Singing Stream Savior · v2.1.7.0

RELEASE HISTORY · FROM NEWEST

Release history

Review every Singing Stream Savior release from newest to oldest, including features, usability improvements, stability fixes, and compatibility changes.

40 public releases 2.0.0.0 → 2.1.7.0
Feature update
Current release

2.1.7.0

12 changes

Profile sharing, application backup and restore, and improvements to YouTube audio and playback.

  • Import and export vocal Profiles to share or move your effects settings.
  • Import OBS audio effects, with support for Gain, Noise Gate, Compressor, and Limiter.
  • Back up and restore application settings, Profiles, the song library, selected projects, and related audio, lyrics, and artwork. Identical audio files are stored only once.
  • Improve the responsiveness of Profile loading and saving.
  • Improve YouTube batch downloads: a failed song no longer affects the others. Successful songs link to local audio files, and failed songs are listed when the batch finishes.
  • Improve MP3/WAV conversion and playback afterward. The progress window now shows which song is being processed.
  • Improve YouTube playback-range editing for some videos that can play but cannot be downloaded. Fix cases where playback had to start before a range could be adjusted.
  • Add a missing-audio indicator to the song list. Click it or use the right-click menu to locate the file while keeping the song's settings.
  • Improve playback stability when switching songs, pausing, and stopping.
  • Long song titles now scroll in one direction and restart from the beginning. Refine several interface details.
  • Stop automatically installing or repairing the OBS connector at startup, reducing unnecessary permission prompts.
  • Happy Mid-Autumn Festival, everyone! ٩(˃̶͈̀௰˂̶͈́)و
View GitHub Release
Hotfix

2.1.6.1

2 changes

2.1.6.1 hotfix

  • Fixes an issue in advanced streaming mode where the Microphone Block's level meter on the Audio Routing page could remain still even though the microphone signal was working normally.
  • Fixes issues that prevented downloading audio from YouTube links and converting downloaded files.
View GitHub Release
Launcher hotfix

1.2.0.8

5 changes

Launcher-only update. The application remains at 2.1.6.0; the existing complete-install ZIP and audio features are unchanged.

  • Update failures now use an inline dark status panel instead of a separate system dialog. A verified safe installation starts automatically; an unsafe installation remains blocked with an explanation and a Close button.
  • Show download speed in KB/s or MB/s, with distinct waiting, retry, verification and installation states.
  • Add bounded retries, reconnect and validated resumable payload downloads. Unsupported resume or a changed asset restarts the download. Complete size, SHA-256 and installation checks remain mandatory.
  • Add button hover feedback while preserving disabled, pressed and keyboard-focus states.
  • Keep bounded download diagnostics locally without automatic uploads. No download-speed improvement is guaranteed for every network route.
View GitHub Release
Feature update

2.1.6.0

8 changes

Local diagnostic export and theme customization

  • Adds Export diagnostic data for reporting software problems. Records stay on your computer and are manually exported as a ZIP, with no automatic upload or email.
  • Adds theme color and background opacity options to Transparent Black and Transparent White.
  • Fixes an issue where some custom font names could not be applied correctly to lyrics, improving handling of parentheses and other special characters.
  • Improves VST3 plug-in compatibility, audio configuration checks and audio bus handling.
  • Fixes hand cursors incorrectly appearing over some non-clickable areas.
  • Improves the display priority of status messages.
  • Improves lyric display in Classic Karaoke.
  • I didn't manage to get any Hans Zimmer merch 🥲
View GitHub Release
Stability update

2.1.5.4

12 changes

Stability update

  • Improves brief BGM dropouts.
  • Improves occasional delayed response when stopping accompaniment.
  • Fixes how Vocal Profiles are applied in player modes.
  • Improves VST3 stability.
  • Fixes an issue where preview could stop unexpectedly or Profile switching could fail after latency adjustments to some VST3 plug-ins.
  • Improves effect preloading performance and fixes an issue where a Profile containing disabled effects could open the wrong effect editor.
  • Fixes an issue where a failed VST3 replacement, removal, or enable operation could still save the wrong selection or checkbox state.
  • Improves project media collection.
  • Improves UVR vocal separation.
  • Improves recording save and finalization.
  • Improves YouTube feature stability.
  • I'm going to see Hans Zimmer in concert today! Can't wait to soak up Guthrie Govan's solos!
View GitHub Release
Stability update

2.1.5.3

8 changes

Several stability improvements across the app.

