The GStreamer team is proud to announce a new major feature release of your favourite cross-platform multimedia framework!
The 1.14 release series adds new features on top of the previous 1.12 series and is part of the API and ABI-stable 1.x release series of the GStreamer multimedia framework.
Highlights:
WebRTC support: real-time audio/video streaming to and from web browsers
Experimental support for the next-gen royalty-free AV1 video codec
Video4Linux: encoding support, stable element names and faster device probing
Support for the Secure Reliable Transport (SRT) video streaming protocol
RTP Forward Error Correction (FEC) support (ULPFEC)
RTSP 2.0 support in rtspsrc and gst-rtsp-server
ONVIF audio backchannel support in gst-rtsp-server and rtspsrc
playbin3 gapless playback and pre-buffering support
tee, our stream splitter/duplication element, now does allocation query aggregation which is important for efficient data handling and zero-copy
QuickTime muxer has a new prefill recording mode that allows file import in Adobe Premiere and FinalCut Pro while the file is still being written.
rtpjitterbuffer fast-start mode and timestamp offset adjustment smoothing
souphttpsrc connection sharing, which allows for connection reuse, cookie sharing, etc.
nvdec: new plugin for hardware-accelerated video decoding using the NVIDIA NVDEC API
Adaptive DASH trick play support
ipcpipeline: new plugin that allows splitting a pipeline across multiple processes
Major gobject-introspection annotation improvements for large parts of the library API
GStreamer C# bindings have been revived and seen many updates and fixes
The externally-maintained GStreamer Rust bindings have many usability improvements and cover most of the API now
There briefly was a 1.25.9 release which was superseeded by 1.25.10 before a public announcement. Both amount to these fixes:
libout123: Fix error messages beginning from OUT123_ARG_ERROR (bug 261).
mpg123: Fix --icy-interval handling to work with stream from stdin. (curl | mpg123 --icy-interval=n -)
libmpg123: Fix another invalid read and segfault on damaged (fuzzed) files with part2_3_length == 0 (set maxband=1, pulled from upcoming 1.26.0).
PyAudio provides Python bindings for PortAudio, the cross-platform
audio I/O library. With PyAudio, you can easily use Python to play
and record audio on a variety of platforms, such as GNU/Linux,
Microsoft Windows, and Apple Mac OS X / macOS.
* Add the GConf missing dependency
* Patch the default cdrom device to /dev/rcd0d for NetBSD
Changes since 3.6.1:
v3.7.0
* Sync Brazilian Portuguese from Translation Project
* Patch by Phil Stracchino: extended genres to genre list
v3.6.3
* Sync ukranian and hungarian translations from Translation Project
* Patch from Adrian Reber: fix compiler warning with gcc 8.0
* Update the FSF postal address in source files
v3.6.2
* Translation updates: German, Italian, Brazilian Portuguese, Serbian
This fixes an out of bound write, but possibly many other changes.
Unfortunately upstream doesn't do releases for tremor, so I manually
fetch the tarball and uploaded it.
abcde 2.9.1
* Merged Debian packaging files updates
* Fixups in the new Musicbrainz code:
+ Cope with not finding an entry for a CD!
+ Fix (again!) dealing with CDs with no release events
-- Steve McIntyre <93sam@debian.org> Fri, 09 Mar 2018 23:01:04 +0000
abcde 2.9
* Add some documentation for the aged CD ripper dagrab. Thanks to
Teika Kazura for the notification and suggested documentation. This
closes Issue 50: https://abcde.einval.com/bugzilla/show_bug.cgi?id=50
* Work by Matthias Andree <matthias.andree@gmx.de> to address the issue
where abcde fails with accented characters from CD-TEXT. The issue and
partial fix applied here documented in Issue 53:
https://abcde.einval.com/bugzilla/show_bug.cgi?id=53
and also in the abcde mailing list:
https://lists.einval.com/pipermail/abcde-users/2017-January/000232.html
* Allow for embedding of album art downloaded by the getalbumart function.
Currently this is available for flac (using metaflac), mp3 (using eyed3),
m4a (using AtomicParsley) and WavPack aka wv (using wvtag).
This can be invoked in 3 ways:
1. Use the commandline '-B' option (this will also call getalbumart)
2. Use the commandline '-a embedalbumart' option to add to list of actions
3. Use 'embedalbumart' in the 'ACTIONS' list in ~/.abcde.conf
Still needs more development but it is perfectly usable at the moment!
* Use md5 rather than md5sum under macOS. Thanks to JCount for the bug
report and also the fix. This solves Issue 59:
https://abcde.einval.com/bugzilla/show_bug.cgi?id=59
* Support added for encoding with Audio Interchange File Format (AIFF).
