Unified Mask

Parameters

Freq Min — Low edge of the kept frequency band in Hz, 20–24000 (default 20). Same role as Frequency Band Mask’s low_freq. Auto-clamps against freq_max.

Freq Max — High edge of the kept frequency band in Hz, 20–24000 (default 24000). Same role as Frequency Band Mask’s high_freq. Auto-clamps to Nyquist at execution.

Freq Feather — Soft transition width on the frequency edges in Hz, 0–500. Smoothstep cross-fade. 0 = hard edges.

Amp Min — Lower amplitude bound in dB, −96 to 0 (default −96 = no lower limit). Same role as Amplitude Mask’s amp_min. Auto-clamps against amp_max.

Amp Max — Upper amplitude bound in dB, −96 to 0 (default 0 = no upper limit). Same role as Amplitude Mask’s amp_max.

Amp Feather — Soft transition width on the amplitude edges in dB, 0–20. Same role as the other masks’ feather.

Additional controls

Time Min — Time range start in seconds, 0–(input duration). Static (not animatable — modulating the time bounds doesn’t make sense for a non-time-varying selection). Auto-clamps against time_max.

Time Max — Time range end in seconds, 0–(input duration). Auto-clamps against time_min. Defaults to the input’s full duration when one is connected.

Time Feather — Soft transition width on the time edges in frames (not seconds), 0–50. Static. The widget shows it as a frame count rather than ms because the smoothing operates per STFT frame internally.

Invert — Swaps the two output ports — same convention as the other mask nodes. Default off (Out1 = Extracted, Out2 = Rejected); on flips them.

Trim — When on, the Extracted output is trimmed to [time_min, time_max] — the resulting buffer is shorter than the input. When off (default), the Extracted output keeps the input’s full duration with bins outside the time window zeroed. Use trim when downstream nodes care about buffer length (e.g. you only want to feed the selected segment into a Slicer); leave off when you want to preserve timing for parallel mixing.

About Unified Mask

Unified Mask is the master mask node — combines all three selection dimensions (frequency, amplitude, time) into one. A bin/frame is selected only when it satisfies all three criteria simultaneously: frequency within [freq_min, freq_max], amplitude within [amp_min, amp_max], time within [time_min, time_max]. Each dimension has its own feather control for soft edges. The combined weight is the product of the three per-dimension weights, so feathered regions cross-fade smoothly in 3D selection space.

Use cases this node uniquely enables (vs. chaining the simpler masks):

• “Loud bass hits in the chorus only” — set freq_max = 200, amp_min = −20, time = chorus range
• “Quiet treble breath in the verse” — freq_min = 5000, amp_max = −30, time = verse range
• “Surgical removal of one vocal phrase” — freq = 200–4000, amp = −20 to 0, time = phrase, then take Rejected output

Time bounds are not animatable by design (they describe a fixed selection window, not a modulation source). Frequency and amplitude bounds are animatable — modulate them with envelopes / LFOs / sidechain for evolving selections. Pair with the simpler Amplitude Mask or Frequency Band Mask when you only need one dimension; this node is heavier (more knobs, slightly more compute) and overkill for single-dimension work.

Generated 2026-05-05 from K2K_Dev@96730bdc by scripts/gen_lexique.py. Edit _intros/ or _overrides/, not this file.