Releases
v1.13.0
Compare
Sorry, something went wrong.
No results found
1.13.0 (2026-09-06)
Features
analyzer: ANALYZER_REPLICAS=0 turns the local analyzer container off (#1586 ) (97f72b4 )
api: expose CLAP sounds-like search over HTTP and MCP (#1575 ) (#1578 ) (048cdb0 )
backup: scheduled, rotating backups (#1570 ) (#1585 ) (9d56bec )
broadcast: "talk only between tracks" scheduling constraint (#1485 FR 5b) (#1562 ) (510ab49 )
broadcast: configurable show handover timing and ordering (#1576 ) (#1581 ) (a859e2d )
broadcast: expose the two duck depths as settings (#1485 FR 6) (#1560 ) (971b997 )
broadcast: fade a long track out at a show boundary (#1574 ) (#1583 ) (2a31319 )
exportable LLM call log + GeoIP fallback for listener country (#1485 FR 15) (#1561 ) (12a9f2a )
library: consolidate the scene vocabulary after a tagging pass (#1580 ) (1489a83 )
one-field DJ Brain setup, wizard preset, and native cloud speed (#1564 ) (f120da1 )
picker: album-level cooldown at the point of choice (#1485 FR 3) (#1563 ) (f5d92e1 )
picker: minimum track length filter, per show and station-wide (#1573 ) (#1582 ) (daf1256 )
scrobble: report plays back to Navidrome so lastPlayed drives rotation (#1559 ) (b4a6910 ), closes #1298
search: allow pinning a SearXNG engines list (#1557 ) (b030345 ), closes #1353
tts-heavy: release idle TTS models instead of pinning ~4GB forever (#1579 ) (#1584 ) (07c1ea4 )
tts: persona engine can inherit the station default (#1566 ) (f141ae1 )
Bug Fixes
cache a missing analyzer backend, forward CHATTERBOX_REFERENCE_WAV (#1591 ) (#1592 ) (6c39c63 )
library: count the library on request, not on every page load (#1570 ) (#1587 ) (e0d7baf )
scheduler: rebuild auto.m3u whenever the active show changes (#1111 ) (#1558 ) (6e4e7c7 )
tts: add a Slovak row to the voice preview table (#1108 ) (#1556 ) (05ff108 )
Documentation
cross-reference open Discord suggestion threads against GitHub (#1572 ) (8b735e9 )
remove stale planning and draft docs (#1588 ) (49de92b )
You can’t perform that action at this time.