Testing the Korg Multi/Poly - A well-behaved little fellow

If there’s one family of synthesizers that today marks the digital music landscape for its depth and breadth of range, it’s the Wavestate/Opsix/Modwave and KingKORG Neo series. Each offers its own specific synthesis engine, injected into a modulation and effects platform that’s quite similar between models, all packaged in a compact, easy-to-carry keyboard. Some models started out as modules, large keyboards (everything else being equal), then upgraded to revision II, depending on age, benefiting from technological advances over time. Then, there were also limited series.

Today, everyone can choose a synth from the range according to their synthesis preferences, available space at home, existing equipment or destination (studio/stage). The latest model, the Multi/Poly promises easy access to complementary forms of synthesis in a compact multitimbral synth, with everything you need for complete control.

Compact synth: design, ergonomics and transportability

The Multi/Poly comes with a lightweight carrying case. It takes its shape from the KingKORG Neo, Modwave, Opsix and Wavestate. The plastic shell is covered by an aluminum front in the Mono/Poly blue and black colors, with the addition of a wooded blade for the sides, which enhances the presentation. Its weight is slightly heavier, at 3.5 kg (7.7 lbs.) on the scale. Its build quality is rather average, and the standard 37-key plastic keyboard is still not up to the asking price: noisy and soft, sensitive to initial velocity and release, but not to pressure. In our memory, however, it seems slightly better than the series’ predecessors.

The panel is covered with controls organized by section (pad, layers, editing/modes, effects, oscillators, filters, envelopes, LFOs, assignable buttons, program selection, sequencer steps or tracks). There are 22 rotary knobs (jump/relative modes), one encoder, 53 rectangular buttons, 16 backlit square buttons, two knobs and one modulation pad on a compact 57 × 32 cm surface. The organization is a little confusing, especially as many of the modules share the same controls: oscillators, filters, envelopes, LFOs, effects. In other words, just about everything you’ll ever need to fiddle with! The Shift key is frequently used as a modifier. While using it as such doubles the number of controls, it adds to the complexity.  

Dedicated keys allow you to select/activate/cut the four sound layers that a Performance —  the mode in which the synth is always in — contains. A fifth allows rotation between layers (forward, backward, alternate, random), giving interesting results reminiscent of the behavior of Mono/Poly oscillators. Alas, this rotation does not apply to the arpeggiator, which triggers all activated layers simultaneously — a disappointment. Four assignable buttons let you create direct modulations by assigning destinations to them; they transmit and receive Midi CCs.
As on the Modwave, you can use the pad to create two simultaneous modulations in Cartesian coordinates (X/Y) or polar coordinates (radius/angle). You can do this in two possible modes: either the classic manually-controlled pad, or a configurable 3D surface (hill, slope, chasm, trough, etc.), on which a virtual marble rolls and is thrown with the finger, progressing along the surface and bouncing off the walls. It’s as fun as ever!

A 128 × 64 dot monochrome OLED screen is used for accessing the many menus. Navigation is by means of the <>/Pages +/- keys, the notched encoder and the Enter key. The encoder is sensitive to rotation speed, and the Enter key can be used to increase its action, which is a great feature!

Most menu pages display all accessible parameters, and the scroll bar rarely gets used; certain parameters can be filtered if necessary. Other important ergonomic points: the screen displays movement sequences and envelope curves in real time; sounds can be selected by category and sorted alphabetically; values are often expressed in their true units (dB, semitones, seconds, Hz). There’s also a context-sensitive help function, a Compare function, and a random sound creation function, acting at different levels and on different parameters you can define. In utility mode, you can control active voices, voice theft, CPU load and temperature.

Comparable to its predecessors and as minimalist as ever, connectivity is integrated into the rear panel: stereo headphone socket (6.35mm/1/4" jack), pair of stereo outputs (2 balanced 6.35mm/1/4" TRS jacks, thank you!), foot pedal socket (6.35mm/1/4" TS jack), USB type B socket (MIDI, program data, firmware update, but no audio), MIDI DIN input/output, mains switch and terminal for external power supply (12 VDC positive center with block on the cheap end). It’s still not that generous.

