Contents: Audio & Routing

Target version: 2026.43.1

Audio

Audio & Routing

CueFlow’s native audio engine supports multichannel output. Every track, the click, the count, and LTC can be routed freely to physical output channels.

Devices and multichannel

Choose the audio device, sample rate, and output channel count under “Devices & Timecode” in Settings. Multichannel devices are detected with their real maximum channel count (e.g. 16ch), selectable in the output channel setting. If the saved device is not present on this machine, CueFlow automatically uses the current default device for playback. Sound routed to output channels that do not exist on the current device stays silent on this machine. The saved device name and routing remain in the project, so the original routing comes back when that interface is connected again. Multichannel output, delay, and EQ run on the native engine. If the native engine is unavailable, CueFlow switches to stereo playback. Multichannel output and timecode sync require Pro or above, or an active organization editor seat (Free is limited to 2ch and 2 audio tracks).

Audio device settings, showing the connected device and its real channel count.
Audio device settings, showing the connected device and its real channel count.

Record onto an audio track

Recording input settings opened by right-clicking REC
Right-click REC to select the input device, mono/stereo mode, and input channel.
  1. Right-click REC on the destination audio track and select the input device, mono/stereo mode, and channel. Click REC to arm it; the track meter now shows input level. On Mac, grant microphone access when prompted the first time.
  2. Enable the transport Record button (●, Shift+R), then play. Enabling Record during playback punches in at that position; pressing it again punches out while playback continues.
  3. Stop or pause to commit the take. Audio is saved in the project’s Audio folder and placed at the recorded position. Esc discards the take; for takes longer than 30 seconds, recording stops before a discard confirmation is shown.
  • Availability — Record one track at a time while playing a song through the native audio engine. Recording is unavailable in Slow Mode, while chasing timecode, or during Web Audio playback. A song change ends the take and does not start another. The maximum take is 2 hours; takes under 0.25 seconds are not saved.
  • Monitoring and alignment — CueFlow does not monitor the input back to your outputs; use the interface’s direct monitoring. Input/output latency is compensated when placing the clip. To adjust it, use Settings → Devices & Timecode → Recording and set the input-device offset (negative if clips land late, positive if early).
  • Storage and overlaps — Input assignments and recording offsets stay on this computer. Recording over an existing clip trims the overlapped region; the original file remains and Undo restores the placement. Separate input and output devices may drift during long takes because their clocks differ.

Separate vocals and instrumentals

  1. Right-click an audio clip and choose Separate vocals / instrumental…. This Pro-or-above feature processes audio on this computer. On first use, review the model size, location, and license, then download the separation model (about 80 MB).
  2. Review the destinations for Vocal and Inst. Matching tracks are suggested; otherwise new tracks are proposed. Resolve any existing-audio or track-limit warnings before starting.
  3. The resulting Vocal / Inst clips retain the source position, trim, and fades. The source clip is muted by default. Files use names such as “Original (Vocal).wav” and “Original (Inst).wav”; one Undo reverses their placement.

Windows prefers a supported GPU through DirectML, with CPU fallback; Apple-silicon Macs use the CPU. Intel Macs are unsupported. Placement onto an existing track crossfades or trims overlaps and reports any fully covered clips that are overwritten. Progress and an estimate of remaining time are shown, and processing can be canceled.

The Audio Routing window

Routing is edited in the standalone Audio Routing window. Open it from the header’s Audio Routing button (under ⋯ in narrow windows), the View → Audio Routing menu, or ⌘/Ctrl+Shift+R. The window has two tabs.

  • Node editor — A graph view where you drag connections from sources (each audio track / Master / Click / Count / LTC) to output channels, making the signal flow easy to see.
  • Matrix — A grid view of the same routing. Clicking a cell cycles how that source feeds the output — Left / Right / Mono — and one source can feed several outputs at once. Buses can also be selected as input sources, enabling bus chaining (bus → bus).

Per-output delay and EQ

Each output can carry a delay for speaker alignment (10 ms steps), a graphical 3-band EQ, and a compressor. The matrix tab additionally offers per-output delay/advance in frames.

EQ and compressor

Open the EQ / Comp button on any output node to edit a 3-band EQ and a compressor. Shape and protect the sound for the room or the desk entirely inside CueFlow.

  • 3-band EQ — Adjust the Low / Mid / High bands while watching a response curve. Presets such as “Presence” get you in the ballpark quickly.
  • Compressor — Set THR (threshold) / RAT (ratio) / MK (make-up gain) / ATK (attack) / REL (release) with knobs to even out click and 2mix levels and tame sudden peaks.

Click and count routing

Beat Click and Count are independent sources. A Metronome file added by an AI import is routed as an ordinary audio track. You can send click only to performer monitors and the stereo mix to the PA.

Timecode I/O

Timecode send and receive are configured from the Timecode track inspector in the timeline. CueFlow can output LTC as an audio signal to any physical output, output Art-Net Timecode, and receive/chase LTC, MTC, or Art-Net. LTC routing is set with the Output 1 shortcut or the Audio Routing window.

