Review: Korg Minilogue

In Korg’s impressive lineup of synths today, the products ending in “-logue” are the analogue/analogue-hybrid keyboard synths. On one side of the spectrum, there’s the Monologue (as the name implies, a monophonic synth). Then there’s the two variants of the Prologue, 8/16-voice analogue/digital hybrids.

In between sits the Minilogue, which although the “Mini” in the name might suggest differently, is a proper four-voice analogue polysynth. I say “proper” because it’s not a paraphonic configuration option like with some other products past and present. And with a street price at around €550, it’s far cheaper than any other polysynth available today.

I already had a brief look at it before, and back then identified it as something to further consider. Which I did. I got it.

First Impression

When unpacking the synth, your first impression is probably the brushed metal faceplate with the black knobs and silver levers. At a size of around 50x30x10cm (including knobs), it sits in the same area as e.g. a Minibrute 2 or, to stay within Korg’s legacy, is less wide but deeper than a Prophecy, and a bit higher but otherwise almost identical to a Taktile-25.

Top View.jpg

Looking from the rear, the defining feature is a fake wood panel, which, although its usefulness can be debated, sets a nice contrast to the rear connection panel and other synths you might have.

Connection-wise, there’s audio in/out and headphones in standard 1/4”, sync in/out in 1/8”, a MIDI in/out pair with diode jacks, a USB port (USB2, MIDI/data transfer only) and connection for the external PSU.

Rear View.jpg

Returning to the top, the first thing that catches the attention is the playing aides, or rather the lack of them: instead of the typical baseline of pitch and modulation wheel, we only find a lever which moves in a diagonal angle over the left side of the keyboard. No modulation wheel.

Powering it up, the small display shows the name as well as a progress bar for “tuning”, before you can start to have fun.


Let’s start with the voice architecture: each of the four voices comprises two “almost” identical VCOs, a lowpass filter switchable between 2 and 4 poles (which would lead to roughl 12dB/oct vs. 24dB/oct) and of course a VCA. As for modulation sources, there’s two ADSR envelopes, one of which is hardwired to the VCA, and one LFO.

The second envelope can control filter cutoff, oscillator 2 pitch and either LFO rate or modulation intensity (although that one is only an on/off switch).

I said “almost” identical about the VCOs before, because there’s several modulation options, all of which have VCO2 as the target or analyze source: in addition to the usual sync, there’s also cross modulation (a kind of frequency/pitch/phase modulation) and a ring modulator.

A third source is a noise source without configuration options. And then there’s the audio in.

Before we go into the details, there’s one important feature on this synth: parameters do (internally) not use 7-bit resolution (i.e. go from 0-127 as they did since ancient times), but rather a 10- or more bit resolution. And that is a really important advantage over many competitors, which I’d like to explain using the example of filter cutoff: with a frequency range of ten octaves or 19980Hz, with a 7-bit resolution that would mean that each increment would either mean one semitone or 200Hz (both of which would lead to the well-known jumps when you sweep the frequency). This issue has been resolved here for good, and for all parameters at that.

Another noteworthy detail is that the display, while being rather small, can be configured to show the output oscilloscope-style, which actually really helps (and doesn’t seem to be faked, as it also works with signals from the audio input).



Looking at the VCOs again, you get 16′ to 2′ with the octave switches, fine tuning over a +/-1 octave range in 1 cent steps, a choice of one of triangle, square and saw, and a “shape” parameter.

Following the path many other manufacturers have taken as of late, “shape” allows you to change not only the square wave. For the square, it’s obviously the pulse width we’re changing.

For the triangle wave, the top of the waveform is increasingly folded back (or “warped”, as the effect is often called), until at full value of the shape parameter, you arrive at a (somewhat distorted) triangle wave with double frequency.

For the saw, finally, to stick with the “saw” metaphor, the shape changes from a Japanese saw to a M-tooth crosscut saw. This adds a lower harmonic, so unlike with the triangle, the saw appears to move down one octave when you change the shape.

