Delay times fall into perceptual bands, and the boundaries are sharper than the continuous knob suggests. Above roughly 50 milliseconds you hear a discrete repeat. Below it, you don’t hear a delay at all — you hear timbre, width, or movement, depending on how the delayed copy is modulated.
Verbos Electronics’ Haas Zone Processor lives entirely below that line.
Verbos Electronics’ own introduction to the module.
What’s inside
A multi-tapped digital delay bank, three taps:
| Tap | Range |
|---|---|
| Tap 1 | under 1 ms – 20 ms |
| Tap 2 | 1 ms – 20 ms, chained from Tap 1 |
| Tap 3 | 9 ms – 42 ms |
Plus:
- A voltage-controlled bandpass filter — a 2-pole high-pass and low-pass combination
- A voltage-controlled oscillator with triangle, square and sawtooth, 0–10 V
- An envelope follower, normalled to the delay time CV inputs
- Feedback routed from the bandpass filter output
Around $800, available now. Verbos describes it as usable as an audio processor, a patch tool, or a standalone voice.
Why that time range is a design decision, not a limitation
Mix a signal with a copy of itself delayed by a few milliseconds and you get comb filtering — a series of evenly spaced notches and peaks across the spectrum, at intervals set by the delay time. That is not an effect layered on top of the sound; it is a filter built out of time.
At under 1 ms, the comb’s first notch sits high in the spectrum and you hear a thin, metallic, resonant colouration — the sound of very short delay, and the basis of Karplus-Strong-style string models.
At 1–20 ms, you’re in flanger and doubling territory. Modulate the delay time and the comb sweeps: that is a flanger, and it is why Tap 2 is chained from Tap 1 — the two form a modulatable series rather than two independent voices.
At 9–42 ms, you’re in chorus and early-reflection territory. Long enough that the ear stops fusing it into pure timbre, short enough that it still reads as one event with body rather than two events.
And the module’s name points at the reason the top of that range matters: the Haas effect (or precedence effect) is the psychoacoustic phenomenon where two copies of a sound arriving within roughly 1 to 40 milliseconds fuse into a single perceived event, whose apparent direction is dominated by whichever arrived first — while the later copy contributes loudness and spatial width rather than a separate image. Beyond about 40 ms the fusion breaks and you hear two sounds.
Tap 3’s 42 ms ceiling is that threshold. The module’s entire time range is the fusion window, and its name is a specification.
The list of effects Verbos quotes — “chorus, flange, phase cancellation, doubling, ADT, waveshaping” — is just that window enumerated. ADT is Abbey Road’s artificial double tracking, invented in 1966 to spare John Lennon from singing a part twice; it is a short delay with modulation, and it is on this list because it is the same mechanism.
The two smart bits of routing
Feedback comes from the bandpass filter output, not the delay output. Feeding a filtered signal back means each pass through the loop is band-limited again, so resonance builds in a controlled band rather than accumulating into broadband screech. It is the same reason feedback-path filtering is standard in reverb design, and it is the difference between a feedback control you can use across its whole range and one that is a cliff at 70%.
The envelope follower is normalled to the delay time CV inputs. Normalled means the connection exists until you patch something else into that jack. So out of the box, the delay time tracks the input’s amplitude — louder passages shift the comb, so the timbre moves with the playing. That is a genuinely musical default, and it means the module does something interesting with a single cable in and a single cable out before you patch anything at all.
Where it fits
Verbos modules are the ones people buy for character rather than utility, and the house style is oscillators and processors whose behaviour is idiosyncratic on purpose. A short-delay processor is an unusual thing to build as a standalone module in a format where most delay modules reach for seconds and tape emulation.
The honest case for it: if your patches sound thin, or centred, or static, the fix is very often in this time range rather than in a reverb or a filter. Short delay is how you get width without stereo sources, movement without an LFO on a filter, and resonant body without resonance. $800 is a lot for a delay — and it is not really a delay, which is the argument the price rests on.