  • Frame rates — 24 / 25 / 29.97 / 30, including drop-frame (29.97df) counting.
  • Send settings — Enable LTC Output / Art-Net Timecode Output on the Timecode track. LTC level, Solo, and Output 1 / Audio Routing are shown in the same panel.
  • MTC output — Enable MTC output in Settings → Devices & Timecode and select its MIDI output port. With no specific port selected, it uses this computer’s global MIDI output.
  • Receive protocols — Choose exactly one of LTC Audio / MTC MIDI / Art-Net. The inspector also shows receive status, detected format, LTC input device/channel, and per-channel input meters.
  • Chase modes — Choose Monitor / Freewheel / Chase Stop. While chasing, tune receive offset, frame tolerance, and minimum deadband; CueFlow moves to the song and position matching incoming TC. Small locked drift follows the external clock with continuous rate correction instead of sudden jumps.
  • No double output — While the native engine is playing, the fallback LTC generator is stopped automatically so LTC is never output twice.
The Timecode track inspector. Enable send paths (LTC / Art-Net) and receiving from one place, then expand only the settings you need.
The Timecode track inspector. Enable send paths (LTC / Art-Net) and receiving from one place, then expand only the settings you need.

Sending: LTC / Art-Net

The send section lets you enable LTC Output and Art-Net Timecode Output independently. Active outputs are highlighted in yellow. For LTC, the same panel holds the output channel, Audio Routing shortcut, output level, and Solo. For Art-Net, set broadcast mode, port, and individual destinations. Timecode continuity across songs is on by default: LTC / MTC / Art-Net keep advancing through continuous automatic song boundaries. Intentional gaps and timecode-excluded items instead move to the correct scheduled time in the show. Seeking also moves outgoing TC. Master volume and master mute do not stop LTC; use LTC’s own enable and level controls.

Receiving: LTC Audio / MTC MIDI / Art-Net

Receiving is exclusive: choose one of LTC Audio / MTC MIDI / Art-Net. With LTC Audio, check receive status, chase mode, input device/channel, level, signal waveform and quality, receive offset, and chase correction. The LTC input device and channel always belong to this computer and are not overwritten by collaborators. With default selected, check the actual device name; select the intended device explicitly if no signal arrives.

Chase correction

Chase correction absorbs small lock jitter while following the external clock naturally. Frame tolerance absorbs decoded boundary jitter; minimum deadband ignores tiny differences before correction starts. When drift exceeds the effective threshold, CueFlow corrects with continuous rate adjustment where possible and only resynchronizes directly when the difference is too large. Receive-offset and advanced-correction values apply only on Enter, not by dragging or scrolling. Use the resync button to reapply the current offset. The drift display shows the remaining offset and correction state.

Chase timecode across multiple shows

While incoming TC stays within the current show’s range, CueFlow stays in that show. The range includes its first through last block and roughly 10 seconds of trailing allowance (ending earlier if another show starts there). Outside that range, a matching show is selected after a short confirmation; the first match in show order wins. If several shows use the same TC range, there is no automatic switch while the current show still contains it.

To keep the current show, enable “Do not switch shows to follow timecode” in the Timecode inspector. This setting belongs to this computer. Incoming TC outside the current show is then treated as outside the setlist.

Check timecode input and output

  • Match the sender and project frame rates, including DF versus NDF at 29.97. A persistent mismatch between input and project settings produces a warning.
  • For LTC, inspect the waveform, jitter, and error rate as well as signal level. Select an input if the device is unset; a red output-route warning means the destination or device needs attention.
  • Returning this computer’s output to its own input can drive its playhead. Check that the input is connected to the intended chase source.
  • Timecode input, output, and routing can be edited during playback and apply immediately. Receive-offset and advanced-correction numbers require Enter to apply.

When audio is missing or Web Audio appears

  • Check the file and output — Confirm that the audio is present locally, the clip is not marked unlinked, and mute/solo and destinations are correct. Channels that do not exist on this computer cannot produce sound.
  • Return from the fallback engine — If native audio is unavailable, playback switches to stereo Web Audio and a badge appears beside the header timecode. Native output routing is disabled. Use Retry, stop and play again, or change the output device to attempt recovery. After repeated fallback, the badge shows the wait before another retry.
  • Save diagnostics — Open Settings → Devices & Timecode → Audio Reliability to inspect the state and fallback reason. Press Save stability diagnostics immediately after a problem. Diagnostics stay on this computer and are not sent in the background. Sending an in-app support request includes a summary of recent diagnostics; attach the full file when needed.

Caution

The PANIC button (⇧Esc) forcibly disables timecode I/O and persists that state, so timecode cannot resume unintentionally after an emergency stop. Re-enable the required paths from the Timecode track inspector; enable MTC output under Settings → Devices & Timecode. Timecode I/O is also suspended automatically during Slow Mode.

Tip

A classic show setup keeps LTC and the click on physical channels separate from the PA 2mix. The matrix tab shows every source’s destinations at a glance, which makes pre-show checks much faster. To review receive settings, select the Timecode track in the timeline and open its inspector.