  • Improves data retention when linked audio or attachment files are temporarily unavailable.
  • Improves recording stability when recordings finish, fail, or are interrupted.
  • Improves stability for YouTube playback resolving, background downloads, and fallback playback.
  • Improves Collect Project Media so custom cover art is collected as well.
  • Improves OBS Timeline timestamp accuracy when OBS status queries time out.
  • Improves UVR temporary-file cleanup during batch processing to reduce disk usage on large batches.
  • Fixes several other small issues and improves overall stability.
  • I just finished reading Kaiju No. 8, and it was honestly way too good. Now that I’ve finished it, it feels like I’ve lost some purpose in life. I hope I can find another great manga soon…
View GitHub Release
Stability update

2.1.5.2

6 changes

Several stability improvements across the app.

  • Improves recovery when audio devices disconnect and reconnect.
  • Improves VST3 editor stability.
  • Standardizes the feature name as Lyrics Editor.
  • Improves editing and saving of word-synced lyrics.
  • Improves player behavior.
  • Scrolled through lots of cat videos.
View GitHub Release
Lyrics timing hotfix

2.1.5.1

1 changes

Fixes word timing getting stuck when a word passes the next original plain-LRC line timestamp.

  • Fixes word timing getting stuck when a word passes the next original LRC line timestamp; the next line starts at its first actual word mark.
View GitHub Release
Lyrics, guide vocal, and workspace improvements

2.1.5.0

8 changes

Adds a built-in lyrics editor, guide vocal monitoring, dockable Meter controls, and lyrics improvements.

  • I went out in the rain to buy a raincoat today, and the rain stopped as soon as I put it on...
  • Adds a lyrics editor for creating LRC files with detailed word-by-word timing.
  • Adds guide vocal monitoring from UVR-separated vocals. Guide vocals are heard only in monitoring, never in the OBS mix; their level can be adjusted in Meter and they follow key and speed changes.
  • Lyrics Window and YouTube Video buttons now stay in the accompaniment player, and the duplicate buttons on the Lyrics page have been removed. The YouTube window also offers caption selection when playing a YouTube link.
  • Lyrics timing offset can now be adjusted by up to ±20 seconds.
  • Makes small fixes to some animated lyrics.
  • The Meter window can now dock on the right of the main window or float again. Track order was also adjusted, with Master last.
  • UVR vocal separation adds a Keep Vocal track option. Kept accompaniment and vocal tracks can be linked together for guide vocal use.
View GitHub Release
Audio, video, lyrics, and performance improvements

2.1.4.3

8 changes

Fixes brief audio artifacts when switching effects, adds synchronized YouTube video, and improves lyrics and song library performance.

  • I found a few small issues while testing my own singing livestream, and they are now fixed. No need to worry—these issues had little impact on livestreams.
  • Fixes brief bursts of earlier vocals or reverb that could occasionally occur when switching between chatting and singing effects.
  • Adds a “YouTube Video” window that follows accompaniment playback, pause, seeking, and speed changes.
  • Uses monochrome icon buttons for “Lyrics Window” and “YouTube Video” in Compact mode to save player space.
  • Improves Letter Spread transitions for smoother lyric changes and starts the letter spreading effect earlier.
  • Fixes a brief set list shake when songs are added to Set List in Oblique Stream.
  • Makes dragging lyrics smoother while reducing unnecessary processing.
  • Improves loading, searching, and scrolling performance for large song libraries to reduce stuttering.
View GitHub Release
Theme, lyrics, and display improvements

2.1.4.2

7 changes

Fixes a persistent audio timing warning and adds a new theme, two lyric styles, and clearer one-way text motion.

  • Fixes an issue where audio routing could keep showing “Check audio timing”.
  • Adds the new “Oblique Stream” theme with a minimal transparent design. Place it on the left or right, show Now Playing above or below, and enable timestamps and curved scrolling.
  • Adds two animated lyrics styles: “Letter Spread” and “Stagger Signal”.
  • Long song titles now scroll to the end, pause briefly, then restart from the beginning. Set list scrolling follows the same one-way reading flow.
  • The former numbering button now cycles through “Off, 1., 01, .” in themes that support numbering.
  • Fixes update progress jumping when the launcher switches between downloading, verification, and installation.
  • Had drinks with friends today! Happy!
View GitHub Release
Hotfix update

2.1.4.1

10 changes

Fixes missing new themes after updating from 2.1.3.2 to 2.1.4.0. Fixes enlarged playlist blocks getting stuck when dragged.

