Skip to content

BetterAudio 26.7-beta.4

Pre-release
Pre-release

Choose a tag to compare

View all tags
@rokartur rokartur released this 24 Oct 17:36
26.7-beta.4
1b7524b

Pre-release for testing. Report issues: https://github.com/rokartur/BetterAudio/issues
Get betas in-app: Settings → General → Include beta versions.

Since beta.3

Added

  • Switching output devices crossfades instead of cutting the sound. Settings → Output → Switching sets the length, 50 to 1000 ms (150 ms default). #157
  • Headphone correction per output device: pick an AutoEQ profile or import ParametricEQ.txt, and compare A/B from the menu panel. #155
  • Loudness compensation lifts bass and treble at low volume by ISO 226:2023 contours. #156
  • Dictation profiles: choose the voice model (Parakeet or WhisperKit), the language model, a translation target and the shortcuts per profile.
  • Per-app EQ is a curve with draggable points instead of sliders.
  • Multi-Output in the standard panel: "+" in the Output header adds a device, mirrored devices show a badge and a remove button.
  • Settings → Per-app Volume shows the driver status and version.

Changed

  • Settings are grouped into new tabs: Keyboard Shortcuts, HUDs, Menu Panel, Automation and Integrations.
  • The wide panel's Input, Output and Sound Effects rows use a native layout. The separate Output card is gone.
  • Quit is an icon beside Settings in the panel footer.
  • Now Playing controls are redesigned.
  • The ChatGPT subscription sign-in for dictation is named Codex CLI.
  • Dictation sends an API key only to https servers or to localhost.
  • Stop-to-text with an unloaded voice model: 3234 ms to 2168 ms median. The language model step gives up after 10 s and pastes the raw transcript.
  • Launch is 414 to 521 ms faster: iCloud settings are read in one call.

Fixed

  • The license and trial survive the update from 26.6.2.
  • Per-app audio keeps playing through bad samples, mono outputs and unusual sample rates.
  • Video played through BetterAudio no longer drifts out of sync: drift over 110 s went from 529 ms to under 1 ms.
  • Loudness compensation engages; the device volume curve was read as 0 dB.
  • Apps with several audio streams get their EQ and meter on every stream.
  • A new EQ curve no longer clicks or carries old filter state.
  • iCloud Sync keeps edits, boosts and ignored apps from Macs on older versions.
  • Dictation never loses a recording or a transcript to the clipboard. Expired recordings that cannot be moved to the Trash stay listed.
  • Two dictations no longer sign you out of Codex CLI by refreshing the token at once.
  • A dictation profile keeps a synced voice model and language this Mac cannot select.
  • Old whisper.cpp models are deleted from disk.
  • Now Playing shows the new track on track change and clears it when the player quits.
  • The menu panel stays open on clicks inside its popovers. EQ and Multi-Output switches animate.
  • Remove… in Settings asks only to remove the driver. #198
  • Hearing Health history and the input mute state survive a quit.
  • The CLI reports a blocked profile and a timed-out command instead of success.
  • Dictation, Shortcuts actions and Siri phrases are translated in all 9 languages.

Removed

  • Show Feature Tour in Settings → About.
  • The compact Now Playing layout.

Install

  • The audio driver updates to 26.0: approve one driver update after installing.
Assets 6
Loading