Sensei Caswell: “Our SEM is known as a ‘Voltage Controlled State Variable Filter’ because it provides simultaneous high pass, low pass, and band pass outputs, all with a 12 dB/oct slope. The notch is the inverse of the bandpass, achieved by summing the low pass and high pass outputs. This version is a combination of the original SEM filter and the newer design from the OBX, along with improvements by SE. It has a modest current draw, which it pulls evenly from the plus and minus rails.”

If you’ve heard the SEM filter in our Boomstar desktop (https://soundcloud.com/studio-electronics/sets/mean-sem-ranch/) you also know that this 12 dB favorite is “capable of lighter, more subtle tones than the other three models. This isn’t to suggest the filter’s Low Pass mode can’t cater for fat and punchy, because it can.” – Paul Nagle, Sound on Sound. Looking for deeply gratifying tonal variance and this emotional warmth? Try https://soundcloud.com/alanc3-1/switched-on-sem—our friend Fred Falke’s favorite—it can reach into your head and heart, and stay as long as desired. Thank you Lord Tom Oberheim for this marvelous design/template from which we were able to work our very own Studio Electronics eurorack filter enchantments—we do like to hit our SEM a bit harder though 😉

Greg St. Regis on the harder hitting: “The ‘bonk out’ or sound crash, is a feature, not a fault. We intentionally allow for this to happen by not limiting the input level, but rather boosting it and the resonance level. Some very interesting tones can surface when it pops back and forth between normal and bonk. As most filters do not have an input level, but preset it with a fixed resistor, they are adjusted to eliminate this; it doesn’t hurt anything when bonk happens: it is an extreme overload. If you don’t want ‘blackouts’ to occur, lower the input level. As you add more waveforms or oscillators, the level is increased, so you will have to adjust the input level accordingly regardless for a cleaner tone. All the filters are designed to push the input level this way, but only the SEM and SE88 fall apart with too much level. The beauty of this feature, is that you benefit from all the loveliness of a traditional SEM filter by keeping the input and resonance level short of maximum, and have the power to twist the once boxy, gentle circuits into a demented, distorted, peaking and vanishing, snarlingly wild beastly sound. Full variety is what we have provided. GSR”

Tim Caswell on the easy bonking: “Right. This filter is a ‘State Variable’ circuit, which is an analog computer model of a pendulum. When the resonance is increased, the “damping” or stabilizing of the circuit is actually decreased. This makes the pendulum swing more wildly. If it is hit hard enough, it will swing all the way to one side and ‘stick’ there, what we call ‘bonk out’. Greg likes the sounds that can be gotten as it goes into and out of ‘bonk out’, so, against my objections, we have not made the filter ‘idiot proof’ by limiting the input level and resonance amount. If you don’t like it, limit your input level and/or resonance amount. Otherwise, go wild! It won’t hurt the circuit. TC”

This filter (and the SE88) can be “input driven” to the cutout/cutoff/blackout point, for angry, aggressive unpredictability, and staggered, shell-shocked recovery. We wanted it that way; it’s perfectly harmless OpAmp smashing, for smashing good fun. Think traumatic movie scene after which an actor’s hearing momentarily disappears (extreme wooziness overtakes the balance of his senses—and camera work), high-pitched sounds follow, and then blessed normalcy of awareness and sensory perception gradually, grudgingly return. That. No reboot necessary: just sweep the frequency counterclockwise until the SEM ‘n SE88 snap back to their “right mind,” and rein in the input and resonance levels—resonance levels below 9:00 are safe; unless of course you want “that” to keep happening. Taunting the edge is entertaining too.

* From 7:00 (0) to 9:00 our “Negative Resonance Saturation” adds beefiness, boosting the waveform amplitude and taming waveform transients (most pronounced in the 3003 filter): set to 9:00 to achieve the cleanest tone possible.