The Bark Filter Processor is based on the Bark scale, a psychacoustic perceptual scale of pitches, thus the module's twelve frequency bands originate from the Bark Scale: <100Hz, 300Hz, 510Hz, 770Hz, 1.08kHz, 1.48kHz, 2kHz, 2.7kHz, 3.7kHz, 5.3kHz, 7.7kHz and >10,5 kHz.
The lowest band uses a band pass filter, the highest uses a high pass, all other bands use band pass filters.
Besides an individual audio output each band also features a separate envelope follower CV output.
The levels of the filter bands can be set and modulated in a number of ways. On one hand for each individual band manually with the fader and with a CV which is injected into the socket below the fader. On the other hand there is the scan function konwn from Verbos' own Harmonic Oscillator which dynamically can control several bands, independent of the faders:
"Frequency Scan" sequentially scans through the frequency spectrum, with Center and Width determining the center frequency and bandwidth
"Spectral Tilt" either emphasizes the low or high harmonics.
Each filter band passes an individual envelope follower which emits a CV. The sum of all twelve EV-CVs is available at the all followers output, though.
The envelope followers' decay times are set globally which affects the quickness of the filterbank - longer settings might be fine for voice, shorter more suitable for percussive signals.
The module has two level-adjustable audio inputs, each one for the odd and the even filter bands which makes possible to use the Bark Filter Processor as a dual six-band filterbank. The inputs can be criss-crossed by normalizations in order to process the same audio signal with all frequency bands.
There are three mix audio outputs: one for all even bands, one for all odd bands and one for all frequency bands.
How to construct a six-band vocoder? One half of the bands is used as analysis section and is fed with speech for instance; the other half of the bands makes for the synthesis section and is fed with oscillator and noise signals. Now patch the analysis section's envelope follower outputs to the level CV inputs of the synthesis section - the vocoder signal will be preent at the synthesis output. That's it!
audio inputs: odds in, evens in (normalized criss-cross)
outputs: odd filters, even filters, all filters, all envelope followers
CV inputs: width, center frequency, spectral tilt
per filter band: level CV input, envelope follower outout, filter output
3U Eurorack module, 34HP wide