R3 2025 Release Notes

Prev Next

News, updates, and changes in Composer R3 2025 ("Autodrome")

Composer R3 2025

Release date 2025-10-xx, version: xx

Summary:

This version features a brand-new audio mixer view and an updated audio workflow across the application. This version also supports additional file formats, adds SRT input support, and contains an updated video switcher.

News and Improvements:

  • New audio workflow:

    • Added a new audio mixer view that supports mixing audio from multiple inputs and scenes. The audio mixer uses the concept of "channel strips", AUX/BUS, SEND, and MUTE found in typical audio mixers.

    • Channel Strip features:

    • Pre/Post VU-meter monitoring

    • Stereo (Peak & RMS)

    • 8-channel per audio strip

    • 8-channel audio re-mapping

    • Mute, Solo, and pre-fader listening (PFL)

    • Added support for RMS and Peak audio levels for all inputs, operators, scene layers, and scenes.

  • Switcher input:

    • Added support for audio (cut/crossfade)

    • Added support for a multiviewer output (2+8 or 2+13), including tally indicators and input labels. Preview, PGM, and 8 (or 13) inputs are currently supported.

    • Added support for 4x overlays.

    • Added support for VU-meters (peak stereo level)

  • Added support for SRT:

    • SRT stands for Secure Reliable Transport. It's an open-source protocol designed for low-latency, secure, and reliable video and audio streaming over the internet. SRT aims to minimize latency and packet loss while providing robust security features.

    • New Input: SRT Input

  • RTMP Target:

    • Added support for sending time stamps in the outgoing data stream ("onFI").

    • Improved performance when using the Nvidia NVENC encoder. Reduced system load overall.

    • Support for scaling (1x, 0.66x, 0.5x, 0.33x) the output resolution.

  • RTMP Input: Added option to automatically connect to the RTMP server when the input is started. This option is enabled by default.

  • NDI Target & NDI Input: Added support for alpha channel (BGRA 4:4:4:4)

  • Video File Input:

    • Added support for playback of lossless HEVC (H.265) files, using YUV444P format.

    • Added support for .dv and .mxf file formats.

    • Added support for DNxHD video codec.

    • Added support for QuickTime Animation RGB 4:4:4 codec.

  • File Target:

    • Support for hardware-accelerated encoding using Nvidia NVENC. This reduces system load and improves encoding performance.

    • Support for HEVC (H. 265) encoding using Nvidia NVENC. H.265 is a modern video codec that provides better compression and quality than H.264.

    • Support for lossless HEVC encoding (4:4:4). Lossless HEVC generates very high-quality video files, but the file size is large.

    • Support for AV1 encoding using Nvidia NVENC. AV1 is a modern video codec that provides even better compression and quality than H.265.

  • HTTP API:

    • Added a new API for starting playback of media files from disk. The new API is /api/media/startmedia?input=MyMediaInput&sourceuri=MyFileName.

  • Alignment operator:

    • Added support for adjusting the stroke width of the alignment lines.

  • New operator: Audio Routing Operator. This operator can be used to modify audio layer content (8 channels).

  • Script Engine: Added new helper method IsLinux() to check whether Composer is running on Linux.

  • Text Input: Added support to display media file input properties: Time Code, Time Left, filename (short), and playback End Time.

  • All inputs: added two new advanced options:

    • Invisible - this option will remove the Input from all layers, making it invisible and silent.

    • Do not render - this option will disable the internal render method of the input. Useful when optimizing a high-load multicamera setup.

  • Added a new input for Batch Processing, making camera switching much easier when using Connectors.

  • Connectors: added a timer (ms) display and an auto-generated link to the HTTP API.

  • Project Tuning:

    • Added a new project setting for optimizing the internal renderer workflow. When this new setting is enabled, the internal renderer will only render inputs that are used, plus inputs that are configured to always render. This new option can significantly improve performance in multicamera+branding scenarios. To optimize your project for utilizing this new feature, please get in touch with RealSprint.

Desktop version:

  • New audio mixer:

    • Added support for toggling between the normal layout and the mixer layout. Shortcut: CTRL+M.

  • Backup files: Files are now stored in a subfolder having the same name as the project. This makes it easier to find backup files.

  • Added a setting for dimming/not dimming the scene preview window when the scene preview is disabled.

  • Added support for resizing the name column in the scene layer list.

  • Input list:

    • Added a miniature VU-meter for all inputs.

    • Added a tooltip for media assets, showing the file name and the file path.

  • Operators list: added a miniature VU-meter for all operators.

  • CAV-inputs: added commands for skipping/rewinding 10 frames. Keyboard shortcuts: SHIFT+ALT+RIGHT and SHIFT+ALT+LEFT. To use this feature, select the CAV input in the inputs list, and then use the keyboard shortcuts.

  • Scene layers:

    • Added support for renaming a layer to the file name used by the associated input ("short name"). Use the context menu to rename the layer.    

    • Added support for moving a layer up/down/top/bottom using the context menu.

    • Added an audio volume knob (stereo) for the selected scene layer.

    • Added an audio balance knob (stereo) for the selected scene layer.

    • The order of properties has been changed.

  • Scene audio:

    • Added support for setting the audio volume for the selected scene.

    • Moved the scene vu-meter to the center of the UI, below the scene preview area.

    • Added basic scene audio operators for lowcut, compressor, and limiter. For more advanced configurations, use Audio operators on layers or on a scene.

  • Added support for exporting Decklink Capture snapshots (images) to the Snapshots folder. The new feature is located in the File > Export sub-menu.

  • When exporting a project and media assets, any Decklink Capture snapshots will be included in the export.

  • Audio Preview:

    • Only audio from channels 0 and 1 (= left and right, stereo channel 1 ) will be played in the audio preview by default.

  • VU meters: added a configuration for scale (0,+3,+6,+9,+12,+15, and +18 dB).

  • Zoom in/out using the Scroll wheel: Zoom in/out has been improved, creating a smoother/linear experience.

 Bugfixes:

  • Backup media did not include all model files.

  • Backing up an unsaved project rendered an application exception.

  • BlackMagic Target V4 would crash when loading a project using a BlackMagic Target component using a DisplayMode not supported by the installed capture card.

  • Squashed two bugs: deleting the last layer within a scene could make the audio go into "repeat mode", and the scene content was not always cleared (the last rendered image was shown instead of an empty image).

  • NDI Target would crash when used in a scene with a portrait aspect ratio (height > width).

  • Preview Audio may sometimes become out of sync or produce unexpected audio.

  • Stereo VU meters did not always display correct values (visual scale error).

       

  Other changes/news: