As previously discussed, the CrystalHD hardware returns exceptionally
useless information about interlaced h.264 content - to the extent
that it's not possible to distinguish most MBAFF and PAFF content until
it's too late.
In an attempt to compensate for this, I'm introducing two mechanisms:
1) Peeking at the picture number of the next picture
The hardware provides a capability to peek the next picture number. If
it is the same as the current picture number, then we are clearly dealing
with two fields and not a frame or fieldpair.
If this always worked, it would be all we need, but it's not guaranteed
to work. Sometimes, the next picture may not be decoded sufficiently
for the number to be known; alternately, a corruption in the stream may
cause the hardware to refuse to return the number even if the next
intact frame is decoded. In either case, the query will return 0.
If we are unable to peek the next picture number, we assume that the
picture is a frame/fieldpair and return it accordingly. If that turns
out to be incorrect, we discard the second field, and the user has
to live with the glitch. In testing, false detection can occur for
the first couple of seconds, and then the pipeline stabalizes and
we get correct detection.
2) Use the h264_parser to detect when individual input fields have
been combined into an output fieldpair.
I have multiple PAFF samples where this behaviour is detected. The
peeking mechanism described above will correctly detect that the
output is a fieldpair, but we need to know what the input type was
to ensure pipeline stability (only return one output frame per input
frame).
If we find ourselves with an output fieldpair, yet the input picture
type was a field, as reported by the parser, then we are dealing with
this case, and can make sure not to return anything on the next
decode() call.
Taken together, these allow us to remove the hard-coded hacks for
different h.264 types, and we can clearly describe the conditions
under which we can trust the hardware's claim that content is
interlaced.
Signed-off-by: Philip Langdale <philipl@overt.org>
Now that we know the type of the input picture, we have to bring
that information to the output picture to help identify its type.
We do this by adding a field to the opaque_list node.
Signed-off-by: Philip Langdale <philipl@overt.org>
As the hardware is unreliable, we will have to use the h.264 parser
to identify whether an input picture is a field or a frame. This
change loads the parser and extracts the picture type.
Signed-off-by: Philip Langdale <philipl@overt.org>
In preparation for adding additional fields to the node, return
the node instead of the pts value. This requires the caller to
free the node.
Signed-off-by: Philip Langdale <philipl@overt.org>
I found another MBAFF sample where the input:output pattern is
the same as mpeg2 and vc1 (fieldpair input, individual field output).
While I'm not sure how you can output individual fields from MBAFF,
if I apply the mpeg2/vc1 handling to this file, it plays correctly.
So, this changes the detection algorithm to handle the known cases.
Whitespace will be fixed in a separate change.
Signed-off-by: Philip Langdale <philipl@overt.org>
* ffmpeg-mt/master:
DUPLICATE mingw32 compilation after 'unbreak avcodec_thread_init'
pthread: validate_thread_parameters() ignored slice-threading being intentionally off
DUPLICATE Remove unnecessary parameter from ff_thread_init() and fix behavior
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
proto: include os_support.h in network.h
matroskaenc: don't write an empty Cues element.
lavc: add a FF_API_REQUEST_CHANNELS deprecation macro
avio: move extern url_interrupt_cb declaration from avio.h to url.h
avio: make av_register_protocol2 internal.
avio: avio_ prefix for url_set_interrupt_cb.
avio: AVIO_ prefixes for URL_ open flags.
proto: introduce listen option in tcp
doc: clarify configure features
proto: factor ff_network_wait_fd and use it on udp
Conflicts:
ffmpeg.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
Fix parser: mark av_parser_parse() for removal on next major bump
swscale: postpone sws_getContext removal until next major bump.
fate: add AAC LATM test
mmst: get rid of deprecated AVERRORs
lxfdec: use AVERROR(ENOMEM) instead of deprecated AVERROR_NOMEM.
Reemove remaining uses of deprecated AVERROR_NOTSUPP.
REIMPLEMENTED in 2 lines of code: lavf: if id3v2 tag is present and all else fails, guess by file extension
Conflicts:
libavformat/utils.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
psymodel: extend API to include PE and bit allocation.
avio: always compile dyn_buf functions
Remove unnecessary parameter from ff_thread_init() and fix behavior
Revert "aac_latm_dec: use aac context and aac m4ac"
configure: tell user if libva is enabled like the rest of external libs.
