ffmpeg/libavformat/xmv.c

587 lines
16 KiB
C
Raw Normal View History

/*
* Microsoft XMV demuxer
* Copyright (c) 2011 Sven Hesse <drmccoy@drmccoy.de>
* Copyright (c) 2011 Matthew Hoops <clone2727@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
* Microsoft XMV demuxer
*/
#include <inttypes.h>
#include "libavutil/intreadwrite.h"
#include "avformat.h"
#include "internal.h"
#include "riff.h"
#define XMV_MIN_HEADER_SIZE 36
#define XMV_AUDIO_ADPCM51_FRONTLEFTRIGHT 1
#define XMV_AUDIO_ADPCM51_FRONTCENTERLOW 2
#define XMV_AUDIO_ADPCM51_REARLEFTRIGHT 4
#define XMV_AUDIO_ADPCM51 (XMV_AUDIO_ADPCM51_FRONTLEFTRIGHT | \
XMV_AUDIO_ADPCM51_FRONTCENTERLOW | \
XMV_AUDIO_ADPCM51_REARLEFTRIGHT)
#define XMV_BLOCK_ALIGN_SIZE 36
typedef struct XMVAudioTrack {
uint16_t compression;
uint16_t channels;
uint32_t sample_rate;
uint16_t bits_per_sample;
uint32_t bit_rate;
uint16_t flags;
uint16_t block_align;
uint16_t block_samples;
enum AVCodecID codec_id;
} XMVAudioTrack;
typedef struct XMVVideoPacket {
/* The decoder stream index for this video packet. */
int stream_index;
uint32_t data_size;
uint32_t data_offset;
uint32_t current_frame;
uint32_t frame_count;
/* Does the video packet contain extra data? */
int has_extradata;
/* Extra data */
uint8_t extradata[4];
int64_t last_pts;
int64_t pts;
} XMVVideoPacket;
typedef struct XMVAudioPacket {
/* The decoder stream index for this audio packet. */
int stream_index;
/* The audio track this packet encodes. */
XMVAudioTrack *track;
uint32_t data_size;
uint32_t data_offset;
uint32_t frame_size;
uint32_t block_count;
} XMVAudioPacket;
typedef struct XMVDemuxContext {
uint16_t audio_track_count;
XMVAudioTrack *audio_tracks;
uint32_t this_packet_size;
uint32_t next_packet_size;
uint32_t this_packet_offset;
uint32_t next_packet_offset;
uint16_t current_stream;
uint16_t stream_count;
XMVVideoPacket video;
XMVAudioPacket *audio;
} XMVDemuxContext;
static int xmv_probe(AVProbeData *p)
{
uint32_t file_version;
if (p->buf_size < XMV_MIN_HEADER_SIZE)
return 0;
file_version = AV_RL32(p->buf + 16);
if ((file_version == 0) || (file_version > 4))
return 0;
if (!memcmp(p->buf + 12, "xobX", 4))
return AVPROBE_SCORE_MAX;
return 0;
}
static int xmv_read_close(AVFormatContext *s)
{
XMVDemuxContext *xmv = s->priv_data;
av_free(xmv->audio);
av_free(xmv->audio_tracks);
return 0;
}
static int xmv_read_header(AVFormatContext *s)
{
XMVDemuxContext *xmv = s->priv_data;
AVIOContext *pb = s->pb;
AVStream *vst = NULL;
uint32_t file_version;
uint32_t this_packet_size;
uint16_t audio_track;
int ret;
avio_skip(pb, 4); /* Next packet size */
this_packet_size = avio_rl32(pb);
avio_skip(pb, 4); /* Max packet size */
avio_skip(pb, 4); /* "xobX" */
file_version = avio_rl32(pb);
if ((file_version != 4) && (file_version != 2))
avpriv_request_sample(s, "Uncommon version %"PRIu32"", file_version);
/* Video track */
vst = avformat_new_stream(s, NULL);
if (!vst)
return AVERROR(ENOMEM);
avpriv_set_pts_info(vst, 32, 1, 1000);
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
2014-06-18 18:42:52 +00:00
vst->codecpar->codec_type = AVMEDIA_TYPE_VIDEO;
vst->codecpar->codec_id = AV_CODEC_ID_WMV2;
vst->codecpar->codec_tag = MKBETAG('W', 'M', 'V', '2');
vst->codecpar->width = avio_rl32(pb);
vst->codecpar->height = avio_rl32(pb);
vst->duration = avio_rl32(pb);
xmv->video.stream_index = vst->index;
/* Audio tracks */
xmv->audio_track_count = avio_rl16(pb);
avio_skip(pb, 2); /* Unknown (padding?) */
xmv->audio_tracks = av_malloc(xmv->audio_track_count * sizeof(XMVAudioTrack));
if (!xmv->audio_tracks)
return AVERROR(ENOMEM);
xmv->audio = av_malloc(xmv->audio_track_count * sizeof(XMVAudioPacket));
if (!xmv->audio) {
ret = AVERROR(ENOMEM);
goto fail;
}
for (audio_track = 0; audio_track < xmv->audio_track_count; audio_track++) {
