Output & Recording
This page covers everything about where your visuals go: sending them to projectors and other apps, recording your live set, and exporting a polished high-res render afterwards.
Watch: Output and recording — open the guided tour
Fullscreen & Projector Output
The simplest way to output: press F for Output View. The UI disappears and the master output fills the screen. Press Esc to come back.
To make the whole window fullscreen with the normal UI still showing, press F11 (again to restore the previous size and position). The two combine: F11 for a fullscreen window, then F for Output View on top of it.
For a second display or projector, use Preferences → Video & Display to configure a secondary output window you can drag to the projector screen and go fullscreen there independently.
Choosing the output screen: the Output → Secondary Output Window item fills one monitor. By default (Output → Output Monitor → Automatic) that is the first monitor that is not the system's primary one. If the video lands on the wrong screen, which can happen on Wayland where there is no true primary monitor, pick your projector under Output → Output Monitor. The choice is remembered by monitor name.
Output scaling: - If your shaders are demanding and your GPU is struggling (especially on a laptop), lower the internal render resolution in Preferences → Video & Display. Dropping from 1080p to 720p or 540p cuts GPU load significantly while the output still fills the screen. - Use Fit / Fill / Stretch to handle non-standard aspect ratios — useful for LED walls or vintage 4:3 projectors.
Sending Video to Other Apps (Spout / Syphon / PipeWire)
Liquid LSD can share its video feeds live to other VJ apps, media servers, and streaming tools — without any encoding overhead or latency.
Go to Preferences → Video & Display → Live Video Sharing and toggle which feeds you want to share:
- Deck A, Deck B, Deck BG, Deck PV (Preview), Master Output — enable any combination.
- Set a custom stream name for each (e.g.
LiquidLSD-Master,MainStage-Feed). - Choose a resolution for the shared stream, or keep it synced to the master.
- Pick how the image is scaled if the resolutions don't match (Fit, Fill, or Stretch).
The sharing method depends on your platform:
- Windows: Spout2
- macOS: Syphon
- Linux: PipeWire (DMA-BUF GPU sharing, with a memory fallback if the GPU doesn't support it)
On Linux with OBS: Add a PipeWire Screen Capture source and select the Liquid LSD video node. Use qpwgraph or Helvum to inspect and route video nodes.
Status: Check the Preferences → Video & Display panel for a live status indicator next to each shared feed showing which sharing method is active and whether it's working.
Recording Your Set (REC)
Press Ctrl+R (or go to Output → Record Output) to start recording. A red REC mm:ss badge appears on the title bar while recording is active, along with a dropped-frame counter.
Press the same shortcut again, or click Stop Recording, to finish. The output file lands in the recordings/ folder as an MP4, named with the date and time (e.g. liquid_lsd_2026-09-01_20-15-00.mp4).
What you get: H.264 video at your current render resolution and frame rate, with synchronized audio from your input. On capable hardware, NVENC, QuickSync, or VAAPI hardware encoding is used automatically.
You can also map the REC toggle to a MIDI controller button so you never have to touch the keyboard during a performance.
Offline Render Studio
The Offline Render Studio is for making video content — music videos, promotional visuals, or anything where quality matters more than real-time speed. It renders frame by frame at whatever resolution and frame rate you like, fully decoupled from the clock.
Open it via Output → Render Video (Offline)....
What you can set
- Audio file — A
.wav,.flac,.mp3, or.oggfile that drives the visual modulation. - Preset source — The current active deck, or any saved preset.
- Resolution — 1080p, 1440p, 4K UHD (3840×2160), or custom.
- Frame rate — Typically 60fps, but configurable.
- Motion blur — 2×, 4×, or 8× temporal oversampling. Each frame is blended from multiple sub-frame renders, producing smooth cinematic motion on fast-rotating mandalas and feedback loops. More samples = more render time.
- Output destination — Where to save the file.
What happens
The audio file is decoded up front and the audio analysis runs mathematically per frame — so the beat tracking, frequency bands, and envelope followers all behave exactly as they would in real time, just calculated offline. You get a frame-accurate render even on complex shaders.
A progress bar shows render percentage, elapsed time, estimated time remaining, and output file size estimate. There's also a live thumbnail so you can confirm it looks right before committing an hour to a 4K render.
Hardware Preferences & Persistence
Your audio routing, MIDI mappings, display preferences, and broadcast preferences all save automatically to lsd-preferences.properties in the app folder (with automatic backward compatibility for existing lsd-settings.properties files). You don't need to reconfigure anything between sessions.