Hey everyone!
Ok, this is a big one. It is the first public build since 0.9.0, and a lot has landed since then: the whole record / mix / master chain got deeper, plugins run safer, and the timeline got a real workflow pass. Full rundown below.
If this is your first time seeing this, Dusk Studio is a deliberately constrained, portastudio-style DAW for Linux, macOS, and Windows. Fixed signal chain, finite track count, one page per stage. You commit, you move on.
Recording and editing
Audio region editor with non-destructive trim, fade, gain, and 8 takes per region.
Take cycling and comping (cycle through previous takes in place).
MIDI tracks, instrument plugins, and a full piano roll editor.
Bar and beat grid in both the audio and MIDI editors, with adaptive sub-beat ticks that get denser as you zoom.
Chase / Follow playhead: the timeline and editors scroll to keep the playhead in view during playback. Toggle per editor, or set it on by default in Settings.
Import TASCAM DP-24 / DP-24SD / DP-32 songs straight off the SD card (experimental): each fragment lands on its own track with the right rate, bit depth, and stereo pairing, plus recovered clip positions, mixer settings, tempo, and markers where the format allows.
Multi-file audio and MIDI import with a target-track picker.
Broad undo: piano-roll notes, automation breakpoints, tempo-map edits, and track / bus / aux rename and color changes.
Mixing and mastering
True-peak brickwall mastering limiter, rebuilt from scratch: 4x oversampled lookahead with a hard inter-sample-peak ceiling and a clean hold / release. The Threshold control now does what you expect (pull it down, push more into the ceiling).
Limiter Mode (Modern / Transparent / Punchy) and a Stereo-link toggle.
Mastering-EQ real-time spectrum analyzer (FFT overlay behind the response curve).
DP-24-style multiband compressor presets: nine genre starting points in the mastering comp.
Mastering section visual pass: higher-contrast knobs, depth tiers, a proper waveform with played / unplayed colouring and a time ruler.
Piecewise tempo map: songs can change tempo over time, and the ruler, grid, BPM control, piano roll, MIDI scheduler, and metronome all follow it.
Multi-output routing: Open more than two outputs and route each aux return to its own physical output pair for an independent headphone mix. Pick which output pair carries the master.
Console automation (Write / Read / Touch) on channels, aux, and master, with automatic cross-track plugin delay compensation so latency plugins stay aligned.
MP3 bounce and export at 320 kbps (name the output .mp3). WAV stays default; stems stay WAV for sample-accurate re-import.
Plugins and performance
Optional out-of-process plugin sandbox on all three OSes: run third-party plugins in a child process so a crashing plugin cannot take the session down. In-process stays the default for the most responsive editors; opt in with DUSKSTUDIO_USE_OOP_PLUGINS=1.
Plugin scanning is always sandboxed, and a bad plugin gets timed out and blacklisted instead of hanging the app on first scan.
Plugin-scan validation drops dead entries so the picker never offers a plugin that cannot load.
Multicore DSP setting: Auto, Off, or a pinned worker count, saved per machine.
Native soundfont playback (.sfz and .sf2 via the built-in sfizz engine, no external synth).
External hardware inserts per channel and per aux, with an auto-latency ping that runs while the transport is stopped.
Workflow and UI
Top-row redesign: the RECORDING / MIXING / MASTERING / AUX stage selector moved up next to the File and Settings menus; the timeline controls (Snap, zoom, Fit, Chase) sit in the transport row grouped with the bank buttons under the stage selector, and only show when the timeline is open.
Keyboard shortcuts cleaned up: number keys 1 to 4 switch channel banks, Cmd/Ctrl+1 to 4 switch stages, T shows / hides the timeline, and Cmd/Ctrl+E is the one split key everywhere. Press ? for the full in-app reference.
Native audio device selector with in-window errors (no native popups), tempo at the playhead shown on the transport, and the current song-section name next to the clock.
Marker rename prompt on creation, tape-strip track labels that auto-size to the longest name, and a cleaner timeline interaction pass (left-click to move the playhead, drag the ruler to set loop / punch).
Accessibility: every channel-strip control has a screen-reader title and value text (VoiceOver and Orca read "Track 1 Fader, -4.2 dB").
Update notice on launch, open a session from the command line or file manager, and an unsaved-changes prompt on every open path.
Platform and stability
Portable sessions: audio paths are stored relative to the session folder, so moving or renaming a session between machines just works.
Portable Linux tarball plus an install.sh that registers the launcher, icon, and file association (replaces the old AppImage). Raspberry Pi (arm64) builds now ship alongside x86_64.
Windows ASIO-first device selection, Windows MSI with Start Menu and desktop shortcuts, hot-unplug detection that force-stops cleanly and keeps the session intact.
A long list of crash, real-time-safety, and data-integrity fixes: atomic session save, autosave recovery that actually persists, MIDI delivered off the audio thread to kill xruns, send-effect tails no longer cut off, and a fixed bounce dialog that used to freeze on cancel.
Linux fixes: ALSA real-time priority, a crash on sub-32-channel interfaces, Wayland edit cursors, the device-busy fallback so a session always opens on a working device, and the app icon now shows in the dock instead of a generic one.
How to get it
The build is live for all patrons. Here is how to grab it:
Head to https://builds.duskaudio.com/latest
Click Sign in with Patreon and authorize
Pick your platform (macOS, Windows, Linux) and download
As long as your $1+ membership is active, you are in. No codes, no special links, just sign in with the same Patreon account you are reading this on. This is my first attempt at this, so let me know if you run into issues trying to download from the link.
First launch: the builds are unsigned, so macOS and Windows will warn you the first time. On macOS, right-click the app and choose Open. On Windows, click More info, then Run anyway. About 30 seconds, once. Full walkthrough: https://builds.duskaudio.com/manual
Links
Download the binaries (patrons, $1+): https://builds.duskaudio.com/latest
Source code (build it yourself, free, GPL-3.0): https://github.com/dusk-audio/dusk-studio
Found a bug? Let me know: https://github.com/dusk-audio/dusk-studio/discussions