Add to that the fact that you also get oscillator sync, cross modulation and a ring modulator, and you can already guess that the potential of the oscillator section is far beyond what you’d expect if you just had two oscillators each only capable of one of three standard waveforms at any given time.


After the flexible oscillator section, the filter is almost a letdown: lowpass with a choice between two and four poles, and that’s it. The filter can be set to track key (pitch) by 0, 50 or 100%, and the same can be done for the envelope amount’s control by velocity.

I started to examine the four-pole version first, and immediately found it to be a rather neutral affair, not dissimilar to the DFAM‘s filter. And similar to this, it also appears to be normalized, i.e. when you turn up resonance, rather than adding the peak on top until it starts to distort, the passband is strongly attenuated.

The resonance itself is, also as a consequence of this behaviour, a clear sine, and considering that 100% tracking is really that (due to the autocalibration routines of this synth), allows you to use it as an additional, higher-pitched oscillator – just like on the Minimoog, just more deterministically. Interestingly, if you play around with the combo of cutoff, resonance and VCO level for settings where the filter frequency is not a harmonic of the oscillators, you can get it into non-deterministic territory where the filter frequency may jump around between three pitches. This is not a limitation, as adjusting a parameter one step will resolve this without any other audible effect, but it’s perhaps the most “analogue” (as in: hard to tame) behaviour this synth offers.

Switching to two-pole mode, I immediately found the result more interesting compared to the four-pole mode, so if you want a patch to sound as if it was sent through an analogue filter (as opposed to just have the higher frequencies removed), two-pole is your way to go. Interestingly, the filter resonates just as nicely in two-pole mode as it does in four-pole mode. Which either means the filter is very well-designed and uses high-quality components, or the filter can only do four poles to begin with, and what’s labelled “two-pole” is in reality a bypass around the filter.

Either way, the choice between both configurations gives you a convincing alternative and greatly adds to the variety that is possible with this synth.


There’s two envelopes (called “Amp EG” and “EG”), which are straightforward ADSR affairs, and one LFO per voice. All of them are semi-hardwired, i.e. every modulation source has one or more destinations to which it is either directly connected, or connected via an attenuator, or where you can switch between them.

The Amp EG is hardwired to the VCA. All the times go from extremely short and snappy to around three seconds, so there’s the space between electronic clickly sounds and slowly evolving swells. The Minilogue initially suffered from an issue where the envelopes would always “snap”, even at longer times; fortunately, this has been fixed with a software update.

The EG is hardwired two filter frequency and oscillator 2 pitch, both via an attenuator. Optionally, it can affect either LFO rate or depth.

That LFO offers saw, square and triangle waveforms, and has three alternative destinations (pitch and shape of both oscillators together, and filter cutoff), where you can adjust the depth/modulation intensity. The rate goes from rather slow (maybe around 0.1Hz) well into the audible spectrum.

A noteworthy option are the LFO sync possibilities, allowing you to sync it to BPM, to key trigger and between voices.

The modulation section really is a weak area on this synth: while two ADSRs are (sadly) typically not exceeded for more affordable (analogue) synths, and the EG allows some individually controllable choices for its destination, the LFO is a letdown: only three waveforms and three alternative destinations are weak.


Now that I’m done lamenting the lack of modulation options, we arrive at the sequencer. It’s a 16-step sequencer (or rather “up to 16”, as the sequence length can be from one to 16 steps), with sequences stored per patch. What makes this interesting is that it can record up to four notes per step, plus four separate controllers for what is called “motion control”, which can be either played back stepped or smoothed. Recording is possible both in realtime and step mode. In addition, step length (from 1/16 to whole note), swing and default gate time can be adjusted.

Now this (especially the four motion control tracks) is interesting. The motion control tracks can control any parameter, and together with the smoothing option, this basically gives you a four-output arbitrary waveform generator with a cycle time between 0.06Hz and (for simple waveforms) around 30Hz. It’s there for deterministic sample-and-hold thingies, or for shape changes every 13th beat, or…you get the idea.

