WHAT'S DIFFERENT ABOUT THE NF-1?

In our search for innovative sounds, we wanted to add new waveforms to the classic three: sawtooth, square/pulse and triangle, and we wanted to create new filter algorithms. We wanted to explore new sound creation possibilities in the oscillator and filter sections of the classic (virtual) analog synthesizer. And we think we succeeded.

WHAT'S DIFFERENT IN THE FILTER SECTION?

In the filter section we realized that real innovation lies not in meticulously recreating the ultimate sound of some legendary synth of the past. We leave that to other synth constructors, who are probably far better in this domain. These (virtual) analog legend synths truly are legendary, no doubt about that. But the future has to create new legends in order to stay interesting.

When creating digital filter algorithms, there is much more possible than only recreating resonant lowpass filters. Literally thousands of other combinations of peaks and holes in the frequency response curves of digital filters are possible. We soon found out that a decent and easy-to-use hands-on formant filter is a very rewarding sound sculpting tool. It took us some time, but we created a three-vowel morphing formant filter, with 10 preset 'vowels' or formant combinations to choose from, and the possibility to user-adapt these presets or create whole new formant combinations.

formantfilter.mp3 // An example of the formant filter acting on a sawtooth wave, and next mixed parallel to a resonant hipass, playing some chords.

Of course we also gave the Modor NF-1 a classic resonant filter. It can be used apart from, or in combination with the formant filter. We made a simple but decent 12dB/oct resonant filter selectable as lowpass, hipass, bandpass and bandstop. It's resonance goes up to self-oscillation, and the filter's exit can be driven into distortion.

filterex.mp3 // A little playing around with the lowpass filter, into self oscillation and adding some output drive.

WHAT'S DIFFERENT IN THE OSCILLATOR SECTION?

We also wanted to give our oscillators new 'waveforms', as a lot more of the character of a synth lies in the oscillator section than usually told. But we had to admit very soon that the classic three: sawtooth, square/pulse and triangle are not for no reason the ever returning classics. They are the basic building stones of a vast number of possible synthesizer sounds. Trying to make other wave'forms' is very unrewarding, and we tried a lot of them! Circular formed waves, sawish pulse forms, peaks, triangles, tilting squares, pentagonal or hexagonally morphing geometric stuff... They all start sounding like a mix of bad imitations of the classic three. Very innovative, but it just doesn't sound good.

We realized that there is a large open space in creating sounds that have a certain 'grain' or roughness in it. The most beautiful singing voices, or the most fascinating genuine instrument performances often contain a form of noisy speckled special 'character' in it. And that is where classic synthesizers fail.

 
Just adding a drop of white or pink noise is not sufficient to achieve this. Some synthesizers in the late '80s and early '90s made a certain approach to this by adding some recorded looping 'waves' of choirs and windblown things to add to your sound, but they are and they sound like nothing but static recordings. Probably also due to the impossibility of programming synths of that age, with their over-complex menu systems and very limited data controls, they just disappeared again a decade later.

We created three noise 'waveforms' that are not recordings of some noisy sounding stuff, but real-time sound creation algorithms. Here are examples of each of them. You can hear they don't sound too well on their own, but they are made to be layered with the other oscillators to create noisy sounds.

sonarnoise.mp3 // Sonar noise: This creates a resonant bandpass-filtered noise, of which the resonance parameter can be adjusted.

windnoise.mp3 // Wind blown noise: This makes a comb-filtered form of noise, with the possibility of further removing some of the lowest frequencies.

arcadenoise.mp3 // Arcade noise: This is a recreation of the noise creation algorithms of former arcade video games' sound chips, such as the SID or the SN76489.

We also added two FM-'waveforms', which are 2-operator blocks as found in the classic FM-synths. This enables you to add very typical FM-synthesis elements.

  1. FM creates a sinus wave (=carrier) which' frequency is modulated by a second sinus wave (=modulator) that you don't hear, but of which you hear the effect upon the first carrier wave.
  2. FBFM or 'feedback FM' creates the same thing but with the modulator wave first modulating itself, and than the carrier wave.

In both algorithms it is possible to set the carrier: modulator frequency ratio, using the FM CARRIER and FM MODULATOR buttons. These FM-waveforms are not really innovative, as FM-synthesis already existed before. But the combination of FM-elements in a typical Virtual Analog-setup proofs to be very interesting and expands the domains of both FM- and VA-style synthesis. And what's more: the classic FM-synths also proved to be incredibly tedious to program, with almost no sound editing controls on their front panels, and extensive menu systems. Which made them also disappear after a decade. This is implemented way easier in the Modor NF-1.

GlassyFM.mp3 // An example of a sawtooth-based pad, with a typical "glassy" FM-attack.

A last innovative element in the oscillator section is the additive oscillator. In this oscillator basic sine harmonics are added to each other to create waveforms with unusual harmonic combinations. You can choose their separation to be 1x or 2x the base frequency, creating waveforms containing all or only the odd harmonics such as in saw- or square-waveforms. But you can choose also 3x, 4x ... 16x the base frequency, with the additional possibility of choosing the first, second, ... eightest harmonic to start from. It gives you the possibility of enhancing or adding a certain harmonic or to make unusual combinations of harmonics.

twinkels.mp3 // In this example, you first hear an atmospherical pad playing some chords. After that we hear the same chords with the separate elements one by one. First some 'twinkling' by the additive oscillator, secondly an FBFM-oscillator and finally an overlay of wind noise. Not exactly a type of sound you'd expect from a virtual analog synth ...

