medical devices.
- Archiving, research and statistics : You may process sensitive data if it’s done for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes.
In addition, you must adhere to all data handling requirements set by the GDPR.
Important : Note that for any data sent that you are processing, you always need to identify a lawful basis under Art. 6. In addition, if the data sent contains sensitive data, you must comply with Art. 9.
Explicit consent
While consent is a valid lawful basis for processing non-sensitive personal data, controllers are permitted to process sensitive data only with an “explicit consent” of the data subject.
The GDPR does not define “explicit” consent, but it is accepted that it must meet all Art. 7 conditions for consent, at a higher threshold. To be “explicit” a consent requires a clear statement (oral or written) of the data subject. Consent inferred from the data subject’s actions does not meet the threshold.
The controller must retain records of the explicit consent and provide appropriate consent withdrawal method to allow the data subject to exercise their rights.
Examples of compliant and non-compliant sensitive data processing
Here are examples of when you can and can’t process sensitive data :
- When you can process sensitive data : A doctor logs sensitive data about a patient, including their name, symptoms and medicine prescribed. The hospital can process this data to provide appropriate medical care to their patients. An IoT device and software manufacturer processes their customers’ health data based on explicit consent of each customer.
- When you can’t process sensitive data : One example is when you don’t have explicit consent from a data subject. Another is when there’s no lawful basis for processing it or you are collecting personal data you simply do not need. For example, you don’t need your customer’s ethnic origin to fulfil an online order.
Other implications of processing sensitive data
If you process sensitive data, especially on a large scale, GDPR imposes additional requirements, such as having Data Privacy Impact Assessments, appointing Data Protection Officers and EU Representatives, if you are a controller based outside the EU.
Penalties for GDPR non-compliance
Mishandling sensitive data (or processing it when you’re not allowed to) can result in huge penalties. There are two tiers of GDPR fines :
- €10 million or 2% of a company’s annual revenue for less severe infringements
- €20 million or 4% of a company’s annual revenue for more severe infringements
In the first half of 2023 alone, fines imposed in the EU due to GDPR violations exceeded €1.6 billion, up from €73 million in 2019.
Examples of high-profile violations in the last few years include :
- Amazon : The Luxembourg National Commission fined the retail giant with a massive $887 million fine in 2021 for not processing personal data per the GDPR.
- Google : The National Data Protection Commission (CNIL) fined Google €50 million for not getting proper consent to display personalised ads.
- H&M : The Hamburg Commissioner for Data Protection and Freedom of Information hit the multinational clothing company with a €35.3 million fine in 2020 for unlawfully gathering and storing employees’ data in its service centre.
One of the criteria that affects the severity of a fine is “data category” — the type of personal data being processed. Companies need to take extra precautions with sensitive data, or they risk receiving more severe penalties.
What’s more, GDPR violations can negatively affect your brand’s reputation and cause you to lose business opportunities from consumers concerned about your data practices. 76% of consumers indicated they wouldn’t buy from companies they don’t trust with their personal data.
Organisations should lay out their data practices in simple terms and make this information easily accessible so customers know how their data is being handled.
Get started with GDPR-compliant web analytics
The GDPR offers a framework for securing and protecting personal data. But it also distinguishes between sensitive and non-sensitive data. Understanding these differences and applying the lawful basis for processing this data type will help ensure compliance.
Looking for a GDPR-compliant web analytics solution ?
At Matomo, we take data privacy seriously.
Our platform ensures 100% data ownership, putting you in complete control of your data. Unlike other web analytics solutions, your data remains solely yours and isn’t sold or auctioned off to advertisers.
Additionally, with Matomo, you can be confident in the accuracy of the insights you receive, as we provide reliable, unsampled data.
Matomo also fully complies with GDPR and other data privacy laws like CCPA, LGPD and more.
Start your 21-day free trial today ; no credit card required.
Disclaimer
We are not lawyers and don’t claim to be. The information provided here is to help give an introduction to GDPR. We encourage every business and website to take data privacy seriously and discuss these issues with your lawyer if you have any concerns.
Try Matomo for Free
21 day free trial. No credit card required.