Voice Modes

Analogue polysynths had had different voice modes since the dawn of time, typcially “poly” and “mono” or “poly” and “unisono”. The Minilogue expands on that concept with no less than eight voice modes, each of which has a parameter adjustable with a dedicated knob.

Poly should be clear – up to four voices. The parameter here is invert, which will invert any chord (or combination of more than one key) you’re playing. While I at first didn’t think much of this, it works really great if you record movement of this parameter in the sequencer and then hold chords for a kind of chord-arpeggiator effect.

Duo and Unison has two or four voices stacked per key, turning it into a duo- or monophonic synth. The parameter here is detune.

Mono has only one oscillator playing at first. Upon increasing the parameter, you first add two -1oct sub oscillator, then further down the path, the fourth voice two octaves down. This allows for typical lead parts, where you can fatten up the sound for those key notes.

Chords plays chords (selected with the parameter) based on the note you play for those techno-style chord sequences. For added fun, try sequencing the parameter to switch among the more simple structures (e.g. 5th, sus2 and minor).

Delay is a software delay, where the additional voices are triggered after you play your mono part. The parameter is the delay time, which can for some reason only be set according to the synth’s tempo.

Arp is – you guessed it – an arpeggiator. You can latch it, and switch between different orders, including some poly ones.

Sidechain is a ducking effect: if you hold one (or more) notes and then play others on top of them, then the older notes will be attenuated until you release the newer additions. Adjusting the parameter here changes that from barely noticeable to an effect that appears like note-stealing.

Summing this up, Korg really have added some creative options here, some of which weren’t seen before, and some have gained an important option with the ability to automate the parameter.

Global Processing/Effects

For global processing, that is for all voices, there’s the combination of an analogue-style delay and a highpass, accessible with dedicated cutoff, time and feedback knobs and an output routing switch.

The configuration is the highpass fed by the voices’ output, followed by the delay, and its feedback path added before the highpass’ input. Output routing allows you to tap the output from before the feedback is added (essentially bypassing both highpass and delay), after the feedback is added (not sending the dry signal through the highpass) or after the highpass (sending the dry signal through the highpass as well).

The delay time goes up to roughly 350ms, i.e. one beat in the 180bpm range.

If you’re wondering where the “delay amount” knob is: there is none: due to the way this configuration is done, you essentially listen to the feedback, not the delay out, so that the feedback knob affects both feedback and delay level. Which means endless delays hidden in the background are not possible. But it’s possible to only have the highpass but not the delay.

On interesting aspect is that the “feedback greater unity” is only possible for a small range of settings of the highpass, with its knob roughly between 10 and 11 o’clock (or 300 to 340 for the parameter value), independently of the delay time.This is also the range where for a typical synth playing range, the highpass becomes clearly audible in the synth’s signal.

While both an additional highpass and a delay are a welcome addition in any synth, there’s two negative effects. One is that delay is, although it’s made to sound that way, not analogue. And the second one is that it’s too much made to sound that way, meaning it’s horribly noisy. There’s guys that can do such a delay without the noise but being analogue for a little over twenty bucks. We’d want that in this synth, too.

Under the Hood

Even with all the knobs and switches, there’s still some parameters under the hood, i.e. accessible via menus. They are grouped into program (for the current patch – patches are named “programmes”, or is it “programs”?), sequencer (for the current patch’s sequence) and global. The parameters are then grouped into pages, which you select with the eight sequencer buttons.

In the “Program” section, you get program name, slider assignment, portamento settings, LFO sync, program init and dump, and – neatly hidden there – how velocity affects the VCA.

For the “Sequence”, you get BPM, sequence length and step size, gate duration and swing, enabling and smoothing the four “motion” lanes, and clearing notes or one of the motion tracks.

Finally, “Global” holds things regarding MIDI/USB configuration, display brightness, if knobs jump or catch, and similar settings. Unfortunately, velocity curve and audio in on/off is set globally, not per patch.