Excellent sound: impressive versatility

The Multi/Poly starts up in around ten seconds, which is handy for live use. The synth comes with over 300 multitimbral Performances and more than 500 ready-to-use programs. The memory is managed in the form of a database; Korg doesn’t publish it’s capacity, but it’s likely to be several GB, with several thousand locations for the various elements (Performances, programs, favorites lists, sequence lines, temperaments, effects). Programs and associated items are automatically saved within Performances, so it’s instantaneous and not headache-inducing.

Separate backups can be made, although you risk their destinations being over the place. Fortunately, the database can then be tidied up, notably with the supplied editor/librarian. For the stage, you can create lists of favorites pointing to 64 Performances, quickly accessible with the 16-key row. The limit on the number of lists is not published either, but there’s plenty of available space. 
The audio quality is excellent, the levels are high, and there’s no background noise. We also appreciate the symmetrical jacks. Still, with a scene orientation, the smooth transition between different Multi-Timbral Performances (MTP) is most welcome. Sometimes, you can launch one sequence, hold it, and launch a second without interruption (provided you don’t exceed multitimbrality or polyphony, of course). The sound quality is excellent, as you might expect. It features what been Korg’s signature since the mid-90s: an ample, wide, deep tone, occupying the entire sound spectrum and stereo space without being intrusive (as long as you keep ensemble effects and reverb to a minimum). Even at high frequencies, there’s virtually no aliasing.

There are many areas of sonic interest, from ethereal textures to analog brass, big synth basses, sharp electric pianos, haunting jazz organs, cutting metal percussion, evolving wavetables, hallucinogenic West Coast ambiences and spectacular special effects, all realized in pure synthesis. Multitimbrality and a four-layer sequencer allow for elaborate rhythms. The Multi/Poly truly amazes with its sonic versatility, revisiting different eras of synthesis, while also being capable of producing contemporary sounds.

Multi_Poly_1audio 01 Wake up

00:0001:34

Link

Code

• Multi_Poly_1audio 01 Wake up01:34
• Multi_Poly_1audio 02 Gorgeous pad00:54
• Multi_Poly_1audio 03 So string00:47
• Multi_Poly_1audio 04 Acid combo01:06
• Multi_Poly_1audio 05 Toto brass00:24
• Multi_Poly_1audio 06 Digital motion01:00
• Multi_Poly_1audio 07 Multi tech01:18
• Multi_Poly_1audio 08 Machinery layer00:38
• Multi_Poly_1audio 09 Pad max01:25
• Multi_Poly_1audio 10 Dodo pad00:41

 WAV

Global Organization: a well-thought-out architecture

The Multi/Poly is a four-layer, 60-voice (maximum) polyphonic analog modeling and digital synthesis synth. It’s voice allocation is dynamic, but limits can be set per layer (1 to 8 voices). The sounds are organized by Performance, each containing four program layers (ABCD) and two global effects (reverb, EQ).

Each layer includes synthesis modules, modulations, three effects and a motion sequence. The arpeggiator is global. For each layer, you can import a program and then adjust volume, tessitura (texture) and velocity (with fade up and fade down for both). This is done conveniently via the menu or keyboard. You can transpose by octave/half-tone/hundredth, define the type of trigger (on key press or release — nice!) and force key hold.

Next comes voice assignment: mono with play priority + unison (number of voices, detune, stereo thickness and width) or poly (stacking or retriggering of the same note, useful for percussion) with dynamic allocation or voice reserve. We then move on to the MIDI receive channel (global or 1–16), hold pedal reception and MIDI CC reception. All this is quick to edit, thanks to the clear organization of menu pages. At the top level of a Performance, you can adjust volume, transposition, tempo and keyboard temperament (several presets with import capability via the supplied editor/librarian). In the “Extreme” position, the oversampling function, available per program, enhances quality at the expense of polyphony. In practice, it’s especially useful for certain high-frequency modulations.

.

