2.4

Important Note

This release is identical to the 2.3, but we had to revert #366 since it caused freezing issues when using GDK_SCALE

2.3 Release Notes, adjusted for the reverted PR

Presets are now once again the default view. Also we've added support for AJA capture cards.

What's Changed

• add support for AJA capture cards by @lukegb in #353
• Modernize build setup & tests by @justjanne in #362
• Type Checking by @justjanne in #363
• Preset reorder by @sophieschi in #369
• replace python3-sdnotify with vocto.sd_notify, use in voctogui by @Kunsi in #373

New Contributors

• @lukegb made their first contribution in #353

Full Changelog: 2.2...2.4

v2.2

We have reworked the presets generation and fixed some bugs. Find a list below:

What's Changed

• gstreamer release 1.18 deprecated the yadif plugin. by @baldurmen in #331
• Improve voctocore's default config paths by @baldurmen in #332
• TCPAVSource: listen on socket.gethostname() instead of localhost by @olasd in #333
• avdec_h264 requires h264parse in front of it by @stefanor in #345
• Python 3.12 support: use ConfigParser by @stefanor in #343
• The OPUS encoder is called opusenc by @stefanor in #344
• Fix previews for v4l sources by @stefanor in #342
• Fix typo in debug message by @baldurmen in #338
• just enough ini so voctocore can run. by @CarlFK in #337
• Fix audio only sources by @justjanne in #350
• register preset hotkeys by @justjanne in #355
• Fix unstable audio channel order by @justjanne in #358
• only use vaapi decode on the client if vaapi rendering is enabled by @justjanne in #354

The issues #317 (voctomix assumes client is vaapi capable if vaapi is used on the server side) and #359 (custom backgrounds are always decoded as mpegts) were also resolved. It turned out this was a configuration issue in both cases.

New Contributors

• @baldurmen made their first contribution in #331
• @olasd made their first contribution in #333
• @stefanor made their first contribution in #345
• @justjanne made their first contribution in #350

Full Changelog: 2.1.0...2.2

v2.1

Features

• voctogui is able to have preset buttons. The list of presets will get auto-generated by getting the source and composite lists from the core.

Known bugs

The list of v2.0 still applies

v2.0

We are very pleased to announce the release of voctomix 2.0, the first release of a completely rewritten voctomix!

This version of voctomix requires atleast python 3.9!

Notable changes:

• Source handling was reworked, so the past issues with sources slowly drifting over time should no longer appear.
• Voctomix has switched away from the "preset" style mixer to a freely configurable video mixer. You can now freely choose sources for each mix.
  • The missing presets feature has been noted in past events where we have run a prerelease version. Pull request #320 will re-add a tab to add presets to voctogui. It is expected to see this feature in voctomix 2.1
  • The lecture mode has been configured to automatically crop the "A" source to the middle third to allow for a bigger image without compromising on the size of the "B" source.
• Voctomix can now output the mixed video onto a gstreamer sink of your choice. See the documentation for details
• Voctomix now supports a lot more input sources, including support for generic video4linux2 sources and the raspberry pi camera.
• Voctomix can now utilize VAAPI to reduce the amount of CPU load by offloading graphics processing to a connected GPU.
  • Please note that some consumer GPUs have limits on how many VAAPI encodings you can actually run on them. Voctomix runs a bunch of encodings in parallel, so you might run into issues there.
• Voctomix can now display overlays onto a running mix. See the documentation for details

Known bugs:

• Audio issues when inputs have different frame (or sampling?) rates:
  Users are advised to run all inputs that use audio on the same frame and sampling rate. This also includes the blinder source.
• get_composite_modes control command does not work:
  There is no workaround available.
• voctomix assumes client is vaapi capable if vaapi is used on the server side:
  There is no workaround available.
• ControlServer does not broadcast preview changes:
  There is no workaround available. Program changes get broadcasted correctly.
• DPI scaling does not work:
  If you need scaling, you can use the GDK_SCALE environment variable to scale the GUI. GDK_SCALE only supports integers though, so specifying 2 means the GUI elements are double the size.

