Review – Xen_FMTS2, a FREE Softsynth That Opens the Way to Wild Explorations of Tunings and Timbres

Xem-FMTS2 is a free softsynth which has been designed specifically for those who want to explore the farther realms of ultra-accurate tuning, and the relationship of tuning to timbre.

by Warren Burt, July 2014


Xem-FMTS2 is a free softsynth from Jacky Ligon’s, which has been designed specifically for those who want to explore the farther realms of ultra-accurate tuning, and the relationship of tuning to timbre.  Built on the Synth-Edit platform, it’s an amazingly flexible and efficient (fairly light on the CPU) synth which allows one to do many things that other synths can’t do. 

The unusual things this synth can do are: it can accept tuning files in the MTS standard, it can allow partials (harmonics) to be tuned to be in any existing scale, and it has a large number of algorithms that allow the different operators (there are four of them) to frequency modulate, ring modulate, and mix with each other.

Some explanation for each of these might be in order.

Microtonality:  This usually means tuning that uses anything but the equal division of the octave into 12 tones that is standard in Western classical and popular music today.  As a matter of fact, most of the rest of the world does not use the division of the octave into 12 equal tones, and Western music didn’t adopt that tuning as standard until the mid-19th century.  (Myth: Johann Sebastian Bach wrote “The Well-Tempered Clavier” to demonstrate the modern 12 tone equal temperament.  In fact, there were several scales in Bach’s time called “Well Temperaments” which had each semitone on the keyboard tuned to a different size, and which allowed you to play in any major or minor key and they would all sound good – but they would all sound different.  Those were the scales that Bach wrote his collection to demonstrate.)  Indian music, Thai music, Chinese music, Indonesian music, African music, European Medieval music all used scales different than the one we use today.  The world of microtonality is both incredibly ancient and filled with potential for the future.  And to have “microtonal” music does not imply any musical style.  Today, there are microtonal folk groups, microtonal rock groups, microtonal classical ensembles (hearing Bach, Mozart or Beethoven in the meantone tunings they wrote for is revelatory), and even microtonal country and western bands (JC and the Microtones album “Cow People” is a C&W classic!).  Using microtunings simply means tuning differently than is mostly done today, and says very little about the musical style used.

MIDI TUNING STANDARD: The MIDI specification specifies that every semitone on the MIDI keyboard should be able to be divided into 16384 parts, with the division of the octave being 196608 equal divisions to the octave. (14 bit numbers are used for this.)  How synthesizers implement this is another matter.  Most softsynths that allow microtuning use the Scala .scl or .tun formats.  These use 7 bit numbers to divide the semitone into 128 parts for a tuning of 1536 equal divisions to the octave.  For most purposes this is quite adequate, but there are a number of people, such as the experimental composers Harry Partch (1901-1974) and La Monte Young (1935- ) whose work has demanded greater precision.  Hence, in the early 1980s, composers Robert Rich and Carter Scholz created the MIDI Tuning Standard, which became part of the MIDI specification.  Finally, in the past decade, synthesizers have appeared that use all 14 bits of the MIDI Tuning Standard to give very precise tuning. 

The Matching of Tuning and Timbre: As people began to use synthesizers to explore different tunings, it was quickly noticed that some scales sounded horribly dissonant, while others had very consonant intervals.  It was then noticed that if the timbre of the note were changed, the dissonances in the intervals of a particular scale seemed to smooth out quite nicely.  An example would be 13 tone equal temperament – dividing the octave into 13 equal parts instead of 12.  In this scale, there is no interval that even comes close to a normal Perfect 5th, or any other harmonic series interval, and the intervals of 13 tone tuning, when played with a piano timbre (or any other timbre made of harmonic series intervals, like a guitar, clarinet or trumpet) sound really awful.  However, William Sethares, electrical engineer and composer, noticed that if the harmonics of the piano sound were retuned to be in 13 tone equal temperament as well (producing a more gong-like timbre than the normal piano sound), the out-of-tune Perfect 5th of 13 tone tuning sounded quite mellow.  His work is summed up in his book “Tuning, Timbre, Spectrum, Scale,” which was one of the most exciting music theory books of the 1990s.  Implementing Sethares’s ideas usually involved the use of programming languages such as CSound.  Very few synthesizers allowed one to tune partials and tuning separately.

