1. How are Sonicsmith’s synths different from other synths before them?
Sonicsmith’s synthesizers are different in the way you can control them.
They are based on a brand new kind of oscillator, the Audio Controlled Oscillator (the ACO).
Regular oscillators until today (VCO’s and DCO’s) were controlled either using voltage control or digital control. Audio Controlled Synths means you can control the frequency of the oscillator (its pitch) by an incoming analog audio signal. Then you can also influence the ACO’s pitch with two, separate CV inputs that allow you to arpeggiate the pitch of the ACO for automatic melody effects. While the pitch and “gate” of the audio are extracted by the ACO chip, two ENV followers will extract both the main audio input level changes and the side-chain input’s level changes for an alternative envelope path. To find out more about the ACO chip, read this. To find out more about our Audio-Controlled Synths read this.
2. How is using Sonicsmith’s Audio Controlled Synths different from using CV / MIDI controlled synths?
In synthesizer history there have only been 2 common kinds of oscillators: the VCO and DCO. Those are still in use today. Throughout the decades there have been a few attempts to make audio pitch to CV to control a VCO but these have never been practical and satisfactory for musicians for a few reasons. The ACO chip solves these problems and offers several other distinct advantages.
- Tune: The ACO is “frequency-locked” to the input so it doesn’t need to “warm up” and never loses tune, unlike analog VCOs
- The ability to modulate (arpeggiate) the oscillator’s frequency via CV while the main pitch is controlled by another source. This has never been possible until now.
- Size: The ACO is performing a few functions that previously needed large amounts of electronic parts and space to perform. It does that with a 5x5mm footprint.
- Power requirement: Much like the size advantage, the power supply to the ACO chip can be as low as 4.5V or in the case of Sonicsmith’s synths, a 9V standard PSU. It makes it much smaller and simpler.
- Power consumption: The ACO consumes a mere 20mA max, which makes it ideal for battery-power applications. Sonicsmith’s synths consume around 30mA (ConVertor) and 50mA (Squaver P1).
- Latency: The ACO chip’s latency is difficult to spec in terms of raw milliseconds. It locks to the input frequency within a matter of cycles, so the locking speed depends on the input frequency — but there are no perceptible negative time-domain artifacts that could be termed “latency” in the traditional sense, even with very low-frequency input like from bass guitar or cello.
3. How can I use Sonicsmith’s Audio Controlled Synths such as the ConVertor and the Squaver P1?
There are many ways one can use the Audio Controlled Synths. Here are just a few of the obvious ones:
- Instrumentalists can plug their electric instrument or microphone into the main audio input of the ACS allowing them to play the internal synth and letting them also control other analog synths with its pitch CV, envelope CV, gate CV and trigger CV outputs.
- Studio producers can hook-up their ACS much like an “external effect”. Choosing an audio output from the sound interface to the main inputs of ACS’s and the Synth output of the ACS back to an input of the sound interface (or straight to PA in a live show). Then controlling the ACS using any software instrument as a tone generator that is routed to that ACS input.
- Analog synthesizer musicians can use the ACS plugged to any audio output (sound interface out or oscillator output) to mix the ACO signal and add more voices or to convert any soft-synth to CV signals without the need for a MIDI-to-CV module.
4. Can I really “play” Sonicsmith’s Audio Controlled Synths with any instrument?
Yes, you can! We have tested the ACO technology with guitar (of course), voice via an unamplified microphone, trombone, kalimba (african thumb piano), fretless guitar, electric cello, bass guitar, and even acoustic harmonica.
Some instruments benefit more from the dedicated low-pass filter we have provided for improved isolation of the fundamental frequency of their input audio. For best tracking, it may be helpful to use the dedicated expression pedal input dynamically to optimize the input-LPF setting over the range of notes the musician plans to play. In this case, it’s easy to calibrate the LPF expression pedal to cover exactly the desired range of notes and not difficult to learn to “ride” this expression pedal while playing.