  • The spicy Weili fried sauce noodles are seriously delicious—the night owls working late are in for a treat!
  • Fixes missing new themes after updating from 2.1.3.2 to 2.1.4.0.
  • Fixes enlarged playlist blocks getting stuck when dragged.
  • Defaults YouTube downloads to MP3 320 kbps and keeps only MP3 and WAV. Downloaded audio can be converted locally while preserving originals and avoiding same-name overwrites.
  • Adds a Download all YouTube files entry beside song statistics; right-click to choose format, open download settings, or open the download folder. Library update messages now appear in the main status bar.
  • When returning to BGM or chat, Now Singing / Now Playing becomes Now. Chat themes are enabled by default while existing disabled preferences are preserved.
  • Clicking elsewhere clears the Meter knob highlight and commits and ends percentage editing.
  • Waits for audio output readiness before playing local BGM to avoid occasional autoplay failures.
  • Lyrics style cards now consistently show a clear style description below the name.
  • The new launcher supports normal, recommended, and emergency hotfix updates; recommended and emergency updates can prompt at next launch.
View GitHub Release
Animated lyrics and playlist themes

2.1.4.0

6 changes

Adds seven animated lyrics styles: Kinetic Type, Prism Cut, Lumen Drift, Ink Cascade, Silk Script, Verse Stack, and Glyph Motion, with different text compositions, entrances, and transitions.

  • Improves lyrics synchronization and multilingual display: fixes occasional horizontal lyrics appearing in Vertical Verse's single-area layout, and limits romanization to Basic Lyrics.
  • Adds animation previews directly inside playlist theme and lyrics cards, and improves previews of Basic Lyrics scrolling and Classic Karaoke color fill.
  • Improves layout editing: lyrics can be moved and resized beyond canvas edges, resized from all four corners, and identified more clearly with playlist and animated lyrics area labels.
  • Adds five playlist themes: Signal Veil, Signal Tide, Cue Line, Margin Set, and Oblique Plate. Both Signal themes use actual playback audio to drive waveforms and light points. Removes some older themes, retaining more versatile ones. Quiet Ledger now combines left and right layouts in one theme.
  • Refines existing themes: removes the selected extra themes and fixes unnecessary scrolling of short titles in Transparent v2, backgrounds that fail to extend with custom layouts in Transparent, and default placement and dragging issues in some themes.
  • Other display fixes and asset improvements: fixes the blurry and partially clipped accompaniment playback button.
View GitHub Release
Karaoke, preview, and playback improvements

2.1.3.2

6 changes

Fixes the issue where BGM sometimes fails to resume automatically after accompaniment ends.

  • Adds Classic Karaoke with per-character color fill and Vertical Verse with vertical animated lyrics, supporting font and text color adjustments.
  • Lyrics and playlist themes can be dragged and proportionally resized in the preview, with live OBS synchronization.
  • Improves editing bounds, fixes preview canvases at 16:9, and reduces unnecessary layout calculations.
  • Updates the launcher to 1.2.0.5, reuses unchanged files, and reduces downloads when differential packages are available.
  • Improves YouTube fallback playback and video range previews.
  • I ate delicious mala chicken rice today, and I am very happy!
View GitHub Release
YouTube playback hotfix

2.1.3.1

3 changes

Improves YouTube accompaniment reliability during output switching and keeps the fallback display and play history accurate.

  • Prevents valid YouTube sources from being sent to fallback during output switching.
  • The fallback video panel no longer appears unnecessarily or remains after playback stops.
  • Accompaniment that did not start successfully is no longer incorrectly recorded as played.
View GitHub Release
Monitoring, lyrics, and launcher improvements

2.1.3.0

5 changes

Greatly reduces monitored vocal latency, improves lyric updates, and speeds up launcher checks while adding chat topics to the BGM area.