Until now.  Jacky Ligon, at, has developed several free softsynths designed to use the full accuracy of the MIDI Tuning Standard, and now, with his latest, Xen-FMTS2, he’s implemented a way to tune the partials of a sound into any tuning system whatever.  So experimentation with fantastically accurate tuning of both the scale a note is playing in, and the tuning of the harmonics of that note is now easily possible.

But wait, as the TV ads say, there’s more.  The central tone generator of XEN-FMTS2 consists of four separate oscillators (known in FM parlance as Operators).  Extrapolating from Sethares work, Ligon reckoned that if partials were tuned to match (or mis-match) the scale being used, then those partials would make harmonically interesting timbres if they were to modulate each other.  So with this synth, the operators can Ring Modulate, Frequency Modulate or simply be mixed with each other.  57 different algorithms are available, along with 11 different waveforms for each operator.

Here’s a listing of the algorithms:

All  of  the  Algorithms  are  displayed  in  a  mathematical  equation  style.  Functions within brackets are performed first, as is usual, and from left to right in other cases. 

‘ * ‘ Indicates the Ring-Mod function, i.e., (A*B) Wave_A Ring-modulates Wave_B

‘ > ‘ Indicates the FM function, i.e., (A>B) Wave_A FM-modulates Wave_B

‘ + ’  Indicates simple mixing.

 Where “A>B>C” = Wave_A FM-modulates Wave_B, the resulting Wave then FM-modulates Wave_C

The Algorithm List:

01     A+B+C+D

02     (A+B+C)*D

03     (A+B+C)>D

04     (A+B)>C+D

05     ((A+B)>C)*D

06     (A+B)>C>D

07     (A+B)*C+D

08     (A+B)*C*D

09     ((A+B)*C)>D

10     (A>B)+C+D

11     ((A>B)+C)*D

12     ((A>B)+C)>D

13     (A>B)>C+D

14     (A>B>C)*D

15     A>B>C>D

16     (A>B)*C+D

17     ((A>B)*C)>D

18     (A>B)*C*D

19     (A*B)+C+D

20     ((A*B)+C)*D

21     ((A*B)+C)>D

22     ((A*B)>C)+D

23     ((A*B)>C)*D

24     (A*B)>C>D

25     A*B*C+D

26     (A*B*C)>D

27     A*B*C*D

28     (A+B)*(C+D)

29     (A>B)+(C>D)

30     (A*B)+(C*D)

31     (A*B)+(C>D)

32     (A>B)*(C>D)

33     (A*B)*(C>D)

34     (A>B)+(A>C)+D

35     (A>B)+(A>C)+(A>D)

36     (A>B)+(A*C)+D

37     (A>B)+(A*C)+(A>D)

38     (A>B)+(A*C)+(A*D)

39     (A*B)+(A*C)+D

40     (A*B)+(A*C)+(A*D)

41     (A+B)+(A>C)+D

42     (A+B)+(A*C)+D

43     (A+B)+(A>C)+(A>D)

44     (A+B)+(A*C)+(A*D)

45     (A+B)+(A>C)+(A*D)

46     ((A+B)>C)+((A+B)+D)

47     ((A+B)>C)+((A+B)>D)

48     ((A+B)>C)+((A+B)*D)

49     ((A+B)*C)+((A+B)*D)

50     ((A>B)>C)+((A>B)+D)

51     ((A>B)>C)+((A>B)>D)

52     ((A>B)>C)+((A>B)*D)

53     ((A>B)*C)+((A>B)*D)

54     ((A*B)>C)+((A*B)+D)

55     ((A*B)>C)+((A*B)>D)

56     ((A*B)>C)+((A*B)*D)