A MODULATION MATRIX?

Can I modulate the volume with an LFO? Can I modulate the filter envelope amount by velocity? The stereo panning by a note's keyboard position? A random cutoff frequency? A noisy pitch attack? A ... ?

Yes of course! There are a number of 'hardwired' modulation possibilities for the most popular modulations, such as the filter LFO amount, or pulse width modulation on the square and saw waveforms (PWM). But next to that, you can attach up to 7 virtual modulation wires between 19 modulation sources and 86 modulation destinations. These sources are:

  • 3 LFO's
  • S&H and LS&H
  • 4 Envelopes
  • Velocity and Aftertouch
  • Pitchbend
  • Modwheel(CC#01)
  • BreathCtrl (CC#02)
  • Controller 3 (CC#03)
  • An expression pedal (CC#04)
  • Keyboard position
  • A random value chosen at the beginning of each note
  • Slide, used in Midi Polyphonic Expression (MPE) keyboards as made by Roli and Roger Linn Design.

The modulation wire's possible destinations are:

  • Every oscillator's pitch, level and MOD parameter
  • The carrier & modulator frequency ratios of the FM waverforms
  • The filter's cutoff, resonance and drive
  • The formant filter's FORMANT and MIX parameters
  • The speed of every LFO
  • Each of the 7 parameters of all 4 envelopes
  • Every parameter in the Chorus/Flanger and Delay effect sections

But also (since OS009) :

  • Every modulation wire's modulation amount... !
 

So you can also modulate the intensity of other modulations, the hardwired ones and the 7 freely assignable wires.

Sounds too good to be true? What's missing here, where's the hidden drawback? Well, we'll be honest: there are no audio-rate modulations. You can't cross-modulate the oscillator's pitches, and there's no audio-rate filter frequency modulation. Audio rate modulation are very difficult to achieve in digital synthesis for some complex theoretical reasons. Some other digital synth manufacturers do offer some audio rate modulation, mainly pitch X-mod. We chose to keep our valuable CPU time for other tasks, such as our special Noisy waveforms and our Formant filter. Hence the NF-1...

 

IS THE NF-1 FULLY MIDI-RESPONSIVE?

Yes, it is. We made the NF-1 respond and send out a maximum of possible midi messages.

  • Control changes: All the 'continuous' controls (turning knobs) of the Modor NF-1 do send out a midi control change message if they are turned. Also some of the 'discontinuous' controls (with pushing knobs) do the same. This can be enabled or disabled in the System Settings-menu. The Modor NF-1 of course also reacts to incoming control changes, which can eventually also be disabled.
  • Program changes: The Modor NF-1 reacts when a program change message comes in. This option can be disabled.
  • Pitch bend, aftertouch, modwheel, breath controller and CC#03: These messages are received and their values can be used as modulation sources to alter any sound creation parameter in the modulation matrix.
  • Sysex messages: The contents of the memory of the NF-1 can be backed up by sending out midi sysex messages. You can choose to dump only the active patch and/or the complete memory contents. The NF-1 can also receive future firmware updates via Midi Sysex. Sysex reception can also be disabled.

WHAT'S INSIDE THE BOX?

The Modor NF-1 always comes with a 230V to 9V adapter, a printed manual, removable massive wooden side panels and removable rack ears. Midi and audio cables are not included.

WHAT CHIPS ARE USED IN THE NF-1?

Want to know what's inside? The NF-1 contains two processors:

  • one DSP56725 dual core DSP-processor at 250MHz dedicated to sound synthesis,
  • and one 16MHz PIC18F67K22-microcontroller to collect data from the knobs & buttons, control the LCD-screen & leds and manage midi communications.

Furthermore there's a DA-converter, Cirrus Logic's CS4334 and a few flash memorychips 24LC256.

The DSP56725 is a member of the famous DSP56-family of DSP processors by Freescale (Motorola) that populated the Access Virus, some of Waldorf's synths and Clavia's Nord Leads. But that doesn't mean that the NF-1 sounds identical to these synths. The sound of a digital synth is defined by the program inside the chip, not by the chip itself.

WHERE DO I FIND THESE ANGLED MIDI CABLES?

 

The Modor NF-1 can be built into a standard 19" rack using the (always included) rack ears. This means the connectors for power, audio and MIDI are at the top side. The lower part of the backside of the Modor NF-1 is recessed such that you don't lose precious rackspace above the NF-1 in this case.

But that space is not big enough to fit any cable. Angled audio cables are rather easy to find, but you also need MIDI cables with a 90° angle at one end. These cables are manufactured for example by Manikin Electronics: http://www.manikin-electronic.com/en/products_accessories.html.

One place to buy them is here: http://www.schneidersladen.de/en/manikin-midi-kabel-2m-gewinkelt.html

 

           
all rights reserved © 2015 Modor Music | website by MIST