XMVAudioTrack *track = &xmv->audio_tracks[audio_track];
XMVAudioPacket *packet = &xmv->audio [audio_track];
AVStream *ast = NULL;
track->compression = avio_rl16(pb);
track->channels = avio_rl16(pb);
track->sample_rate = avio_rl32(pb);
track->bits_per_sample = avio_rl16(pb);
track->flags = avio_rl16(pb);
track->bit_rate = track->bits_per_sample *
track->sample_rate *
track->channels;
track->block_align = XMV_BLOCK_ALIGN_SIZE * track->channels;
track->block_samples = 64;
track->codec_id = ff_wav_codec_get_id(track->compression,
track->bits_per_sample);
packet->track = track;
packet->stream_index = -1;
packet->frame_size = 0;
packet->block_count = 0;
/* TODO: ADPCM'd 5.1 sound is encoded in three separate streams.
* Those need to be interleaved to a proper 5.1 stream. */
if (track->flags & XMV_AUDIO_ADPCM51)
av_log(s, AV_LOG_WARNING, "Unsupported 5.1 ADPCM audio stream "
"(0x%04X)\n", track->flags);
if (!track->channels || !track->sample_rate ||
track->channels >= UINT16_MAX / XMV_BLOCK_ALIGN_SIZE) {
av_log(s, AV_LOG_ERROR, "Invalid parameters for audio track %"PRIu16".\n",
audio_track);
ret = AVERROR_INVALIDDATA;
goto fail;
}
ast = avformat_new_stream(s, NULL);
if (!ast) {
ret = AVERROR(ENOMEM);
goto fail;
}
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
2014-06-18 18:42:52 +00:00
ast->codecpar->codec_type = AVMEDIA_TYPE_AUDIO;
ast->codecpar->codec_id = track->codec_id;
ast->codecpar->codec_tag = track->compression;
ast->codecpar->channels = track->channels;
ast->codecpar->sample_rate = track->sample_rate;
ast->codecpar->bits_per_coded_sample = track->bits_per_sample;
ast->codecpar->bit_rate = track->bit_rate;
ast->codecpar->block_align = 36 * track->channels;
avpriv_set_pts_info(ast, 32, track->block_samples, track->sample_rate);
packet->stream_index = ast->index;
ast->duration = vst->duration;
}
/** Initialize the packet context */
xmv->next_packet_offset = avio_tell(pb);
xmv->next_packet_size = this_packet_size - xmv->next_packet_offset;
xmv->stream_count = xmv->audio_track_count + 1;
return 0;
fail:
xmv_read_close(s);
return ret;
}
static void xmv_read_extradata(uint8_t *extradata, AVIOContext *pb)
{
/* Read the XMV extradata */
uint32_t data = avio_rl32(pb);
int mspel_bit = !!(data & 0x01);
int loop_filter = !!(data & 0x02);
int abt_flag = !!(data & 0x04);
int j_type_bit = !!(data & 0x08);
int top_left_mv_flag = !!(data & 0x10);
int per_mb_rl_bit = !!(data & 0x20);
int slice_count = (data >> 6) & 7;
/* Write it back as standard WMV2 extradata */
data = 0;
data |= mspel_bit << 15;
data |= loop_filter << 14;
data |= abt_flag << 13;
data |= j_type_bit << 12;
data |= top_left_mv_flag << 11;
data |= per_mb_rl_bit << 10;
data |= slice_count << 7;
AV_WB32(extradata, data);
}
static int xmv_process_packet_header(AVFormatContext *s)
{
XMVDemuxContext *xmv = s->priv_data;
AVIOContext *pb = s->pb;
uint8_t data[8];
uint16_t audio_track;
uint32_t data_offset;
/* Next packet size */
xmv->next_packet_size = avio_rl32(pb);
/* Packet video header */
if (avio_read(pb, data, 8) != 8)
return AVERROR(EIO);
xmv->video.data_size = AV_RL32(data) & 0x007FFFFF;
xmv->video.current_frame = 0;
xmv->video.frame_count = (AV_RL32(data) >> 23) & 0xFF;
xmv->video.has_extradata = (data[3] & 0x80) != 0;
/* Adding the audio data sizes and the video data size keeps you 4 bytes
* short for every audio track. But as playing around with XMV files with
* ADPCM audio showed, taking the extra 4 bytes from the audio data gives
* you either completely distorted audio or click (when skipping the
2012-12-19 17:48:21 +00:00
* remaining 68 bytes of the ADPCM block). Subtracting 4 bytes for every
* audio track from the video data works at least for the audio. Probably
* some alignment thing?