5. So your audio-controlled synths are basically like a vocoder, right?
Audio-controlled synths may remind you of some vocoders at first glance, but PROBABLY only because we have demoed them using voice + microphone and the internal VCF can impart a “vowel voicing” (or wah-wah type) sound when you modulate its cutoff frequency. However, the underlying technology making them work is completely different than what’s inside a vocoder.
First of all, a vocoder takes two audio inputs — input A is usually a voice input and input B is usually a “synth” input (from an external synth source). A vocoder, in contrast with our synths, does not generate its own synth sound internally. Input A is analyzed through a bank of fixed band-pass filters, which basically detects the energy levels present in input A at each of the filter frequency bands. The level envelopes are detected from these band-pass signals and used to control input B which has its own set of band-pass filters tuned to the same frequencies as the first set. Input B (the synth signal) is passed through these 2nd set of band-pass filters and each band goes through an independent VCA. Finally, the VCAs on input B are all controlled by the envelopes extracted from input A (the voice). That means they impose the same levels as detected on input A so the synth sound will have similar frequency response as input A (hence the synth will sound like the voice) If this sounds complicated and expensive, that’s because it usually is. Two sets of band-pass filter banks is a lot of analog hardware. This explains the high cost of most analog vocoders. Keep in mind that a vocoder doesn’t analyze nor care what is the fundamental frequency (AKA pitch) of any of its audio inputs.
Our audio-controlled synths, on the other hand, have ONE main audio input with the purpose of detecting its fundamental frequency. This input can be anything — voice, guitar, cello, brass etc. They do have a second audio input (the side chain input) but that input is not relevant to this discussion and is only an option. The internal ACO (audio-controlled oscillator) inside our synths generates square and sawtooth waves that are immediately in sync with the fundamental frequency of the main audio input (well, not IMMEDIATELY, but latency is so low as to be undetectable). We do also have an envelope follower which you can then use to control the VCA but it performs this operation over the entire audio bandwidth (no complicated band-pass filter bank required like in a vocoder). This ENV follower plus VCA is the one feature our synths have in common with vocoders, so our synths can reproduce the dynamics of voice or other instrument intonation like a vocoder does. We do have a VCF that we think you’ll like, but it’s similar to the VCFs you’ll find on a lot of analog synths and also doesn’t have much in common with a vocoder.
Our philosophy is that if you want to use our synths to make sounds that seem similar to what vocoders can do, we encourage you! Making this kind of experimentation possible is exactly what makes electronic music great in our opinion. Please note that this is just a small fraction of what our synths can do. We think that the most creative possibilities open up when you configure our synths to make sounds that aren’t similar to anything else you’ve heard — and we think you’ll agree.
6. So your ACO is really just a PLL, right?
The ACO is NOT based on a PLL but rather a digitally-assisted analog FLL (frequency-locked loop). The circuit was awarded a patent this year and it will be published before the end of June so all the gory details will be known then. We have heard several claims that the ACO sounds “similar” to a PLL — perhaps that is true for very simple playing — but here are three main reasons why a PLL will never compete with the ACO:
- To make PLL stable, the loop bandwidth (which is directly related to the tracking time) has to be maximum 1/10 the lowest reference frequency you will use. Let’s say the lowest reference frequency is 25Hz. So the loop bandwidth can be 2.5Hz MAX. This means tracking time is on the order of 1/(2*pi*2.5) = 64ms and most likely longer. This is gonna be easy to hear, especially at high audio frequencies.
- PLL depends on PHASE locking (hence PHASE-locked loop). This means when you have a large jump in reference frequency, you can get huge accumulation of phase error and your PLL will SLEW. This means in most practical cases, the locking time is going to be WAY longer than the 64ms quoted above! If you tried to drive a PLL with alternating low and high frequencies, you would most certainly hear this. It can be used for pleasant musical effect but only if you want such an effect — and we have found that the majority of musicians who like our ACO-based solution love it precisely because it doesn’t add any coloration or slewing or non-idealities to their playing, it duplicates EXACTLY their dynamics, gestures and agility.