  • Greatly reduces monitored vocal latency (ASIO is recommended; its latency is lower than WASAPI)!
  • You can now set a chat topic in the BGM area and display it while BGM is playing.
  • Fixes lyric fill progress and some subtitle timing displays, and improves lyric updates when switching songs or changing Traditional/Simplified Chinese settings.
  • The launcher checks files faster and continuously shows verification progress, reducing the chance that its window appears stuck.
  • I went to the gym today, and I think I'm great!
View GitHub Release
Launcher update

1.2.0.4

4 changes

This update is for the launcher only. The application remains at version 2.1.2.1.

  • Reduce waiting during repeated update checks.
  • Keep the current item and progress visible while checking files, so you can see that the launcher is still working.
  • Improve responsiveness during update checks.
  • Mucheng is sleeping, so this launcher update was published by AI Mucheng.
View GitHub Release
Launcher repair hotfix

2.1.2.1

3 changes

Fixes unnecessary resource downloads when launching a fresh full installation of the current version.

  • A complete, unchanged full ZIP now passes the launcher's file checks and starts directly, without downloading about 525 MB of resources again.
  • Missing or damaged files still trigger repair. The launcher shows a compact repair status instead of announcing a new version with empty update notes.
  • Mucheng is sleeping, so this version is being published by AI Mucheng.
View GitHub Release
HP models, MP3 export, and reliability

2.1.2.0

12 changes

Adds HP vocal separation and MP3 export, improves UVR controls, and fixes audio transitions and update reliability.

  • Fixes stretched or repeated sound when switching to a Profile or VST3 effect with higher latency. Backing tracks and vocals now share the same transition; adding latency uses a brief fade-out, silent interval, and fade-in.
  • Prepares the effects signal before switching to Effects Only monitoring, reducing incomplete sound at the moment of switching.
  • When the monitor audio route is interrupted, monitor recording stops with a clear status message and retains the recorded content instead of appearing to keep recording.
  • Fixes stale connection state after OBS audio output fails to start.
  • Keeps stream and monitor volume settings consistent when several gains are applied together.
  • Strengthens update verification so the application remains complete and consistent after an upgrade.
  • UVR now remembers the default vocal-handling mode and output format correctly. Click the pencil to edit a song title; processing is marked only by the left accent strip.
  • Adds HP Preserve Harmony and HP Remove All Vocals alongside MDX. Each queued song can use a different model; new installations default to HP Preserve Harmony.
  • Adds MP3 export alongside WAV and FLAC, with a choice of 48 kHz or 44.1 kHz. The default remains 48 kHz.
  • Fixes processing issues for MDX Preserve Harmony.
  • Polishes interface details by softening the divider before CPU/RAM usage and removing the dotted outline when clicking launcher update notes; mouse-wheel and keyboard scrolling remain available.
  • I really want sushi today...
View GitHub Release
Audio switching and monitor reliability

2.1.1.2

7 changes

Improves the reliability of Profile and VST3 effect changes, monitor-device switching, and audio startup.

  • Profile and VST3 effect changes are prepared before the audible transition, reducing interruptions, clicks, and timing shifts when switching between effects with different latency.
  • Changing the monitor output no longer restarts the rest of the audio path. Streaming, OBS output, and recording can continue, and a failed device change safely restores the previous monitor output.
  • Monitor On or Off now remains consistent after temporary Profile or Block monitoring ends instead of unexpectedly returning to an older setting.
  • When a saved audio device is unavailable, the application falls back safely and shows a clear status message instead of leaving the controls and actual routing out of sync.
  • Audio startup and monitor activation are better protected against brief gaps or unstable sound while the buffers become ready.
  • VST3 scanning and effect windows are more resilient. A plug-in that crashes or stops responding during scanning or editing is less likely to affect the rest of the application; effect windows also close more reliably across display scaling and interface languages.
  • Hungry in the middle of the night and not sure what to eat—it felt like McDonald’s was the only option… Wait! I’ll cook some noodles and have them with scallop sauce. I love scallop sauce!
View GitHub Release
UVR GPU acceleration and playback fixes

2.1.1.1

8 changes

Adds MDX GPU acceleration and fixes UVR output, long-YouTube preview playback, and volume adjustment.

