Korg Prologue

Prologue, a new hybrid analog synth from KORG.

  • Full-Size Keyboard, 49 or 61 keys
  • 8 or 16 voice options
  • Hybrid system
  • Bi-timbral
  • 2 VCOs + Digital OSC (Noise, VPM, FM and user uploaded)
  • 1 Analog Bipolar Lowpass Filter
  • 2 Envelope + 1 LFO
  • Ring-Mod, Cross-Mod and Sync
  • Distortion
  • Digital-Effekte (Modulation und Delay/Reverb)
  • Poly-, Mono-, Unison- and Chord-Modi
  • Arpeggiator
  • 500 patches
  • 1500 and 2000 Euro

Best thing about the Prologue: Developers can write their own digital oscillator and effects for it!! It has 16 user slots for both digital osc and effect section. Korg will release a developer kit/APIs later for it.

If the developer kit is flexible enough and the Prologue has enough processing power, somebody will port all the braids oscillators to the Prologue for sure.

A 16 voice braids synth with analog filter and some custom effects. :smiley:

Korg really kind of surprised me by opening this to third party software… though they have a history in this and probably it proved to be a good ROI thing for them :smiley:
But I can’t complain about it, since it is an enrichment for the musical community.


Finally :slight_smile:

They probably already put similar sounding models :slight_smile:


Pretty. Very pretty.

That low bass booster / compressor on the 16-voice looks interesting.

1 Like

This looks nice! I like the idea of the reprogrammable oscillators a lot.

okay this looks tasty as hell. korg is really doing good work.

Very curious as to what type of microprocessor the user-defined oscillators run on, and what language(s) the SDK will support for writing new oscillator types? But yeah, that’s a refreshing and laudable move on KORG’s part.

Probably an ARM given they used ARM on the Minilogue.

As for the API/SDK, it will be interesting to find out. You’d expect it to be C/C++ but they might try to make it “user friendly” and use something higher level that just calls heavily optimised C functions. Depends who they target the SDK at, I’d expect it to be more low level and for experienced programmers.

I wonder how they want to protect users from malware. I know that ARM chips from ST have a write protection that can be enabled. Any code running on the unit could actually enable this write protection while at the same time bringing the processor to a crash - making the synth unusable with no way to recover other than physically replacing the chip.

I think malware is probably not the best term? do you mean badly coded oscillators that may brick the unit?

I would imagine the software to install the osc code would defend against this.

I mean oscillator code that is written to brick a device on purpose. There’s always going to be someone who tries out some oscillator code they found on some website. For KORG repairing a bricked device is expensive and telling a user that it’s their fault is not very clever either.

Enabling the write protection is just a matter of writing a specific value to a specific memory location (AFAIK). Detecting this in the binary code before it is uploaded would be rather complicated, especially since memory access can be dependent on register values.

The same goes for detecting code that doesn’t work. There are a million ways to create an endless while() loop and all of them look different.

The only way to somewhat check for these kinds of malicious/bad code is to simulate it before uploading it to the synth. That is a significant undertaking because for a bad guy there are - again - a million ways to make the code look harmless to a simulator (e.g. by detecting if the code runs on a hardware unit and if not, simply bypassing the bad code)

Korg has developed technology to solve the halting problem.

My bet: you write your routine in C or C++ and compile on your machine (they might provide a little GUI that dynamically load the code, to listen and play with the results). And there’s a tool to upload it to their servers where it’s compiled into a binary for the target machine. User-created oscillator code is manually vetted before being published on their community platform.


Very likely. They might even go all “app store” and allow developers to charge for their osc and fx models.

You’ll need to write oscillators for it in Haskell, then?

1 Like

Someone finally did it. Porting Plaits OSCs to Prologue.



Given that the Prologue has only two timbre control parameters (SHAPE and shift+SHAPE), I’m curious how they handled the third dimension offered by all Plaits models!

I would love a mini diagram showing how SHAPE and shift+SHAPE are mapped to Plaits’ knobs, and the position of whichever Plaits’ knobs are left untouched.

1 Like

Prologue can have upto 6 additional parameters in the oscillator menu.

The code is extremely short: https://github.com/peterall/macro-oscillator2-prologue/blob/master/macro-oscillator2.cc

I just checked and he also uses the aux mix as parameter (nice!) and added a “LFO target” selection in the menus. I wrote the menu item names in brackets.

Pair of classic waveforms (va)

Harmonics = Menu Parameter 1 (detune)
Timbre = Shift + Shape
Morph = Shape
Mix = Menu Parameter 2 (sync)

Waveshaping oscillator (wsh)

Harmonics = Shift + Shape
Timbre = Shape
Morph = Menu Parameter 1 (asymmetry)
Mix = Menu Parameter 2 (curve)

FM Oscillator (fm)

Harmonics = Shift + Shape
Timbre = Shape
Morph = Menu Parameter 1 (feedback)
Mix = Menu Parameter 2 (sub)

Granular formant oscillator (grn)

Harmonics = Shape
Timbre = Shift + Shape
Morph = Menu Parameter 1 (width)
Mix = Menu Parameter 2 (filter)

Harmonic oscillator (add)

Harmonics = Menu Parameter 1 (bumps)
Timbre = Shape
Morph = Shift + Shape
Mix = Menu parameter 2 (hammond)

Wavetable oscillator (wta - wtf)

Harmonics = Menu parameter 1 (Interpolate)
Timbre = Shape
Morph = Shift + Shape
Mix = Menu Parameter 2 (Low-Fi)

1 Like

If the rumours regarding Minilogue DX are true, this port definitely makes me getting it. 4 voice poly synth with MI oscillators + Prologue filters + custom effects for 600-700 sounds lovely.