Already announced before NAMM (but the video below) is from NAMM 2017 here’s the Bark Filter from Verbos Electronics. It’s named the Bark Filter due to the scaling of the frequency bands used which come from the Bark scale. I’m sure (as with every other Verbos module I’ve used) that this will be killer and you can hear it in the Analogue Zone video below but I wanted to delve into the Bark Scale and using non standard spacings for filter banks and EQs.
I first came across an unusual (or rather uncommon) frequency spacing in the Serge Resonant EQ. Without having an original I jumped at the chance to get one in eurorack format from Low-Gain, Clark and Manhattan Analog with the module pictured to the left. It’s frequency aren’t split like a normal EQ that may have the bands 55Hz, 110Hz, 220Hz, 440Hz, 880Hz and so on. The reason in the case of the Serge EQ is that normal EQs reinforce a musical scale or key. In the numbers listed above those frequencies are all the note A with the frequencies split at octaves all reinforcing the same scale or note. There’s some great blurb for the Serge EQ on the Thonk store page for the DIY kit (not available at the minute) … erm … that page isn’t around anymore and I can’t find it. Which is gutting as it was a good read about the scale. But you do get the idea from the random*source Res EQ module blurb quoted below.
Except for the top and bottom frequency bands, the bands are spaced at an interval of a major seventh. The Resonant Equalizer is designed to produce formant peaks and valleys similar to those in acoustic insstruments.
Until the Verbos module came to life I’d not heard of the Bark scale so doing some digging (god I love the learning that the modular world continues to invite you delve into) and here’s a summary. The Bark Scale was first proposed in 1961 by Eberhard Zwicker (not a Mr. Bark as you may think) however it is named the Bark scale as it was named after Heinrich Barkhausen who proposed the first subjective loudness measurements. Above 500Hz the scale is very similar to a logarithmic curve but below and increasingly becomes more and more linear. The scale has 24 bands that correspond to the 24 critical bands of hearing. The idea of these critical bands was introduced to audiology and psychoacoustics by Harvey Fletcher invention of the in 1933 before it was refined in 1940. Interestingly Harvey Fletcher is known as the father of stereophonic sound and credited with an early electronic hearing aid and alongside Wilden A. Munson they determined the Fletcher-Munson curves … which is something I imagine some of you have heard of. But back to the critical bands … The critical bands describe the frequency bandwidth of the “auditory filter” created by the cochlea which is an organ within the inner ear. Each of the 24 critical bands has a varying bandwidth but a rounded version (so the quote below to see why these are rounded) of the center frequencies is as follows. 60Hz, 150Hz, 250Hz, 350Hz, 450Hz, 570Hz, 700Hz, 840Hz, 1000Hz, 1170Hz,1370Hz, 1600Hz, 1850Hz, 2150Hz, 2500Hz, 2900Hz, 3400Hz, 4000Hz, 4800Hz, 5800Hz, 7000Hz, 8500Hz, 10500Hz and 13500Hz.
In his letter “Subdivision of the Audible Frequency Range into Critical Bands”, Zwicker states:
“These bands have been directly measured in experiments on the threshold for complex sounds, on masking, on the perception of phase, and most often on the loudness of complex sounds. In all these phenomena, the critical band seems to play an important role. It must be pointed out that the measurements taken so far indicate that the critical bands have a certain width, but that their position on the frequency scale is not fixed; rather, the position can be changed continuously, perhaps by the ear itself.”
Thus the important attribute of the Bark scale is the width of the critical band at any given frequency, not the exact values of the edges or centers of any band.
The Verbos Bark Filter doesn’t work with varying the bandwidth but uses the idea of these rounded/approximated center frequencies to form the 12 band fixed filter that the module is. The filters used here are steep 6th order band pass filters and you get a huge range of control and data out of the module. Bands are split into even and odd where you can take in one signal and split the output into the odd and even bands or you can even use two different inputs into the odd and even bands. There’s an envelope follower per band (awesome for extracted CV data out of dynamic and broad band sources such as a drum kit or even a full patch) and also a CV over each band. You can control the decay of the envelope followers and also shift and tilt the EQ in the same way as you can control the first 8 harmonics on the Verbos Harmonic Oscillator. As I said above every bit of Verbos kit I’ve touched is awesome, super high build quality and amazing sound quality. One trick mentioned by Mark Verbos and also brought up regarding the 4ms SMR (although that’s only 6 bands in total, but a very different module I should add) is that you can do basic vocoding with the module. A vocoder is simply an analysis system for extracting information and a processor for applying that to something else. So using the odd input as analyzers you can then use the envelope follower outputs to control the amplitude through the CV inputs of the even bands. There’s even switching to do this on the module which will save a mound of cable spaghetti. I’m excited to check out the module in person hopefully sometime soon and also to see Mark Verbos again at Superbooth in April. It was great to meet him and chat about ideas and his work in person and even better to get him on the Modular Podcast show to talk about origins and applications of Random which I’ve linked below too.
I should add that quotes are from Wikipedia and are all linked through. Like most my digging involved google, Wikipedia, clicking on and checking sources etc. I hope that inspires some digging and a bit of your own research too.