* The video data has (always?) lots of padding, so it should work out...
*/
xmv->video.data_size -= xmv->audio_track_count * 4;
xmv->current_stream = 0;
if (!xmv->video.frame_count) {
xmv->video.frame_count = 1;
xmv->current_stream = 1;
}
/* Packet audio header */
for (audio_track = 0; audio_track < xmv->audio_track_count; audio_track++) {
XMVAudioPacket *packet = &xmv->audio[audio_track];
if (avio_read(pb, data, 4) != 4)
return AVERROR(EIO);
packet->data_size = AV_RL32(data) & 0x007FFFFF;
if ((packet->data_size == 0) && (audio_track != 0))
/* This happens when I create an XMV with several identical audio
* streams. From the size calculations, duplicating the previous
* stream's size works out, but the track data itself is silent.
* Maybe this should also redirect the offset to the previous track?
*/
packet->data_size = xmv->audio[audio_track - 1].data_size;
/** Carve up the audio data in frame_count slices */
packet->frame_size = packet->data_size / xmv->video.frame_count;
packet->frame_size -= packet->frame_size % packet->track->block_align;
}
/* Packet data offsets */
data_offset = avio_tell(pb);
xmv->video.data_offset = data_offset;
data_offset += xmv->video.data_size;
for (audio_track = 0; audio_track < xmv->audio_track_count; audio_track++) {
xmv->audio[audio_track].data_offset = data_offset;
data_offset += xmv->audio[audio_track].data_size;
}
/* Video frames header */
/* Read new video extra data */
if (xmv->video.data_size > 0) {
if (xmv->video.has_extradata) {
xmv_read_extradata(xmv->video.extradata, pb);
xmv->video.data_size -= 4;
xmv->video.data_offset += 4;
if (xmv->video.stream_index >= 0) {
AVStream *vst = s->streams[xmv->video.stream_index];
assert(xmv->video.stream_index < s->nb_streams);
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
2014-06-18 18:42:52 +00:00
if (vst->codecpar->extradata_size < 4) {
av_free(vst->codecpar->extradata);
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
2014-06-18 18:42:52 +00:00
vst->codecpar->extradata =
av_malloc(4 + AV_INPUT_BUFFER_PADDING_SIZE);
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
2014-06-18 18:42:52 +00:00
vst->codecpar->extradata_size = 4;
}
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
2014-06-18 18:42:52 +00:00
memcpy(vst->codecpar->extradata, xmv->video.extradata, 4);
}
}
}
return 0;
}
static int xmv_fetch_new_packet(AVFormatContext *s)
{
XMVDemuxContext *xmv = s->priv_data;
AVIOContext *pb = s->pb;
int result;
/* Seek to it */
xmv->this_packet_offset = xmv->next_packet_offset;
if (avio_seek(pb, xmv->this_packet_offset, SEEK_SET) != xmv->this_packet_offset)
return AVERROR(EIO);
/* Update the size */
xmv->this_packet_size = xmv->next_packet_size;
if (xmv->this_packet_size < (12 + xmv->audio_track_count * 4))
return AVERROR(EIO);
/* Process the header */
result = xmv_process_packet_header(s);
if (result)
return result;
/* Update the offset */
xmv->next_packet_offset = xmv->this_packet_offset + xmv->this_packet_size;
return 0;
}
static int xmv_fetch_audio_packet(AVFormatContext *s,
AVPacket *pkt, uint32_t stream)
{
XMVDemuxContext *xmv = s->priv_data;
AVIOContext *pb = s->pb;