  • The existing Preserve backing vocals and Remove all vocals models can now use GPU acceleration on supported Windows DirectX 12 hardware.
  • UVR output can now be selected as 48 kHz (default) or 44.1 kHz, and the selected rate is saved for future batches.
  • If GPU acceleration is unavailable, the current batch automatically retries with CPU processing instead of losing the job.
  • Corrects reversed (Instrumental) and (Vocal) outputs so importing a completed UVR result adds the actual instrumental track to the song list.
  • The long-YouTube range editor now keeps video playback, the play/pause button, and the timeline synchronized, including seeking. Ads may appear when the fallback player is used.
  • YouTube selection now prefers the video's original audio language instead of an automatically dubbed track.
  • UVR processing load is included in the application's CPU reading, while fallback and completed-batch notices appear in the main window's lower-left status area.
  • The mouse wheel now adjusts both BGM and accompaniment volume by 2 percentage points per notch for finer control.
View GitHub Release
Vocal removal and playback workflow update

2.1.1.0

8 changes

Adds integrated MDX vocal removal, visual range selection for long YouTube videos, new Reserve playback modes, and safer Output Bus updates.

  • Adds a dedicated UVR Vocal Removal workspace between Playlist Appearance and Profile Editor. Drop audio files or YouTube links into its queue, edit titles, monitor each item and the whole batch, stop processing, and import completed instrumentals directly to the song list.
  • Includes two MDX choices with Preserve backing vocals as the default. Each queued item can instead remove all vocals, and the chosen setting is snapshotted when a batch starts so later edits cannot change work already queued in that batch.
  • Writes paired (Instrumental) and (Vocal) files as 48 kHz FLAC or WAV. The UVR output folder is managed centrally in Settings > Files & Project, while the UVR page links directly to that setting.
  • YouTube videos longer than 15 minutes can open a video-preview range selector with a timeline and Capture start/end controls. The selected segment is also available when importing accompaniment, persists in the project, and plays from 00:00 for the selected duration after reopening.
  • Opening the YouTube range selector pauses accompaniment and BGM. If BGM was playing beforehand, both Apply and Cancel restore that original BGM state.
  • Adds Play once, Repeat current, Play Reserve in order, and Shuffle Reserve modes. The same synchronized choices are available from the Windows notification-area menu, while direct song-list playback safely returns to Play once.
  • Improves the stability of Output Bus setting changes. When Mix, Compressor, or Limiter controls are adjusted together, playback applies one complete set of values instead of briefly mixing old and new settings.
  • Because finding an audio file, removing its vocals, and then dragging it into the song list every time felt too cumbersome, I stayed up late and built vocal removal with YouTube support too.
View GitHub Release
Playlist, status, and launcher hotfix

2.1.0.3

4 changes

Hotfixes playlist reordering and status visibility, and includes Launcher 1.2.0.1 with steadier update rendering.

  • Fixes the playlist drag indicator and the actual dropped position disagreeing when a song is moved between other songs.
  • Moves YouTube loading and playback messages to the main window's lower-left status area instead of showing a loading label below the player, and restores other playback status messages there.
  • Adds a one-way, start-to-end marquee for long status messages in Mini mode. The Monitor source selector now uses the same reset-to-start behavior instead of bouncing backward.
  • Launcher 1.2.0.1 reduces icon flicker while update data is processed and card-area flicker while release notes are scrolled.
View GitHub Release
Standard Player audio-routing improvements

2.1.0.2

3 changes

Makes Standard Player Mode safer and simpler, and preserves Advanced Streaming routing choices when switching modes.

  • Improves Audio Routing in Standard Player Mode. It now always uses Windows Audio and the system default output, while hiding driver and output-device choices that do not apply to this mode.
  • Fixes Advanced-only Buffer, latency, and stability information remaining visible in Standard Player Mode.
  • Improves the first Advanced Streaming Mode setup by preferring an installed OBS plugin or detected virtual output, and preserves existing OBS plugin, Monitor device, and ASIO settings across repeated mode switches.
View GitHub Release
Audio device, channel, and device-list hotfix

2.1.0.1

4 changes

Fixes audio-device startup and Windows Audio microphone routing after the 2.1 update, and improves long device-name display.

