ExtensionEvery voice SuperSampler spawns is registered in SamplerQuery: *playing as a SamplerVoice: a lightweight proxy holding the underlying Synth together with the metadata of what it is playing. Voices remove themselves from the registry when their synth ends.
Unknown messages forward to the synth, so code written against raw Synths — typically MIDI-CC control over playing sounds — keeps working unchanged:
.set(\amp, ...) and .set(\pan, ...) on a playing voice take effect immediately (lagged 50 ms). .set(\gate, 0) fades the voice out in 0.1 s and frees it.
Used internally when a voice is spawned.
The underlying Synth.
The voice's total amplitude factor in output units (the caller's amp × normalization gain × the section's recorded peak).
The absolute logical time the voice started.
The SampleDescript being played.
The section index within the sample.
Playback rate (negative when reversed).
The voice's expected wall-clock duration in seconds.
Identifies which .key / .playEnv call (gesture) this voice belongs to. See SamplerQuery: *releaseGesture.
Forwards to the synth (frees it immediately).
Forwards to the synth.
| releaseTime |
Fade time. |
Gate the voice off through the kill-gate (0.1 s fade, then the synth frees itself). The message travels with server error posting suppressed for that bundle, so gating a voice that just ended by itself does not print a "Node not found" FAILURE.