Changes in Avidemux v2.7.6 – v2.7.8
New Features
Many new video filters added by courtesy of
szlldm A dedicated MOV muxer is now available, replacing silent fallback to
MOV muxing mode within the MP4 muxer.
A subset of color info relevant for HDR support is retained in copy
mode when both input and output video are stored in Matroska / WebM
containers. Settings to add or override color info have been added to
MKV and WebM muxers.
The indexer in the MPEG-TS demuxer now detects resolution changes in
MPEG-2 and H.264 streams which are entirely unsupported and suggests
aborting indexing. If the user chooses to continue, it records file
offsets of all resolution changes to the log. This allows advanced
users to split the file at these offsets using other tools like
head and tail into chunks which can be processed
with Avidemux.
YUV varieties of the lossless Ut Video codec are supported now for
decoding and encoding via bundled FFmpeg libraries.
Uncompressed 24-bit audio is now supported for a subset of file formats.
Limited support for 7.1 channel layout.
Decode and stream copy support for XLL substream in DTS audio tracks.
Audio properties are now updated on resuming playback. This allows to
start saving after an audio format change. When audio format changes
during playback, the FFmpeg-based audio decoder will output silence
and keep format rather than producing garbage.
An option to keep the identity of markers A and B instead of swapping
them automatically has been added to Preferences.
The keyboard shortcut for "Delete" action has been made customizable
with alternative keyboard shortcuts enabled.
Other Enhancements
Multi-threaded video decoding is now available for the bundled FFmpeg.
On powerful multi-core CPUs, this can drastically improve decoding
performance. A maximum of 8 threads can be created, but a conservative
custom upper limit of 4 is recommended. Values above 8 cause decoding
failures. Multi-threaded decoding and HW accelerated decoding are
mutually exclusive, changes require application restart to have
effect.
The maximum supported resolution has been increased to 8192x8192
The bundled FFmpeg has been updated to the 4.2.4 release.
Non-standard display aspect ratios can be configured in MP4 muxer by
specifying display width.
MP4 muxer accepts now MPEG-1 and MPEG-2.
The configuration dialog of the x265 video encoder plugin enjoyed
massive cleanup and refactoring by
Trent Piepho , more Adaptive
Quatization Variance methods have been added.
Python interface has been substantially extended, includes now methods
to query segment properties, the number and filenames of loaded
videos, PTS of keyframes, methods to seek to time with preview
updated, to open file dialogs filtered by extension and a built-in
method to split filename extension from a given path.
Scripting shell now supports unicode characters.
On macOS, files can be loaded via Finder context menu or by dropping
them onto Avidemux icon in the dock.
On Linux, Jobs GUI and the CLI executable can be launched from
AppImage by creating a symlink with _jobs or _cli in filename
respectively and running the AppImage via this symlink.
Bugfixes
Rate control was broken in some video encoders, especially in VP9 and
in NVENC-based encoders, fixed now.
Multiple issues in cut point checks for H.264 and HEVC streams in copy
mode, leading to incorrect check results, were identified and
resolved.
The first two frames of external AC3 or EAC3 audio tracks are not
dropped anymore. This regression from enhanced protection against
false positives, added 3 years ago, has been finally identified and
fixed.
Indexing of MPEG-TS files with HEVC video was broken when video stream
lacked access unit delimiters. The issue was identified and fixed
right after 2.7.7 had been tagged as released and is the reason why
version 2.7.7 was superseded by 2.7.8.
VU meter was grossly ahead of audio output on Windows, now it is just
a tiny little bit ahead.
Numerous other fixes and overall polish.