Six sources: oscillators and sound generators in detail

Each synthesized voice consists of four oscillators, a noise generator, a ring modulator, two filters, an amp, four envelopes, five LFOs, three keyboard tracking generators, a modulation matrix and a motion sequence — phew! Each oscillator can operate in three modes: Classic, Digital or Wave Shaper. Classic oscillator mode models analog oscillators: sawtooth, variable pulse, triangle, sawtooth/pulse mixes, sawtooth triangle, pulse/triangle, square/triangle, detunable double sawtooth with different phases and Minimoog-style trapezoidal wave. Mixing/untuning/phasing of complex waves (depending on the case) can be modified and modulated, notably via a directly assigned LFO (but not only; see modulation matrix later).

Digital oscillator mode lets you work with tables of 64 waves (each composed of 2048 32-bit floating-point samples) by scanning the playback position in real-time (via a source to be defined, such as an envelope or an LFO), modifying the harmonic content of the table (filtering/amplification/static quantization of harmonic content according to about thirty presets) and applying morphing in real-time (compression, expansion, inversion, mirror effect, assignable — among other things — to an LFO). You can even set the fade time between consecutive waves from the same table. In short, lots of fun! The synth comes with over 200 ready-to-use tables, but you can import others via the supplied editor/librarian.

The third oscillator mode is the Wave Shaper, especially for fans of West Coast synthesis. The principle is to deform a sine or triangle wave with a more or less pronounced gain and offset. There are many different types of wave shapers: triangle, square, sine, pickups, saturators, additive series, multipliers, FM, tubes, diodes, fuzz, Berkeley (a model of Don Buchla’s original Wave Folder) and 60 models based on the Korg 01/W, a Workstation from the early 90s, successor to the M1, which was already capable of distorting waves and multisamples in the audio domain.

Other oscillator settings are essentially pitch-related: octave, semitone, hundredth, keyboard tracking slope, pitch envelope action (bipolar), portamento (trigger type, time/speed, duration) and pitch modulation (common) by LFO.
In addition to the four oscillators, there’s a noise generator with adjustable density and color (from wind effects to waves, and a ring modulator (with choice of carrier and modulator among the four oscillators + noise injection dosage).

The oscillators can then be synchronized in single mode (1 on 2–3–4) or double mode (1 on 2 and 3 on 4). They can also be cross-modulated (in frequency) with the same possibilities and fine modulation dosage. Noise can also be added to modulations. All this means that, even before using the filters, you can enjoy lots of sounds. Adding to the variety is that everything is available per oscillator, per voice, in four simultaneous programs. It’s a pity that the direct front-panel controls are limited in number and shared.

Pair of filters: modeling and sound diversity

The six sound sources are carefully mixed before attacking the two filters: volume, panning, polarity and balance to the A/B filters. All this can be done in mono or stereo, with stereo obviously consuming more polyphony since it is preserved throughout the entire chain filters => saturators => amp. Filters can be routed in series or in parallel, increasing the possibilities with independent balance control for each sound source. Each can be deactivated to reduce polyphony consumption.

There are 17 models for each filter, the most emblematic of which can be selected directly from the front panel. Models include Mono/Poly (4-pole low-pass resonant), Minimoog (4-pole low-pass resonant in transistor ladder), MS-20 (2-pole low-pass or ultra-high-pass resonant), Prophet-5 (4-pole low-pass resonant), SEM (2-pole medium-resonant with continuous state variables). Band-pass and band-rejection modes complete the picture.

Modeling is one of Muli/Poly’s strengths (along with its oscillators). You can adjust the input level to the filter, the cutoff frequency (alas, by semitone, which generates audible steps when the resonance is high), the resonance (with self-oscillation more or less soft or shrill depending on the model, except for the SEM mode) and the output level (compensation allowing you to adjust level differences between filters in parallel mode). In some modes, a setting lets you control the low-frequency behavior of the filter when the resonance is pushed (restricted, as on a Minimoog, or round, as on a Prophet-5).

There’s also a Multi mode where you can switch progressively between two types of filtering, choose preset combinations or manually adjust the balance between HP/HP/BP modes (as on an Andromeda). Powerful! Each filter has its own envelope (with velocity action), its own LFO and a keyboard follower to modulate its frequency. The latter is very comprehensive, with three separation points and four slopes.