Creating Sounds

In my age-old article about the “perfect synthesizer”, three of the requirements had to do with sound design: being able to create a wide array of sounds quickly and intuitively.

To sum it up: the Minilogue both motivated and enabled me to create a wide spectrum of great patches in a short amount of time.

First of all, the potential and the limitations are pretty easy to understand if you have at least some basic knowledge of synthesis. While of course having those hidden parameters available with a dedicated knob/switch would be beneficial, it’s not too limiting, as you’re typically not bound to adjust two of those hidden parameters against each other, as you sometimes do e.g. with the combo of filter cutoff and envelope amount to find the right spot. Having an arpeggiator and sequencer on board also helps to literally keep your hands free when applying the finishing touches to a sound. And last but not least, the aforementioned limitations regarding modulation options notwithstanding, this synth is capable of a really wide spectrum of sounds, from your staple analogue leads/basses/pads over convincing keyboard instrument implementations up to (if you want that) a wide array of analogue-style drum sounds.

So in that category, the Minilogue is easily best-in-class, and among the finest almost generally.

Performance Aspects

Let’s be frank: this area is the number one weakness of this synth.

Let’s start with what we have on the device, which is the three-octave slim keyboard and that slider thing.

The keyboard is slim-size. Which is bad. It also doesn’t offer aftertouch, nor does the synth respond to it via MIDI. You don’t even get a sustain jack (nor, you guessed it, does the synth respond to it via MIDI, or rather not the way you’d expect it).

Let’s start with the keyboard being slim-size. I know that Korg have done that for quite some time, and I can also understand why a two-octave keyboard (like on the Taktile-25) would not make that much sense for a polysynth. However, assuming that the width of the device was non-negotiable, you could also have fit two-and-a-half octaves (or 32 keys) into the same width, and have proper keys here.

Aftertouch and sustain have been playing mechanics in the synth world for ages, and even synths with only a very basic MIDI implementation (such as Arturia’s original Minibrute) offer it.

For ages (or rather, since Moog did the first keyboard synth back in the day), two wheels were the standard controls in addition to the keyboard, and almost every synth had at least those or something that took their function (although I was never happy with Roland’s odd thingie they introduced in the D-50 era). The Minilogue offers that odd slider, that sits in a strange place, travels in an odd angle, and while it can be assigned to any sound parameter, is simply one control less than the competition.

I perfectly understand that the mechanical design of the device might have been in part responsible for this, and that it’s not possible to have something in that free space between the keyboard and the synth’s controls, but it’s a shortcoming nonetheless.

This is to a degree made up for by the amount of knobs and switches: with 25 knobs and 14 switches, most parameters (and all the important ones) are available with a dedicated input devices. Add to that the fact that you can assign any single parameter (not just pitch) to the so-called slider, and that you can move to another menu parameter of your choice so you can control it with the program/value knob while playing, and this means that the typical “let the arpeggiator run and twist knobs” scenario is perfectly possible.

The situation is better with the interfaces to other devices, where we only face one minor issue.

First of all, having everything in the physical format you’d hope for (meaning the 5-pin diode connectors for MIDI, 1/4” for audio and 1/8” for sync) is greatly welcomed, as well as having both MIDI connectors at all. The Sync jacks, while using a somewhat Korg-specific signal format, work well in combination with my Eurorack thingies, so their availability is an asset not only for Volca users.

While an external wallwart power supply is not the first choice for most of us, it can be considered “expected” in this price category. Fortunately, the power specification follows the boss stompbox standard (9V DC, ground in center), the connector is of a different (albeit standard) type. Which means, if you’re using a shared PSU in your system, you will most probably able to use this for the Minilogue as well.