- The tracking range of our ACO is 25Hz to above 6kHz. You CAN design a PLL to track over such a wide range (with some variant of the 4046 chip for example), but if you do that chances are you’ll have to make a lot of compromises in the design that will make the tracking time issues explained in (1) and (2) even worse.
Gabriel Marin’s guitar playing in our NAMM 2017 booth is a good example of where our ACO exceeds the performance of what a PLL could do.
7. What are the limitations of the ACO chip and the Audio Controlled Synths?
The ACO chip can follow the audio’s pitch between 25 Hz – 6.4 kHz. That is still much larger than natural instruments frequency range existing today.
The ACO is monophonic, meaning that it can only track one frequency at a time. One can play “chords” into our audio-controlled synths and one of the following things will happen: (a) If the input LPF is set appropriately and there is enough filtering of the higher notes, the ACO might track the lowest note; (b) the ACO will try to track a signal with non-uniform zero crossings and will “frequency modulate” between two or more notes (not necessarily the notes of the chord). The results of (b) can be pleasing or not, depending on individual taste. We encourage experimentation and are sure that many musicians will learn how to create pleasing results with polyphonic inputs!
Another limitation has to do with the ACO gate output generator. The gate CV turns “on” when the audio amplitude exceeds the threshold of about 3mV and will only turn “off” when the audio falls back down below this threshold. That means that string picks that do not start below that threshold might not get a “new note” gate output and will continue the previous note gate. In order to retrigger the gate CV the audio must fall below the threshold before a new note starts.
8. Can I control Sonicsmith’s Squaver P1 and ConVertor synths with MIDI?
The only way to control the PITCH of the synths is by using an analog audio input.
Of course you can use a MIDI controller to control any hardware or software tone generator or sampler and connect their analog audio outputs into the audio controlled synth.
The ENV, filter PWM and most other parameters of the synth can be controlled either from the main input, CV inputs as well as with an alternative audio input (the side-chain input signal) converted to CV with the second ENV follower.
9. Can I use Sonicsmith’s Squaver P1 and ConVertor with my Eurorack modules?
The pitch CV output of our synths is calibrated to be 1V per octave.
Notice though that our synths work on a 9V power supply so our outputs will have the following voltage ranges:
Pitch CV: 0-8V
ENV CV: 0-9V
Gate/Trigger CV: 0-9V
10. How are the new ConVertor+/Squaver P1+ units different from the older ConVertors/Squaver P1s?
- The biggest difference between the ConVertor and ConVertor+ is the conversion of the input low-pass filter frequency shift knob (on the left directly below the IP GAIN, or preamp gain, knob) into a gate threshold adjustment knob. This is most useful in situations with lots of background noise (like microphones in a live performance situation). The same change was made on the Squaver P1+ vs. the Squaver P1.
- The 3-way input high-pass filter (HPF) switch has been replaced with a 2-way 12dB/24dB per octave selection switch for the input auto-tracking filter. This allows the user to select more filtering for instruments with higher harmonic content or less filtering for faster transient response. The input HPF in the plus versions has been fixed at 16Hz.
- The main ENV out on the plus versions incorporates an additional smoothing filter which makes it sound smooth enough even to modulate the amplitude of a sine wave.
- The Squaver P1+ PWM control now spans the full range from 0-100% duty cycle, with 50% roughly at 12 O’Clock. This gives a much more dramatic sound when ENV is routed to PWM, for example.
- We added 2.5dB gain to both the main ENV and side-chain ENV outputs in the plus versions. This allows us to clip the envelope at close to +9V and also allows the ENV to fully reverse to a silent (almost 0v) state when the envelope is clipped and ENV AMT is set fully counter-clockwise.
- The plus versions are painted using a much more durable paint process, which we expect will last for many years.
- The original versions will be offered for sale at 40% off until stock is exhausted, whereas the plus versions will be introduced at higher prices ($419 for ConVertor+ and $729 for Squaver P1+ in North America; in Europe prices will be higher because of VAT).
- Both the original versions and the plus versions now have the improved higher gate threshold for the ACO chip so are equally able to reject low-level noise in the input. The plus versions just have the additional gate threshold adjustment available which is described in point (1) above.