Resona v1.4.1
·
82 commits
to main
since this release
Immutable
release. Only release title and notes can be modified.
v1.4.1 : 2026-08-22
A major visual overhaul since v1.3.1: a complete UI redesign around a dynamic accent pulled from album art, a new Settings screen, reliable voice search, and the ability to download APKs directly from the update banner.
Highlights
- Dynamic accent from album art: Home, Library, Search, and Now Playing screens now share a floating pill bottom nav whose active tab and accent colors are derived from the currently playing track's album art, falling back to black/white when nothing is playing.
- Shared top bar: Home, Library, and Search now share a unified top bar with logo, downloads, settings, and voice search access.
- New Settings screen: a full About section and a Developers section that pulls real contributor avatars from the GitHub API, with bots and AI accounts filtered out.
- Voice search rebuilt: voice search now uses SpeechRecognizer directly instead of handing off to the system UI, with Resona's own listening overlay featuring pulsing rings in the current accent color, a live transcript, and immediate feedback on tap including failure states.
- Improved search: Search now shows real Trending/Suggested shelves sourced from the same feed Home uses, replacing static mock data.
- Larger artist pages: Artist detail screens now show up to 20 songs instead of 5.
- Direct APK download: the "new version available" banner now has a Download button that fetches the APK straight from the GitHub release via the system download manager, no browser trip required.
- Dark-only: the light theme has been removed; the app is now dark-only with no in-app theme toggle.
- Typography rebuilt: Montserrat (stretched display voice) and Roboto Flex (everything else), both bundled locally. Fixed a pre-existing bug where Montserrat's weight axis silently rendered every requested weight as Thin.
- Commits are now SSH-signed and show GitHub's Verified badge.
Known Issues
- Voice search is speech-to-text (say a song or artist name), not melody/humming recognition. That requires a paid third-party audio-fingerprint service not included in this app.
Verification
SHA-256: 5745dc59706a7330b0e967eced3194810371fee2629464f059ca386ce754d8bd