You can also inject a frequency modulation audio source (1/2/3/4 oscillators, ring modulator and noise), which is an excellent feature!
At the filter output is the amplification section, which includes input overdrive amount (with anti-aliasing), volume, direct modulation by an LFO, velocity and keyboard follower (multi-segment generator comparable to those on filters, with limitation to twice the gain). There’s also the dedicated envelope (without direct modulation amount metering, except when using the modulation matrix).

Last but not least, there’s panning — stereo position and random modulation — with the rest done via the matrix. All this is as complete as ever. Voices can be subjected to more or less pronounced drift simulation. Multi/Poly even models variations between virtual voice cards. The affected modules (global pitch, oscillators, filters, envelopes, LFOs) can be dosed according to a voice card configuration chosen at the global level (without further details).

.

Modulations galore: advanced creativity and control

As for modulations, each voice has five LFOs, four DAHDSR envelopes, a modulation matrix, four modulation processors and three keyboard followers (already mentioned for filters and amp, so we won’t go into that again). LFOs are pre-assigned to pitch, wave modulation/PWM of each oscillator, filters and volume. Their speed can be tempo-synchronized or oscillate freely between 0.001 and 32 Hz.

There are 19 waveforms of varying complexity (classical, vibrato, exponential, discretized, guitar bend, random). Phase, vertical offset, input fade, waveform curve (rounded) and cycle mode (free or forced) can also be set. The four envelopes are DAHDSR (with additional release-level adjustment at the end of the segment). They are pre-assigned to pitch, filters and volume. Sustain is bipolar for pitch and filter envelopes. Time segments range from 0.001 to 90 seconds. Their curve can be modified continuously, between linear and exponential/log. The cycle can be retriggered by an assignable source or a note.

Let’s move on to the modulation matrix. The maximum number of routings is unknown, but we’re guessing several dozen or even hundreds, in addition to the pre-assigned modulations. Each routing comprises a main source, a bipolar quantity, a lateral source (modulation modulation) and a destination.

Assigning sources and destinations is intuitive, via direct commands, physical controllers (including keyboard velocity), MIDI CCs or menus (as a last resort, when the parameters to be assigned are hidden there). Sources include physical controllers (keyboard, knobs, pad, sustain pedal), generators (envelopes, LFOs) and MIDI CCs.

On the destination side, almost everything is assignable, including effect parameters and motion sequence lines. Specific menu pages allow you to view and edit routings and modulation quantities. As the list can be long, a function lets you filter modulations by type. We end this grand tour with the modulation processors. These can alternate, shift, quantize, spread, curve, smooth or sum one or more modulations. There are four per voice, two per sound layer and two per Performance. They are differentiated according to the type of destination they modulate. All in all, a very nice modulation section…

Full effects: A powerful and dynamic effects section

The Multi/Poly’s effects structure consists of three serial effects per layer and two global Performance effects, making 14 effects blocks in total. Beyond the number, the functionalities are identical to those of previous synths in the series, so we’ll indulge in a bit of copy-and-paste in this part of the test. Each layer has a Pre-FX, a Mod-FX and a Delay. For each, you choose the algorithm and, if you wish, one of the associated presets, then adjust the parameters. You can adjust the various levels (input, balance, output) and three parameters assigned to dedicated knobs.

Pre-FX algorithms include a decimator (lo-fi digital effect), EQs, compressors, ring modulator, tremolo, Wave Shaper and vintage pedal simulator. Mod-FX algorithms include chorus, phaser, flanger, phaser and wah-wah. Some effects are models of famous vintage products with which Korg has deep experience. Finally, there’s the choice and setting of the delay: LCR, stereo, inverted or tape echo. Sound quality is excellent, and you can feel Korg’s long-standing mastery of effects.

Overall, a Performance features a reverb and a semi-parametric four-band EQ. You can send each sound layer to the master reverb independently, letting you season everything differently. There are two main reverb families, based on some forty editable presets, again with sound-level settings and three variable parameters. The reverb is impressive, and you can embellish sounds by using the different modes. We’d love to have more (a few plates in particular), however. There are fewer editable algorithms and parameters than on a workstation, but the choices made by Korg are judicious and offer a good power/ergonomics ratio. Because the effects parameters can be assigned via the modulation matrix, they’re dynamic. Great!