If you use the MIDI jacks for your control needs, the fact that MIDI Thru (neither with a dedicated jack nor as a Thru-via-Out SW functionality) is not available might be negative in some situations. Considering that the Minilogue is not made to be a controller keyboard, it will never be used at the beginning of a MIDI chain, and lack of Thru forces you to put it at the chain’s end (a position for which it might compete with similar devices). A Thru-over-out software functionality is rather easy to implement and should be on board in this scenario.

That isn’t a concern if you use USB for your MIDI demands, e.g. if you’re using it with a computer. The manual states that you need to install a proprietary driver for this – which already caused raised eyebrows because you might run the risk to no longer be able to use it some years down the road. (How many years? Incidentally, my first and also oldest Synthesizer was a Korg PE-1000 from 1976, so that’s a few years you might continue using a Korg).

Fortunately, when I connected the thing to a computer it installed a generic driver for it – so all is fine.

Integration Examples

As one example, I decided to give the Minilogue another oscillator via its audio in. That then meant that I was limited to monophonic/unisono patches. For the PWM Strings preset, I connected to a Doepfer A-190-2 MIDI interface, which controlled both an A-110-1 oscillator and an ADSR (used to modulate pulsewidth) via its bus. I then sent the oscillator’s output to a Dreadbox VCA (to obtain level control) and from there, went into the Minilogue’s Audio In.

Patch 1.jpg
The Minilogue as a three-oscillator (mono)synth

I switched on Audio In in the global menu, and could immediately confirm that this case of usage scenario works. Even when playing fast lines, the delay over the Minilogue’s MIDI out through the Doepfer MIDI interface to the VCO was not to be noticed.

For the next scenario, rather than having the external audio source augment a patch on the Minilogue, I opted to use its filter in a S&H-kinda configuration on an external source.

I set up a two-voice paraphonic sequence on the DFAM with each sequencer lane controlling one oscillator, with the output going to the Minilogue. On the Minilogue, I turned down all oscillators, programmed a 15-step sequence of notes with 100% gate length and had the filter track pitch 100%. Completing that with a generic amp envelope (full sustain, minimum times), all that was left was to sync the timing.

Patch 2.jpg
Filter processing sequence synced to the DFAM

I sent the DFAM’s trigger out to the Minilogue’s Sync In, configured that to 8th notes (which meant that for each incoming pulse it would advance two steps), and had a nice polymetric playground running, and that without having to take the detour via MIDI.

The Competition

As a keyboard synth with four analogue poly voices and a price tag around €550, the Minilogue doesn’t face any direct competitor.

Price-wise, Behringer’s Deepmind 6 (with 6 voices) comes closest at around 650 bucks. And while the Behringer clearly outperforms the Minilogue with regard to modulation options, I find the oscillator section of the Minilogue more interesting (and also original). Four-voice analogue keyboard synths of today include the DSI Mophox4 at almost double the price, and Elektron’s Analog Keys (at around €1200). If we were to include keyboard-less options in our search, we get Dreadbox’ Abyss, Vermona’s Perfourmer and Elektron’s Analog Four (again, all of them at least twice as expensive, sometimes by a margin) – and once we reach that price region, Korg has its own answer with the Prologue.


This synth is a combo of really great ideas and somewhat big shortcomings. You get two full-blown oscillators, with a great implementation of the “shape” concept and lots of options for modulating each other. And you get audio in and noise at the same time – but you can’t adjust the audio in’s level. You get two flexible envelopes – but only one LFO which is limited in its features and routing options. You get lots of knobs to twiddle – but the keyboard is that dreaded slim design. You get a highpass and a delay – but the delay is noisy and not even analogue. You get lots of connectivity – but no MIDI thru. But on top of that, you get a powerful step sequencer, and a price that is below that of many monophonic synth modules. And finally, you get a synth that is flexible, has its own character, and is really a joy to program.

In the end, if you’re in the market for an analogue polysynth that costs considerably less than a thousand bucks, you have to choose between this and the Behringer – and I guess in the end it then will be a decision based on the synths’ sound, which in my case is clearly a decision for the Minilogue.


2 thoughts on “Review: Korg Minilogue

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s