  • Fixes ASIO and related audio features sometimes appearing unavailable on the first launch after upgrading from 2.0.5.4. They are available from the first launch after updating, without a second launch for repair.
  • Fixes a mono microphone connected to Input 1 being recorded mainly in the left channel when using Windows Audio. Input 1 is now duplicated equally to left and right for centered Dry, Wet, Final Mix, OBS, and recording output.
  • Improves device selection on the Audio Routing page. When Stream Output or Monitor Device is opened, its list expands to fit the longest device name while remaining within the available screen area; the closed field and card layout keep their existing width.
  • Yesterday I got a spoiled clam while eating hot pot, so I probably will not dare eat hot pot again for about a month.
View GitHub Release
Advanced streaming audio and vocal processing update

2.1.0.0

16 changes

Adds Advanced Streaming audio routing, editable vocal Profiles, direct OBS output, a five-track Meter, and complete shortcut and tray controls.

  • In theory this update should not cause problems, but the author is still acting as the first test subject for the new features. If you are concerned, you can wait before updating.
  • Starting an accompaniment automatically applies the microphone effects assigned to that song. Every song can use a different chain, and the app returns to the chat setup when accompaniment ends.
  • Added Advanced Streaming Mode to mix BGM/accompaniment with the Profile-processed microphone and send the complete stream through a virtual audio device or the Singing Stream Savior OBS audio plug-in.
  • Added monitoring and recording for BGM/accompaniment, dry vocal, and processed vocal. Final Mix, OBS, and Full Output recording stay synchronized. The audio interface's Direct Monitor remains the recommended main singing monitor.
  • Changed BGM and accompaniment volume to a perceptual loudness curve, with the same mapping for local audio files and YouTube sources. Existing slider numbers are preserved, but the same value can sound different from an earlier version; recheck accompaniment-to-vocal balance before the first live stream.
  • Added editable vocal Profiles, 15 consistently designed built-in effects, draggable Block ordering, and up to eight VST3 slots; parameters, order, and VST3 state are stored with the Profile.
  • Redesigned Factory Profiles for chat, general singing, lower/higher ranges, airy traditional styles, and other starting points with practical Gate, Air, space, and dynamics processing.
  • Program and vocal monitoring can be balanced separately; Stream Output adds a Mix Bus Compressor, Final Limiter, Master, and live latency/device-health information. Final Limiter defaults off in Normal Playback and on the first time Advanced Streaming Mode is used; later manual states are stored separately for the two modes.
  • Added a five-track Meter that can dock to the main window or float independently, with horizontal/vertical layouts, 0–200% knobs, PEAK, and three-second LUFS-S for Stream Output. Its color and switch control match the main UI, while the floating window uses standard Windows minimize and close controls.
  • Added a borderless CPU/memory status in the lower-right corner. Its tooltip separates system and app use; Advanced Streaming Mode also shows buffer, callback, estimated latency, and underrun/overrun details.
  • Added global shortcuts for Reserve play/pause, stop, restart, key and speed changes/resets, and the lyrics window, grouped by function and hidden when the current audio mode does not support them.
  • Expanded the system-tray menu with playback, restart, key, speed, Profile, microphone, lyrics-window, and Meter controls, plus a setting that chooses minimize-to-tray or full exit for the window close button.
  • Improved live and persisted lyrics Offset behavior, lyrics-window availability, source-row alignment, and the borders, spacing, and capability-dependent controls across Lyrics, Profiles, Library, and Appearance pages.
  • Effect windows stay editable without blocking the main app. They support One Knob adjustments, full advanced parameters, live signal graphs, help, and tooltips, with fixes for Saturation naming and Compressor or other graph redraws.
  • Silenced application notifications, ended Profile audition when minimizing to tray, improved complete shutdown, and hardened VST3 editing, state saving, closing, and reordering.
  • Carried forward staged YouTube fallback and lyrics-search reliability fixes, while strengthening OBS reconnection, OBS 29–32 compatibility, safe install/repair/removal, and virtual-output diagnostics. A successful install automatically selects and applies the “Singing Stream Savior Audio (OBS Plugin)” output.
View GitHub Release
YouTube playback compatibility hotfix

2.0.5.4

3 changes

Urgently fixes recent YouTube playback issues with staged retries and fallback playback.

  • Urgently fixes recent YouTube playback compatibility issues and adds staged retries with fallback playback.
  • Fallback playback still supports tempo, key changes, seek, loop, and automatic BGM switching during karaoke playback.
  • Several major features are being prepared for the next major release. The scope is broad, so completion will take more time; thank you for your patience.
View GitHub Release
Playback and Overlay hotfix

2.0.5.3

6 changes

Improves playback startup and large-list operations, with fixed-speed loop scrolling for Card, Default, and Transparent lists.