Movement sequences: A creative and innovative sequencer

The Multi/Poly takes the motion sequencer from the Modwave. Each voice is driven by a sequence organized into seven lines of 64 steps, each line managing different parameters: Timing, Pitch, Shape and four assignable motion lines. Each line has its own steps, loop point, repeat, playback direction (forward, reverse, alternate, random), note-increment mode and step-trigger probabilities.

An (eighth) master line defines the time or number of beats after which the entire sequence restarts. You can import any of the lines in the database (hundreds are already preloaded for each line type, as well as complete sequences!) or start from scratch.
Each key on the keyboard triggers a mono voice transposed along the root note, so you can play solos, chords or shifted notes. Notes in the pitch line are recorded in keyboard steps, with automatic incrementation. You can immediately correct the notes and move through the steps with the <> keys, a smart move!

The four motion lines are programmed in real time by moving one knob per line. Values are memorized for each step, for the duration of a complete loop, and then smoothed between steps if the user wishes, with a choice of curves. Steps store relative values, whether for the pitch line (deviation from the defined root note and the transposition played on the keyboard) or for each movement line (percentage in relation to the synthesis parameter stored in the program).

Editing is accomplished line by line, step by step. This is how you access the Timing (step duration or time division, swing factor, note/silence/pause type) and Shape (modulation envelope per step, assignable to volume, pitch line and four movement lines) lines. For the pitch line, you can define a global temperament and then modify note offset and semitone tuning in each step. Numerous settings can be modulated (loop points, probability, etc.), adding spice to sequences.

It’s still impossible to edit several steps of a line simultaneously, which is a shame. If you get lost in the menus, you can use the external editor, which gives you a global view of the sequence. User lines can be saved in the database, with no known memory size restrictions, but are also saved with the current Performance, which simplifies things – a good idea! Sequenced notes are not transmitted in MIDI; in fact, in spirit, they’re more pitch shifts than notes per se, but we don’t think one precluded the other.

Global arpeggiator: possibilities yet to be improved

The arpeggiator is global for all four layers of a Performance. It’s impossible to disable some or alternate them at each step, which differs from the oscillators of the venerable Mono/Poly. Improvements in this area are sorely needed!

The arpeggiator can be used in conjunction with the motion sequencer, in particular the step increment mode. Patterns include high, low, alternate (two types) and random. Notes can be repeated in the order they were played, or sorted by pattern. Resolution ranges from 1/32T to 1/4 note and swings from –100 to +100%. Notes can be arpeggiated over one to four octaves, with Gate time from 0 to 100%.

Arpeggios may or may not be synchronized with movement sequences. Latch mode lets you maintain the current arpeggio while removing your fingers from the keyboard. The usual question: does the arpeggiator transmit notes in MIDI? Well, it still doesn’t, so we’ve suggested to the manufacturer that they add this function. To be continued…

FAQ section

• What is the Korg Multi/Poly?

  • The Korg Multi/Poly is a compact synthesizer combining analog and digital synthesis technologies, ideal for musicians in the studio and on stage.

• What kinds of sounds can be created with the Multi/Poly?

  • This synthesizer offers immense sonic variety, from powerful basses to ethereal textures.

• Is the Multi/Poly suitable for beginners?

  • Although it offers advanced features, its intuitive interface makes it accessible to avid beginners.

• What connectivity is available on the Multi/Poly?

  • It includes balanced audio outputs, a USB socket for MIDI and a stereo headphone socket.

Technical features

• Dimensions: 57 × 32 cm
• Weight: 3.5 kg
• Keyboard: 37 keys sensitive to initial and release velocity but not to pressure
• Polyphony: Up to 60 dynamic voices
• Oscillators: four
• oscillators per voice with classical, digital and Wave Shaper modes
• Effects: 14 effects blocks, including reverb, delay and modulation
• Display: Monochrome OLED for clear, fluid navigation
• Connectics: Balanced audio outputs, MIDI DIN input/output, USB jack (MIDI only), stereo headphone jack