-
-
Four Trends Shaping the Future of Analytics in Banking
27 novembre 2024, par Daniel Crough — Banking and Financial ServicesWhile retail banking revenues have been growing in recent years, trends like rising financial crimes and capital required for generative AI and ML tech pose significant risks and increase operating costs across the financial industry, according to McKinsey’s State of Retail Banking report.
Today’s financial institutions are focused on harnessing AI and advanced analytics to make their data work for them. To be up to the task, analytics solutions must allow banks to give consumers the convenient, personalised experiences they want while respecting their privacy.
In this article, we’ll explore some of the big trends shaping the future of analytics in banking and finance. We’ll also look at how banks use data and technology to cut costs and personalise customer experiences.
So, let’s get into it.
This doesn’t just represent a security risk, it also impacts the usability for both customers and employees. Does any of the following sound familiar ?
- Only specific senior employees know how to navigate the software to generate custom reports or use its more advanced features.
- Customer complaints about your site’s usability or online banking experience are routine.
- Onboarding employees takes much longer than necessary because of convoluted systems.
- Teams and departments experience ‘data siloing,’ meaning that not everyone can access the data they need.
These are warning signs that IT systems are ready for a review. Anyone thinking, “If it’s not broken, why fix it ?” should consider that legacy systems can also present data security risks. As more countries introduce regulations to protect customer privacy, staying ahead of the curve is increasingly important to avoid penalties and litigation.
And regulations aren’t the only trends impacting the future of financial institutions’ IT and analytics.
4 trends shaping the future of analytics in banking
New regulations and new technology have changed the landscape of analytics in banking.
New privacy regulations impact banks globally
The first major international example was the advent of GDPR, which went into effect in the EU in 2018. But a lot has happened since. New privacy regulations and restrictions around AI continue to roll out.
- The European Artificial Intelligence Act (EU AI Act), which was held up as the world’s first comprehensive legislation on AI, took effect on 31 July 2024.
- In Europe’s federated data initiative, Gaia-X’s planned cloud infrastructure will provide for more secure, transparent, and trustworthy data storage and processing.
- The revised Payment Services Directive (PSD2) makes payments more secure and strengthens protections for European businesses and consumers, aiming to create a more integrated and efficient payments market.
But even businesses that don’t have customers in Europe aren’t safe. Consumer privacy is a hot-button issue globally.
For example, the California Consumer Privacy Act (CCPA), which took effect in January, impacts the financial services industry more than any other. Case in point, 34% of CCPA-related cases filed in 2022 were related to the financial sector.
California’s privacy regulations were the first in the US, but other states are following closely behind. On 1 July 2024, new privacy laws went into effect in Florida, Oregon, and Texas, giving people more control over their data.
One typical issue for companies in the banking industry is that their privacy measures regarding user data collected from their website are much less lax than those in their online banking system.
It’s better to proactively invest in a privacy-centric analytics platform before you get tangled up in a lawsuit and have to pay a fine (and are forced to change your system anyway).
And regulatory compliance isn’t the only bonus of an ethical analytics solution. The right alternative can unlock key customer insights that can help you improve the user experience.
The demand for personalised banking services
At the same time, consumers are expecting a more and more streamlined personal experience from financial institutions. 86% of bank employees say personalisation is a clear priority for the company. But 63% described resources as limited or only available after demonstrating clear business cases.
McKinsey’s The data and analytics edge in corporate and commercial banking points out how advanced analytics are empowering frontline bank employees to give customers more personalised experiences at every stage :
- Pre-meeting/meeting prep : Using advanced analytics to assess customer potential, recommend products, and identify prospects who are most likely to convert
- Meetings/negotiation : Applying advanced models to support price negotiations, what-if scenarios and price multiple products simultaneously
- Post-meeting/tracking : Using advanced models to identify behaviours that lead to high performance and improve forecast accuracy and sales execution
Today’s banks must deliver the personalisation that drives customer satisfaction and engagement to outperform their competitors.
The rise of AI and its role in banking
With AI and machine learning technologies becoming more powerful and accessible, financial institutions around the world are already reaping the rewards.
McKinsey estimates that AI in banking could add $200 to 340 billion annually across the global banking sector through productivity gains.
- Credit card fraud prevention : Algorithms analyse usage to flag and block fraudulent transactions.