2.0-rc2

This is the second pre-release of voctomix 2.0 (formerly known as voc2mix while in development).

Basically everything has changed between those versions. There is no migration strategy between 1.x and 2.0.

1.3 "35C3"

This Release contains lots of improvements. This is the Code that ran 35C3.

• Voxtomix now starts with the Stream-Blanker enabled
• option to show VU meters for all audiostreams (thanks @sophieschi)
• Many performance-optimizations in the GUI-Code (thanks @fightling)
• option to force a specific window-size for the Voctogui (thanks @fightling)
• new API-Call to store the List of available Composition-Modes
• new API-Call to read an individual Config-Setting (useful for clients with less RAM)
• new API-Call to store and fetch values in the core (for synchronizing external clients)
• Fixes to Voctolight/Tallylight Code (thanks @NHGmaniac)
• Improvements to local testing scripts (thanks to @hayden-t and @nightvisi0n)

1.2 "GPN18"

This Release contains several improvements in Voctogui introduced by @fightling, that have successfully been tested at GPN18. This includes:

• a Studio-Clock shown in Voctogui, giving an indication of the current Time #200
• Performance-Improvements to the Volume Level-Meter: #197
• Options to force specific Window-Sizes for Voctogui #199

1.1 "34C3"

Oh boy, 5 months since 1.0? Really? Anyway, this Version contains a huge stash of features originally developed for 34C3 – and there's a lot more in the pipeline.

Most notably this release contains the following new additions:

• A Tally-Light add-on that runs on Raspberry-Pis and shows the camera operator, if her/his camera is currently live
• A Midi add-on that connects Midi-Keyboards to Voctocore - thanks @tu500
• Accept 1080PsF Video-Sources, which actually are Progressive but cloak as Interlaced (#174)
• Allow Productions with more than two Audio-Channels, for example Live-Translations (#145) - see Readme for more Documentation
• Option to Crop Video-Sources and have, for example, a square PiP (#179)- thanks @MaKess
• Option to add a Border around Side-by-Side Equal Composition
• Easy to remember Shortcuts for Fullscreen-Buttons (#72)
• A docker-compose-File for a even quicker quickstart (#192) - thanks @nightvisi0n
• Colored Buttons for A/B Source (#193) - thanks @nightvisi0n

1.0 "Finally"

The 1.0 Release is primarily a Bugfix and Cleanup-Release. IT closes a lot of older bug- and smaller feature-Tickets, most notably

• Shortcuts for stream-live/blank #91
• Cleanup of commands.py and the command-interface #98
• Problems with the set_videos_and_composite command and default-sources #100
• Adjustable Volume of Pauseloop #104
• Better handling of missing/unparsable Config-Files #116
• Analyzed and added Options to circumvent Problems with 3.x series ffmpegs #137
• Adjustable Volume of the Pause-Loop-Audio #104
• Example-Scripts for YouTube and Facebook (if you really need to) #170 #161

Furthermore we added a Bunch of Unit-Tests which are also run on travis-ci.

1.0RC1

-- this release is identical to 1.0 --

The 1.0 Release, of which this is the first Release Candidate, is primarily a Bugfix and Cleanup-Release. IT closes a lot of older bug- and smaller feature-Tickets, most notably

• Shortcuts for stream-live/blank #91
• Cleanup of commands.py and the command-interface #98
• Problems with the set_videos_and_composite command and default-sources #100
• Adjustable Volume of Pauseloop #104
• Better handling of missing/unparsable Config-Files #116
• Analyzed and added Options to circumvent Problems with 3.x series ffmpegs #137
• Adjustable Volume of the Pause-Loop-Audio #104
• Example-Scripts for YouTube and Facebook (if you really need to) #170 #161

Furthermore we added a Bunch of Unit-Tests which are also run on travis-ci.