Thanks to Massimo Villa for the feature request. FFmpeg is required
for the encoding, the container and suffix are 'aiff'.
* Allow selection of either cddb or musicbrainz from the command line:
-Q Select CDDBMETHOD from the command line. Choice is cddb or musicbrainz.
Command line letters are fast running out but the 'Q' option quite neatly
stands for 'Query'!
* Allow for embedding with do_embedalbumart() for single track encodes when
OUTPUTFORMAT and ONETRACKOUTPUTFORMAT are different. Thanks to Ashley Gittins
for the bug report. This closes Issue 63:
https://abcde.einval.com/bugzilla/show_bug.cgi?id=63
* Experimental support for embedding albumart with ogg files. The slightly
tortuous technique drawn from 2 sources:
1. https://github.com/biapy/howto.biapy.com/blob/master/various/mussync-tools
2. https://github.com/acabal/scripts/blob/master/ogg-cover-art
Testing is strongly encouraged, perhaps the simplest way to test is with:
abcde -o ogg -B
Or the appropriate settings in an ~/.abcde.conf file.
* Massive rework of CD lookup code so support multiple sources
better. Thanks to Gabriel Rosenkoetter for his initial idea in this
area, and to Tomasz Goliński on irc for initial inspiration on how
this should work better.
There are now 3 different options for CD lookup: cddb, musicbrainz and
cdtext. They can all be listed in a comma-separated list for
CDDBMETHOD and the code will now call all of them in the sequence
listed. All the results will be combined into one list at the end for
the user to select, just like would have previously worked for one
source only. Closes Issue 42:
https://abcde.einval.com/bugzilla/show_bug.cgi?id=42
* Fix Musicbrainz ID calculation in makeids()
Apply fix suggested by petecollins24@gmail.com; add PREGAP to
LEADOUT to correct Musicbrainz ID calculation. Hopefully closes
Issue 54: https://abcde.einval.com/bugzilla/show_bug.cgi?id=54
* Fix abcde.mkcue() when handling the --wholedisk option. Thanks to
Nino Burini for the patch. Closes Issue 47 and maybe also 45:
https://abcde.einval.com/bugzilla/show_bug.cgi?id=47https://abcde.einval.com/bugzilla/show_bug.cgi?id=45
* Add more comprehensive examples for filename munging in the example
config file. Closes Issue 49:
https://abcde.einval.com/bugzilla/show_bug.cgi?id=49
* Add a more better fix for the year lookup problem in abcde-musicbrainz-tool
Thanks to Tom Samstag for the patch. Closes Issue 57:
https://abcde.einval.com/bugzilla/show_bug.cgi?id=57
* Redirect stderr on "which" calls to clear up error noise on some
systems. Closes Issue 56:
https://abcde.einval.com/bugzilla/show_bug.cgi?id=56
* Add Irix support, based on a patch by abcde@canavan.de. Closes Issue 29:
https://abcde.einval.com/bugzilla/show_bug.cgi?id=29
* Rework abcde-musicbrainz-tool to work with WebService::MusicBrainz 1.x
Thanks to Nicolas Guillaumin for the patch. Closes Issue 60:
https://abcde.einval.com/bugzilla/show_bug.cgi?id=60
Also added a specific dependency on version 1.0.4 or newer, so
abcde-musicbrainz-tool will abort if the version found is too old.
Since tweaked to deal with multi-artist albums and tracks more
completely.
* Add resume support in do_getalbumart()
* Fix getopts setup for "P". Thanks to Alan W. Kerr for reporting this.
* Major set of code cleanups to fix up lots of warnings found with
shellcheck, and a few other style issues:
+ Lots of quoting fixes
+ Use $( instead of `
+ Stop using -o and -a syntax with if [ - use || or && instead
+ Wrap and shorten some very long lines
* Wrap some output messages so they fit on a standard width console
* Factor out repeated code and make page() more useful
* Add version check before resuming from an old ripping run
aeneas is a Python/C library and a set of tools to automagically
synchronize audio and text.
aeneas automatically generates a synchronization map between a list
of text fragments and an audio file containing the narration of
the text. In computer science this task is known as (automatically
computing a) forced alignment.
0.29 (2018/02/11)
* implement %prio% for --format
* the default --format falls back to "performer" if there is no "artist"
* add command "prio"
* command "outputs" prints attributes
* fix build failure with gcc 7.0 and clang 4.0
* build with Meson instead of autotools
pkgsrc changes:
- reuse the github framework
- disable -Werror
- set -D_NETBSD_SOURCE for NetBSD, this is required for obtaining
pthread_condattr_setclock(3) and pthread_condattr_getclock(3) on NetBSD
compiled on version <8.0 (bug in a header)
- switch from ffmpeg2 to ffmpeg3
- add a workaround for a detector of ffmpeg in a homegrown dependency
detector, this hardcodes paths to headers and libraries of ffmpeg3
upstream changelog
==================
Version 4.3.0 (2015-05-25)
player: Add support for "exact mode" - re-opens the audio device when a buffer is encountered that does not match the open audio device's sound parameters.
build: Simplify cmake find modules.
