v0.5.0

wf-recorder v0.5.0 🎉

Highlight of this version is improved audio capture support, most notably having a pipewire capture backend. There were also improvements to dmabuf support, compatibility with FFMPEG 7.0 and other smaller fixes. See the changelog below for full information.

What's Changed

• Also set buffer.drm_format with shm capture by @nowrep in #233
• Use dmabuf feedback instead of wl-drm to get compositor device by @nowrep in #235
• README: add NixOS, order distros alphabetically by @yu-re-ka in #231
• Set the default pixel format to yuv420p by @soreau in #237
• Get hw_frame_context from buffersink by @nowrep in #242
• Allow region capture with DMA-BUF by @nowrep in #244
• recreate buffer if format changes by @ammen99 in #251
• Add compatibility with FFMPEG 7.0 by @eclipseo in #262
• ensure all status messages use stderr by @nekopsykose in #267
• add -O/--overwrite by @nekopsykose in #266
• PipeWire audio capture by @nowrep in #247
• Require libpipewire >= 1.0.5 for pw_buffer.time by @nowrep in #271

New Contributors

• @yu-re-ka made their first contribution in #231
• @eclipseo made their first contribution in #262
• @nekopsykose made their first contribution in #267

Full Changelog: v0.4.1...v0.5.0

wf-recorder 0.4.1

This is a small bug-fix release which fixes builds from release tarballs. See #226, #230

What's Changed

• Skip git-based version if building from tarballs by @WhyNotHugo in #226
• readme: mention Alpine package by @WhyNotHugo in #227
• Fix typo in main.cpp by @b1rger in #229

Full Changelog: v0.4.0...v0.4.1

wf-recorder 0.4.0

This is the 0.4.0 release of wf-recorder.

Most important user-facing changes since 0.3.0:

• Audio codec, sample rate and format can be changed at runtime (#177)
• Setting buffer/frame rate (#190)
• DMA-BUF support (significantly reduces CPU usage when using GPU encoding) (#206)
• Configurable default codec and container format (#198)

There have also been many other changes and bug fixes, see the list below.
Huge thanks to all contributors and testers!

What's Changed

• Support for configuring audio codec and video framerate by @llyyr in #177
• Remove unused variable by @WhyNotHugo in #178
• docs: fix help and examples regarding -a & --audio by @llyyr in #180
• fix buffersink pix_fmt handling by @llyyr in #182
• Add default codec parameters for libvpx/libvpx-vp9. by @alebastr in #197
• Use VP9+Opus on MKV by default by @Conan-Kudo in #198
• build: Fix compile failure introduced by last commit by @soreau in #200
• README: Flesh out build dependencies for building from source by @Conan-Kudo in #203
• Add GitHub Actions CI to test the build by @Conan-Kudo in #201
• Use VP8+Vorbis on WebM for recordings by @Conan-Kudo in #202
• FrameRate Changes, Implement CFR and estimated buffer framerate by @Quackdoc in #190
• Fix deprecation warnings by @llyyr in #193
• Gracefully terminate on SIGTERM/SIGHUP to avoid corruption by @joanbm in #183
• ci: add to safe directories so that git parse-rev works by @ammen99 in #210
• Only flush when video codec has delay by @nowrep in #209
• Use DMA-BUF with HW encoding by @nowrep in #206
• Reduce memory usage with DMA-BUF capture by @nowrep in #219
• Use gbm_bo_create_with_modifiers + don't encode if capture fails by @nowrep in #220
• Use codec format which best matches input by @mstoeckl in #215
• Go back to libx264 by default by @ammen99 in #222
• Add support for 16 bit deep formats by @mstoeckl in #184
• Add X2RGB10 format mapping for DMA-BUFs by @nowrep in #224
• Dynamically increase number of used buffers by @nowrep in #221
• Break early when exit_main_loop is set in encoder thread by @nowrep in #223
• add --no-dmabuf option by @ammen99 in #225
• don't ask for overwrite for the char device by @denisok in #141

New Contributors

• @llyyr made their first contribution in #177
• @WhyNotHugo made their first contribution in #178
• @alebastr made their first contribution in #197
• @Conan-Kudo made their first contribution in #198
• @Quackdoc made their first contribution in #190
• @joanbm made their first contribution in #183
• @nowrep made their first contribution in #209
• @denisok made their first contribution in #141

Full Changelog: v0.3.0...v0.4.0

wf-recorder 0.3.0

This is the 0.3.0 release of wf-recorder. Notable new features since 0.2.1:

• Support for ffmpeg filters with the -F <filter> option (5384bc9). This allows using many of the effects ffmpeg provides, see https://github.com/schauveau/wf-recorder-x for a good introduction on filters.

• Support for recording new frames only when the screen contents actually change (#82). This is now the default behavior. If this is not desired (because for ex. the recording has variable refresh rate and not fixed), the -D option can be used to get the old behavior.

• wf-recorder now asks the user if the output file will be overwritten (0b74ad4)

• Various bug fixes, see the git log for complete list of changes.

wf-recorder 0.2.1

wf-recorder 0.2.1 is a bug-fix release over wf-recorder 0.2 with no new features added. Changes include:

1. Improved documentation
2. Fixed long recordings (>1h10m) (#78)

wf-recorder 0.2

This is the second release of wf-recorder.

Notable changes since v0.1:

1. Addition of manpage and help options #35
2. Conversion to RGB to YUV with the help of OpenCL (the latter being optional) #30 #32
3. Option to set muxer and pixel format #43 #53
4. Support for loopback devices: #43 #53
5. Update to newer avcodec_send_frame API #47

Thanks to all the people who contributed code and effort into this release!

Initial release

This is the first release of wf-recorder. Supported features:

• Recording an output or a part of an output in wlroots-based compositors
• Various file formats & codecs (mkv, mp4, gif, ...)
• Experimental VAAPI encoding
• Audio recording via PulseAudio