  • Card, Default, and Transparent Set List/Reserve panels now scroll at a constant speed, flash at the end, and restart from the top.
  • List speed is adjustable from 8–60 px/s with a 17 px/s default, and themes can declare shared scrolling or paging rules in their manifest.
  • Local and remote sources are prebuffered by 100/200 ms in the background to reduce brief startup dropouts.
  • Playback markers update only affected rows, and stale markers or lyrics after natural completion are cleared.
  • Added multi-row Reserve removal and Clear Sung History, and simplified the automation status to a lamp and text.
  • After handing the work to Codex, I ate a plate of stinky tofu. It was delicious OuOb
View GitHub Release
Stability hotfix

2.0.5.2

4 changes

Fixes delayed OBS timestamps and improves interface responsiveness around play, stop, and rapid song switching.

  • Fixed automatic timestamps being recorded several seconds late when the OBS WebSocket response was delayed.
  • Moved lyric text conversion to the background to reduce brief stalls during play, stop, and rapid song switching.
  • Launcher 1.0.1.3 follows the Windows display language and retries cleanup of stale update backups and temporary files.
  • I stayed up too late fixing a few things. Hopefully I will not oversleep tomorrow.
View GitHub Release
User experience hotfix

2.0.5.1

2 changes

Improves play, stop, and rapid song switching responsiveness, and silences dialogs during streams.

  • Fixed stale playback, lyric, time, and highlight state after rapid song switches.
  • Unified stop cleanup and silenced application and launcher dialogs.
View GitHub Release
Feature and playlist update

2.0.5.0

7 changes

Adds BGM playlists and a card-style song view, with clearer playback state, YouTube sources, and more reliable cross-version updates.

  • Added a collapsible BGM playlist with multiple local or YouTube sources, notes, drag reordering, Repeat One, Repeat All, and Shuffle All.
  • Added a card-style song list that emphasizes the audience-facing title while retaining source, cover, and lyrics status. It is optional; the traditional table remains the default.
  • The playing BGM or song is now clearly highlighted across the library, queue, and History, making rapid switches easier to follow.
  • YouTube sources now show their resolved video titles. When adding a YouTube playlist to BGM, choose the current video or the entire playlist.
  • Improved drag-and-drop feedback across the song library, playlists, and BGM list, plus consistency between player buttons and repeat icons.
  • Launcher 1.0.1.0 now completes app and required-resource updates in one place, including when upgrading directly from an older version.
  • While Codex was working, I trained at the gym for an hour. I am proud of myself today!
View GitHub Release
Resource update

2.0.4.0

5 changes

Adds a more complete resource update system that can restore every required resource even across skipped versions.

  • Simplified/Traditional Chinese dictionaries, Japanese and Korean readings, shared theme resources, and YouTube downloading support are now managed in one independently versioned resource pack.
  • Startup checks the resource version and integrity, then downloads or restores missing, damaged, or outdated files from the local cache.
  • Updating directly from an older version now restores everything required by the current release, without installing each intermediate version.
  • File lists, sizes, and hashes are verified before resources are applied, so incomplete content is never treated as a successful update.
  • While Codex was working, I did the laundry and cooked dinner. It was a very productive day.
View GitHub Release
Hotfix

2.0.3.1

3 changes

Fixes missing Korean romanization in the main preview and OBS, and improves launcher update checks.

  • Fixed Korean romanization appearing in the separate Lyrics window but not in the main preview or OBS after changing Reading Display (Preview + OBS).
  • The launcher now checks again every 15 minutes, retries after a failed check, and automatically restores Korean romanization files omitted when upgrading from 2.0.2.0.
  • Thought about life for a moment while Codex was working. I really want to get rich.
View GitHub Release
Feature and lyrics update

2.0.3.0

8 changes

Adds unaccompanied performances and improves lyrics discovery, Japanese and Korean reading aids, and OBS setup.