- More accurate forecasting : AI-based tools can analyse a broader spectrum of data points and forecast more accurately.
- Better risk assessment and modelling : More advanced analytics and predictive models help avoid extending credit to high-risk customers.
- Predictive analytics : Help spot clients most likely to churn
- Gen-AI assistants : Instantly analyse customer profiles and apply predictive models to suggest the next best actions.
Considering these market trends, let’s discuss how you can move your bank into the future.
Using analytics to minimise risk and establish a competitive edge
With the right approach, you can leverage analytics and AI to help future-proof your bank against changing customer expectations, increased fraud, and new regulations.
Use machine learning to prevent fraud
Every year, more consumers are victims of credit and debit card fraud. Debit card skimming cases nearly doubled in the US in 2023. The last thing you want as a bank is to put your customer in a situation where a criminal has spent their money.
This not only leads to a horrible customer experience but also creates a lot of internal work and additional costs.Thankfully, machine learning can help identify suspicious activity and stop transactions before they go through. For example, Mastercard’s fraud prevention model has improved fraud detection rates by 20–300%.
Implementing a solution like this (or partnering with credit card companies who use it) may be a way to reduce risk and improve customer trust.
Foresee and avoid future issues with AI-powered risk management
Regardless of what type of financial products organisations offer, AI can be an enormous tool. Here are just a few ways in which it can mitigate financial risk in the future :
- Predictive analytics can evaluate risk exposure and allow for more informed decisions about whether to approve commercial loan applications.
- With better credit risk modelling, banks can avoid extending personal loans to customers most likely to default.
- Investment banks (or individual traders or financial analysts) can use AI- and ML-based systems to monitor market and trading activity more effectively.
Those are just a few examples that barely scratch the surface. Many other AI-based applications and analytics use cases exist across all industries and market segments.
Protect customer privacy while still getting detailed analytics
New regulations and increasing consumer privacy concerns don’t mean banks and financial institutions should forego website analytics altogether. Its insights into performance and customer behaviour are simply too valuable. And without customer interaction data, you’ll only know something’s wrong if someone complains.
Fortunately, it doesn’t have to be one or the other. The right financial analytics solution can give you the data and insights needed without compromising privacy while complying with regulations like GDPR and CCPA.
That way, you can track usage patterns and improve site performance and content quality based on accurate data — without compromising privacy. Reliable, precise analytics are crucial for any bank that’s serious about user experience.
Use A/B testing and other tools to improve digital customer experiences
Personalised digital experiences can be key differentiators in banking and finance when done well. But there’s stiff competition. In 2023, 40% of bank customers rated their bank’s online and mobile experience as excellent.
Improving digital experiences for users while respecting their privacy means going above and beyond a basic web analytics tool like Google Analytics. Invest in a platform with features like A/B tests and user session analysis for deeper insights into user behaviour.
Behavioural analytics are crucial to understanding customer interactions. By identifying points of friction and drop-off points, you can make digital experiences smoother and more engaging.
Matomo offers all this and is a great GDPR-compliant alternative to Google Analytics for banks and financial institutions.
Of course, this can be challenging. This is why taking an ethical and privacy-centric approach to analytics can be a key competitive edge for banks. Prioritising data security and privacy will attract other like-minded, ethically conscious consumers and boost customer loyalty.
Get privacy-friendly web analytics suitable for banking & finance with Matomo
Improving digital experiences for today’s customers requires a solid web analytics platform that prioritises data privacy and accurate analytics. And choosing the wrong one could even mean ending up in legal trouble or scrambling to reconstruct your entire analytics setup.
Matomo provides privacy-friendly analytics with 100% data accuracy (no sampling), advanced privacy controls and the ability to run A/B tests and user session analysis within the same platform (limiting risk and minimising costs).
It’s easy to get started with Matomo. Users can access clear, easy-to-understand metrics and plenty of pre-made reports that deliver valuable insights from day one. Form usage reports can help banks and fintechs identify potential issues with broken links or technical glitches and reveal clues on improving UX in the short term.
Over one million websites, including some of the world’s top banks and financial institutions, use Matomo for their analytics.
Start your 21-day free trial to see why, or book a demo with one of our analytics experts.
Syndiquer tout le site
