mirror of
https://github.com/FFmpeg/FFmpeg.git
synced 2024-12-28 20:53:54 +02:00
6b71ed1678
Originally committed as revision 18733 to svn://svn.ffmpeg.org/ffmpeg/trunk
178 lines
5.2 KiB
C
178 lines
5.2 KiB
C
/*
|
|
* mtv demuxer
|
|
* Copyright (c) 2006 Reynaldo H. Verdejo Pinochet
|
|
*
|
|
* 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
|
|
*/
|
|
|
|
/**
|
|
* @file libavformat/mtv.c
|
|
* MTV demuxer.
|
|
*/
|
|
|
|
#include "libavutil/bswap.h"
|
|
#include "avformat.h"
|
|
|
|
#define MTV_ASUBCHUNK_DATA_SIZE 500
|
|
#define MTV_HEADER_SIZE 512
|
|
#define MTV_AUDIO_PADDING_SIZE 12
|
|
#define AUDIO_SAMPLING_RATE 44100
|
|
#define VIDEO_SID 0
|
|
#define AUDIO_SID 1
|
|
|
|
typedef struct MTVDemuxContext {
|
|
|
|
unsigned int file_size; ///< filesize, not always right
|
|
unsigned int segments; ///< number of 512 byte segments
|
|
unsigned int audio_identifier; ///< 'MP3' on all files I have seen
|
|
unsigned int audio_br; ///< bitrate of audio chanel (mp3)
|
|
unsigned int img_colorfmt; ///< frame colorfmt rgb 565/555
|
|
unsigned int img_bpp; ///< frame bits per pixel
|
|
unsigned int img_width; //
|
|
unsigned int img_height; //
|
|
unsigned int img_segment_size; ///< size of image segment
|
|
unsigned int video_fps; //
|
|
unsigned int full_segment_size;
|
|
|
|
} MTVDemuxContext;
|
|
|
|
static int mtv_probe(AVProbeData *p)
|
|
{
|
|
/* Magic is 'AMV' */
|
|
if(*(p->buf) != 'A' || *(p->buf+1) != 'M' || *(p->buf+2) != 'V')
|
|
return 0;
|
|
|
|
return AVPROBE_SCORE_MAX;
|
|
}
|
|
|
|
static int mtv_read_header(AVFormatContext *s, AVFormatParameters *ap)
|
|
{
|
|
MTVDemuxContext *mtv = s->priv_data;
|
|
ByteIOContext *pb = s->pb;
|
|
AVStream *st;
|
|
unsigned int audio_subsegments;
|
|
|
|
url_fskip(pb, 3);
|
|
mtv->file_size = get_le32(pb);
|
|
mtv->segments = get_le32(pb);
|
|
url_fskip(pb, 32);
|
|
mtv->audio_identifier = get_le24(pb);
|
|
mtv->audio_br = get_le16(pb);
|
|
mtv->img_colorfmt = get_le24(pb);
|
|
mtv->img_bpp = get_byte(pb);
|
|
mtv->img_width = get_le16(pb);
|
|
mtv->img_height = get_le16(pb);
|
|
mtv->img_segment_size = get_le16(pb);
|
|
url_fskip(pb, 4);
|
|
audio_subsegments = get_le16(pb);
|
|
mtv->full_segment_size =
|
|
audio_subsegments * (MTV_AUDIO_PADDING_SIZE + MTV_ASUBCHUNK_DATA_SIZE) +
|
|
mtv->img_segment_size;
|
|
mtv->video_fps = (mtv->audio_br / 4) / audio_subsegments;
|
|
|
|
// FIXME Add sanity check here
|
|
|
|
// all systems go! init decoders
|
|
|
|
// video - raw rgb565
|
|
|
|
st = av_new_stream(s, VIDEO_SID);
|
|
if(!st)
|
|
return AVERROR(ENOMEM);
|
|
|
|
av_set_pts_info(st, 64, 1, mtv->video_fps);
|
|
st->codec->codec_type = CODEC_TYPE_VIDEO;
|
|
st->codec->codec_id = CODEC_ID_RAWVIDEO;
|
|
st->codec->codec_tag = MKTAG('R', 'G', 'B', mtv->img_bpp);
|
|
st->codec->width = mtv->img_width;
|
|
st->codec->height = mtv->img_height;
|
|
st->codec->bits_per_coded_sample = mtv->img_bpp;
|
|
st->codec->sample_rate = mtv->video_fps;
|
|
|
|
// audio - mp3
|
|
|
|
st = av_new_stream(s, AUDIO_SID);
|
|
if(!st)
|
|
return AVERROR(ENOMEM);
|
|
|
|
av_set_pts_info(st, 64, 1, AUDIO_SAMPLING_RATE);
|
|
st->codec->codec_type = CODEC_TYPE_AUDIO;
|
|
st->codec->codec_id = CODEC_ID_MP3;
|
|
st->codec->bit_rate = mtv->audio_br;
|
|
st->need_parsing = AVSTREAM_PARSE_FULL;
|
|
|
|
// Jump over header
|
|
|
|
if(url_fseek(pb, MTV_HEADER_SIZE, SEEK_SET) != MTV_HEADER_SIZE)
|
|
return AVERROR(EIO);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
static int mtv_read_packet(AVFormatContext *s, AVPacket *pkt)
|
|
{
|
|
MTVDemuxContext *mtv = s->priv_data;
|
|
ByteIOContext *pb = s->pb;
|
|
int ret;
|
|
#ifndef WORDS_BIGENDIAN
|
|
int i;
|
|
#endif
|
|
|
|
if((url_ftell(pb) - s->data_offset + mtv->img_segment_size) % mtv->full_segment_size)
|
|
{
|
|
url_fskip(pb, MTV_AUDIO_PADDING_SIZE);
|
|
|
|
ret = av_get_packet(pb, pkt, MTV_ASUBCHUNK_DATA_SIZE);
|
|
if(ret != MTV_ASUBCHUNK_DATA_SIZE)
|
|
return AVERROR(EIO);
|
|
|
|
pkt->pos -= MTV_AUDIO_PADDING_SIZE;
|
|
pkt->stream_index = AUDIO_SID;
|
|
|
|
}else
|
|
{
|
|
ret = av_get_packet(pb, pkt, mtv->img_segment_size);
|
|
if(ret != mtv->img_segment_size)
|
|
return AVERROR(EIO);
|
|
|
|
#ifndef WORDS_BIGENDIAN
|
|
|
|
/* pkt->data is GGGRRRR BBBBBGGG
|
|
* and we need RRRRRGGG GGGBBBBB
|
|
* for PIX_FMT_RGB565 so here we
|
|
* just swap bytes as they come
|
|
*/
|
|
|
|
for(i=0;i<mtv->img_segment_size/2;i++)
|
|
*((uint16_t *)pkt->data+i) = bswap_16(*((uint16_t *)pkt->data+i));
|
|
#endif
|
|
pkt->stream_index = VIDEO_SID;
|
|
}
|
|
|
|
return ret;
|
|
}
|
|
|
|
AVInputFormat mtv_demuxer = {
|
|
"MTV",
|
|
NULL_IF_CONFIG_SMALL("MTV format"),
|
|
sizeof(MTVDemuxContext),
|
|
mtv_probe,
|
|
mtv_read_header,
|
|
mtv_read_packet,
|
|
};
|