Add silence support for AV_SAMPLE_FMT_U8.
avio: make URL_PROTOCOL_FLAG_NESTED_SCHEME internal
avio: deprecate av_url_read_seek
avio: deprecate av_url_read_pause
ac3enc: NEON optimised extract_exponents
Conflicts:
libavcodec/utils.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
CONFIG_VDPAU is the condition on which ff_vdpau_mpeg_picture_complete
is compiled in, so it is more appropriate, particularly since the
separate VDPAU decoder should be removed in the longer term.
thread_count passed to ff_thread_init() is only used to set AVCodecContext.
thread_count, and can be removed. Instead move it to the legacy implementation
of avcodec_thread_init().
This also fixes the problem that calling avcodec_thread_init() with pthreads
enabled did not set it since ff1efc524cb3c60f2f746e3b4550bb1a86c65316.
Signed-off-by: Janne Grunau <janne-libav@jannau.net>
Currently, the parser is buggy and only processes the stream extradata
when the flag is set. This fixes it to actually inspect the frames.
Whitespce will be fixed in a separate change.
Signed-off-by: Philip Langdale <philipl@overt.org>
* qatar/master: (22 commits)
ac3enc: move extract_exponents inner loop to ac3dsp
avio: deprecate url_get_filename().
avio: deprecate url_max_packet_size().
avio: make url_get_file_handle() internal.
avio: make url_filesize() internal.
avio: make url_close() internal.
avio: make url_seek() internal.
avio: cosmetics, move AVSEEK_SIZE/FORCE declarations together
avio: make url_write() internal.
avio: make url_read_complete() internal.
avio: make url_read() internal.
avio: make url_open() internal.
avio: make url_connect internal.
avio: make url_alloc internal.
applehttp: Merge two for loops
applehttp: Restructure the demuxer to use a custom AVIOContext
applehttp: Move finished and target_duration to the variant struct
aacenc: reduce the number of loop index variables
avio: deprecate url_open_protocol
avio: deprecate url_poll and URLPollEntry
...
Conflicts:
libavformat/applehttp.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
The pulldown flags should be communicated to the client of the libavcodec library. Not doing so causes jerky playback with pulldown content. Note that this change requires the patch previously provided here: http://ffmpeg.org/pipermail/ffmpeg-devel/2011-April/110314.html
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
The attached patch fixes the jerky playback of VC-1 content with pulldown. The pulldown flags were incorrectly set. They must be correct in order to display the frames with the correct timing as mentioned in the specifications: "SMPTE 421M: VC-1 Compressed Video Bitstream Format and Decoding Process". More precisely the following tables:
Table 20: Progressive P picture layer bitstream for Advanced Profile
Table 22: Progressive B picture layer bitstream for Advanced Profile
Table 23: Progressive Skipped picture layer bitstream for Advanced Profile
Table 82: Interlaced Frame I and BI picture layer bitstream for Advanced Profile
Table 83: Interlaced Frame P picture layer bitstream for Advanced Profile
Table 84: Interlaced Frame B picture layer bitstream for Advanced Profile
Table 85: Picture Layer bitstream for Field 1 of Interlace Field Picture for Advanced Profile
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
fate: fix partial run when no samples path is specified
ARM: NEON fixed-point forward MDCT
ARM: NEON fixed-point FFT
lavf: bump minor version and add an APIChanges entry for avio changes
avio: simplify url_open_dyn_buf_internal by using avio_alloc_context()
avio: make url_fdopen internal.
avio: make url_open_dyn_packet_buf internal.
avio: avio_ prefix for url_close_dyn_buf
avio: avio_ prefix for url_open_dyn_buf
avio: introduce an AVIOContext.seekable field
ac3enc: use generic fixed-point mdct
lavfi: add fade filter
Change yadif to not use out of picture lines.
lavc: deprecate AVCodecContext.antialias_algo
lavc: mark mb_qmin/mb_qmax for removal on next major bump.
Conflicts:
doc/filters.texi
libavcodec/ac3enc_fixed.h
libavcodec/ac3enc_float.h
libavfilter/Makefile
libavfilter/allfilters.c
libavfilter/vf_fade.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
This makes the AC3 encoder use the shared fixed-point MDCT rather
than its own implementation. The checksum changes are due to
different rounding in the MDCT.
Signed-off-by: Mans Rullgard <mans@mansr.com>
commit c0ec9918b066c6bfca8613ac75b589bd3e405971
Author: Måns Rullgård <mans@mansr.com>
Date: Tue Aug 24 17:47:05 2010 +0000
Remove global mm_flags variable
Originally committed as revision 24909 to svn://svn.ffmpeg.org/ffmpeg/trunk
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