  • Added Unaccompanied Performance items for a cappella, self-accompanied, and other performances without a media file. They can end manually or after an optional planned duration and work with the queue, Now Singing, History, and project saving.
  • Automatic lyrics lookup now shows LRCLIB results and YouTube captions together, so synchronized lyrics remain available even when captions are found first.
  • Improved Korean caption identification and search ordering, and added Korean romanization with readable word spacing.
  • Improved context-sensitive Japanese readings and the consistency and loading reliability of Japanese and Korean readings across the main view, Lyrics window, and OBS.
  • Drag to OBS still supports direct dragging; clicking it now copies the Browser Source path for pasting into the OBS URL field.
  • Hovering a truncated playlist name in the sidebar now shows its full name.
  • Improved lyric display stability at song starts, instrumental gaps, and song changes.
  • Played some League of Legends ARAM while Codex was working and lost terribly.
View GitHub Release
Major update

2.0.2.0

7 changes

Adds an eight-step first-use guide and improves drag-and-drop, lyrics during instrumental gaps, and compatibility with older OBS versions.

  • Added an eight-step first-use guide covering BGM setup, karaoke tracks, display names, the queue, overlay themes, starting a performance, synchronized lyrics, and sending lyrics to OBS.
  • The guide can be reopened at any time from Help → First-use Guide.
  • Local audio files and YouTube links can be dropped across the full BGM or karaoke target area instead of requiring an exact drop on the button.
  • Improved guide highlights so every step points to the correct area of the interface.
  • Lyrics no longer jump to the first metadata line during instrumental gaps. When another lyric follows, the display anchors to it without falsely marking it active.
  • Improved theme compatibility with older OBS versions so transparent backgrounds and theme borders display correctly.
  • Watched Netflix's “Bloody Smart” while Codex was working. It was kind of bad.
View GitHub Release
Stability update

2.0.1.3

9 changes

Improves rapid YouTube switching, playlist imports, and launcher recovery.

  • Improved playback stability when switching rapidly between YouTube songs, preventing the wrong song from taking over.
  • YouTube playback now retries automatically after a failure, reducing cases that require restarting the application.
  • Playback state, the overlay's current song, and played-song history stay synchronized during rapid changes.
  • Dropping a YouTube video while viewing a custom playlist now adds it to that playlist.
  • {"Dropping a YouTube playlist offers four choices"=>"add all videos to the current playlist, create a new playlist, add only to All Songs, or cancel."}
  • Improved connection stability for online lyrics searches.
  • Improved automatic recovery when a launcher update fails or downloads incompletely.
  • Existing users receive the launcher update automatically through the updated main application.
  • Scrolled through a lot of Chiikawa while Codex was working.
View GitHub Release
Stability update

2.0.1.2

7 changes

Improves playback reliability, lyrics usability, and visual consistency.

  • Improved an issue where some YouTube links could stop playing until the application was restarted.
  • Improved song-title detection for some YouTube links.
  • Improved an issue where song-list overlays could show a black background in some OBS environments, preserving transparent themes.
  • Added Simplified and Traditional Chinese lyrics conversion with language-aware defaults; Japanese kana remain unchanged.
  • Lyrics timing controls, visible-line controls, and the detached lyrics window are clearer and more stable.
  • Dialogs and secondary windows consistently use the application's dark appearance.
  • Practiced guitar for a while while Codex was working.
View GitHub Release
Launcher hotfix

2.0.1.1

4 changes

Adds automatic launcher updates through the main application and improves recovery after failures.

  • The main application can update the launcher automatically in the background.
  • Launcher-update progress uses the existing dark bottom status area without a second confirmation dialog.
  • Failed launcher replacement or configuration updates roll back safely and retry on a later launch.
  • Ate a pudding while Codex was working. It was great.
View GitHub Release
Feature and usability update

2.0.1.0

9 changes

The first stability and usability update after version 2.0.0.0.

  • Improved BGM fade-out and karaoke transition coordination.
  • Improved speed, pitch, seeking response, and playback stability.
  • Improved lyrics offset controls, live synchronization, scrolling, and line seeking.
  • Expanded lyrics typography, spacing, reading, and surrounding-line controls.
  • Added Hiragana plus Romaji reading mode.
  • Improved animated theme previews and the appearance of built-in themes.
  • Standardized numeric inputs, button sizes, and interaction hints, and improved translations.
  • Improved the bottom status area to make BGM and karaoke automation visible.
  • Watched a lot of cat videos while Codex was working.
View GitHub Release
First public release

2.0.0.0

1 changes

The first public release of Singing Stream Savior.

  • Shipped as a complete installation ZIP; extract it fully and open the outer “Singing Stream Savior.exe” to begin.
View GitHub Release