Add const qualifier to some fields and functions.
metadata example: print the number of channels.
Add documentation.
Add more channel layouts.
build: Compile in Debug mode by default.
Version 4.2.1 (2014-10-07)
updating metadata: set time_base on stream not the codec
libav lockmgr: set mutex pointer to NULL on destroy
fix build on GNU/hurd
Version 4.2.0 (2014-09-25)
build: remove bundled dependencies
build: simpler cmake find modules, both code and license
player: add shim to fix build failure on OSX
Version 4.1.1 (2014-06-20)
playlist: fix race condition which can cause decoder to hang
dummy player: fix timing issues
Version 4.1.0 (2014-06-13)
playlist: added groove_playlist_set_fill_mode. Allows you to choose between buffering until all sinks are full and buffering until any sinks are full.
Version 4.0.4 (2014-06-03)
Fixed a race condition where seeking to a different playlist item and then back would have a window of time where it could report the wrong position.
Properly play and pause network streams.
Version 4.0.3 (2014-05-31)
build: update bundled libav to latest stable 10 release
build: link player with -lrt for clock_gettime. closes#67
playlist: fix case where filter graph was not being rebuilt. closes#65
playlist: fix race condition segfault when attaching a sink
encoder: properly reset encoding when flush or playlist end is encountered. closes#66
Version 4.0.2 (2014-05-20)
player: thread cleanup only if thread was initialized - fixes potential crash on player detach
build: look for includes in the current source tree. Fixes an issue when a previous version of the library is installed.
build: on unix link with -lm
Version 4.0.1 (2014-05-13)
groove_playlist_get_position: always set seconds even when item is NULL
playlist: correct generation of the sink map
fixes potential error when adding multiple sinks
optimizes some cases where sinks can share filter graph chain
dummy player now uses disable_resample optimization
dummy player: avoid floating point error accumulation
Version 4.0.0 (2014-05-12)
GrooveBuffer struct contains the presentation time stamp
move include statements to outside of extern C
ability to set true peak on playlist items. closes#50
support per-sink gain adjustment. closes#41
GroovePlaylist: volume renamed to gain
groove_playlist_set_gain renamed to groove_playlist_set_item_gain
groove_playlist_set_volume renamed to groove_playlist_set_gain
player: specify device by index rather than name. closes#44
player: ability to attach a dummy player device. closes#60
fingerprinter: encode/decode return 0 on success, < 0 on error
fingerprinter: info struct contains raw fingerprint instead of compressed string. closes#61
pkgsrc changes:
- switch to GitHub as the release source
- set CFLAGS for NetBSD: _NETBSD_SOURCE, this works around a bug that
strdup(3) was not included in the POSIX source namespace (fixed in
NetBSD 8.0)
- switch to ffmpeg 3.x
upstream changelog
==================
Release 2017.08.30
- Support binding to specific network interface, see config option bind_to
- New config option gain_mul softens effect of replay gain
- Replaced deprecated ffmpeg and json-c functions, bumps ffmpeg/libav version
requirement
- Fix use-after-free
- Minor improvements to code and documentation
Release 2016.06.02
- Quit pianobar upon reception of two SIGINT
- Minor bugfixes
ver 0.20.16 (2018/02/03)
* output
- pulse: fix crash during auto-detection
* database
- simple: fix search within mount points
- upnp: enable IPv6
* archive
- iso9660: libcdio 2.0 compatibility
* fix crash in debug build on Haiku and other operating systems
This switch is meant to be used by packages requiring an implementation of the
former libusb (as in devel/libusb). The original implementation can be
chosen by setting LIBUSB_TYPE to "native".
The alternative implementation libusb-compat (as in devel/libusb-compat) wraps
libusb1 (in devel/libusb1). This implementation can be chosen by setting
LIBUSB_TYPE to "compat". On NetBSD, it has the advantage of not requiring root
privileges to locate and use USB devices without a kernel driver.
This second part switches packages using libusb to this framework. It does not
change compilation options or dependencies at this point.
Compile-tested on most packages affected and available on NetBSD/amd64.
- audacity: GCC 6 wants std namespace, found via <cmath>.
- smpeg2: don't initialise unsigned with "0-1".
- libclucene/strigi: match GCC 6 (and 7, and 8, and 9.)