XMVAudioPacket *audio = &xmv->audio[stream];
uint32_t data_size;
uint32_t block_count;
int result;
/* Seek to it */
if (avio_seek(pb, audio->data_offset, SEEK_SET) != audio->data_offset)
return AVERROR(EIO);
if ((xmv->video.current_frame + 1) < xmv->video.frame_count)
/* Not the last frame, get at most frame_size bytes. */
data_size = FFMIN(audio->frame_size, audio->data_size);
else
/* Last frame, get the rest. */
data_size = audio->data_size;
/* Read the packet */
result = av_get_packet(pb, pkt, data_size);
if (result <= 0)
return result;
pkt->stream_index = audio->stream_index;
/* Calculate the PTS */
block_count = data_size / audio->track->block_align;
pkt->duration = block_count;
pkt->pts = audio->block_count;
pkt->dts = AV_NOPTS_VALUE;
audio->block_count += block_count;
/* Advance offset */
audio->data_size -= data_size;
audio->data_offset += data_size;
return 0;
}
static int xmv_fetch_video_packet(AVFormatContext *s,
AVPacket *pkt)
{
XMVDemuxContext *xmv = s->priv_data;
AVIOContext *pb = s->pb;
XMVVideoPacket *video = &xmv->video;
int result;
uint32_t frame_header;
uint32_t frame_size, frame_timestamp;
uint32_t i;
/* Seek to it */
if (avio_seek(pb, video->data_offset, SEEK_SET) != video->data_offset)
return AVERROR(EIO);
/* Read the frame header */
frame_header = avio_rl32(pb);
frame_size = (frame_header & 0x1FFFF) * 4 + 4;
frame_timestamp = (frame_header >> 17);
if ((frame_size + 4) > video->data_size)
return AVERROR(EIO);
/* Create the packet */
result = av_new_packet(pkt, frame_size);
if (result)
return result;
/* Contrary to normal WMV2 video, the bit stream in XMV's
* WMV2 is little-endian.
* TODO: This manual swap is of course suboptimal.
*/
for (i = 0; i < frame_size; i += 4)
AV_WB32(pkt->data + i, avio_rl32(pb));
pkt->stream_index = video->stream_index;
/* Calculate the PTS */
video->last_pts = frame_timestamp + video->pts;
pkt->duration = 0;
pkt->pts = video->last_pts;
pkt->dts = AV_NOPTS_VALUE;
video->pts += frame_timestamp;
/* Keyframe? */
pkt->flags = (pkt->data[0] & 0x80) ? 0 : AV_PKT_FLAG_KEY;
/* Advance offset */
video->data_size -= frame_size + 4;
video->data_offset += frame_size + 4;
return 0;
}
static int xmv_read_packet(AVFormatContext *s,
AVPacket *pkt)
{
XMVDemuxContext *xmv = s->priv_data;
int result;
if (xmv->video.current_frame == xmv->video.frame_count) {
/* No frames left in this packet, so we fetch a new one */
result = xmv_fetch_new_packet(s);
if (result)
return result;
}
if (xmv->current_stream == 0) {
/* Fetch a video frame */
result = xmv_fetch_video_packet(s, pkt);
if (result)
return result;
} else {
/* Fetch an audio frame */
result = xmv_fetch_audio_packet(s, pkt, xmv->current_stream - 1);
if (result)
return result;
}
/* Increase our counters */
if (++xmv->current_stream >= xmv->stream_count) {
xmv->current_stream = 0;
xmv->video.current_frame += 1;
}
return 0;
}
AVInputFormat ff_xmv_demuxer = {
.name = "xmv",
.long_name = NULL_IF_CONFIG_SMALL("Microsoft XMV"),
.priv_data_size = sizeof(XMVDemuxContext),
.read_probe = xmv_probe,
.read_header = xmv_read_header,
.read_packet = xmv_read_packet,
.read_close = xmv_read_close,
};