57     ((A*B)*C)+((A*B)*D

Of course, if you turn any operator’s amplitude to 0, you’ll have even more types of algorithms to play with.  You can spend hours exploring the effects of the different algorithms.  I’ve been enjoying hearing whole families of timbres which are produced by different partials, different tunings, and different algorithms (for example 8 tone tuning with 9 tone partials and then listening to all the different algorithms in sequence).  I did a study piece recently with the tuning in 13 tone equal temperament, and a ring modulation algorithm (A*B)+(C*D).  Each section had different partials from 8 tone equal temperament to 17 tone equal temperament.  The difference between the timbres was not trivial.  Some sections (all sections used the same pitch material) sounded very dissonant, others, pleasingly consonant.

Not a Meat and Potatoes Softsynth

As you can see, this is a tool for the serious musical explorer.  Although the machine comes with a set of 115 presets, and is capable of producing your average bass, keyboard, percussive and pad sounds, its more interesting use comes in exploring the many possibilities of its unique architecture.  The 32-bit VSTi plugin synth includes 55 microtunings and 56 partials files to get you started, and there are a number of sources, such as which will give you more tunings to play with.  And for those who want the ultimate roll-your-own resource for exploring and developing microtonal scales, there is Manuel op de Coul’s free program Scala ( which will allow you to make scale files in every available tuning format, including MTS, .scl, and .tun.

The synth has only one master sound generator, so it can only play one timbre at a time, but it’s 12-note polyphonic.  To control this synth, there are seven modulation generators.  That’s Ligon’s term for a combination LFO-ADSR module. Each Operator has its own modulation generator, as well as one for the filter, one for pitch, and one for overall amplitude control.  Each modulation generator contains a separate LFO and ADSR, and the output of these can be mixed.  Each ADSR has a number of shapes – linear, exponential, inverse exponential, square root, etc., and the LFOs have 24 wave shapes and the ability to function as both LFOs and Audio Rate Oscillators (using them as Audio Rate Oscillators adds further complex possibilities to the already rich array of modulations available).  The maximum time on the Attack and Release slopes is currently 10 seconds.  Ligon says he is working on a more complex Modulation Generator with longer slope times to be included in the next version.

The Filter is a combination filter and saturation unit.  There are 6 different kinds of filters with two filters per module (different kinds of Lo, Hi, Band, and All-Pass filters and a Band Reject filter), and 20 different kinds of saturation, and the module can function as a filter, a saturator, both, or in bypass mode.  There is also another Low Pass Filter (called “warm”) on the output stage, as well as Chorus and Ensemble effects. 

The most interesting part of the synth, though, is the 4-operator FM-RM Oscillator.  This has controls for selecting the tuning the synth plays in, as well as a control for selecting a “partial” file.  This is a list of pitches that the operators will be tuned to, based on the scale selected.  Below that are a number of controls designed to tune the operators to a particular member of the list of partials.  These can be chosen in a number of different ways, called “Quantized Sliders,” “Iso-Index,” “Ratio Sliders,” and “Random Index Sliders.”  These can be a bit tricky at first, but with a little work allow you to choose families of partials for your operators in very useful ways.  Consulting the manual and experimenting here will yield great rewards.  Once you’ve developed your sets of partials, you can then proceed to the 57 different algorithms to experiment with your tuned partials modulating or mixing with each other.   I spent quite a while just exploring the possibilities of tuning the partials without any modulation, treating it as a four oscillator additive synthesizer, before I began to experiment with the FM and RM possibilities.  Just in this additive mode, I heard a lot of timbres I hadn’t heard before and yes, was able to tame the fierce “pseudo-fifths” of 13-tone and 8-tone equal temperaments.

MIDI automation is extensive here.  There are 321 different modulation targets.  Just about every function of the synthesizer can be externally controlled by MIDI.  Ligon said that he wanted to be as extensive as possible with the modulation possibilities for those composers who wanted to plunge in deeply and make the synth as dynamically controlled as possible.

The manual is extensive and well worth reading.  Before I explore a new softsynth, I usually spend a while just reading the manual, trying to intuit the compositional possibilities of the new machine.  As I was reading the manual, my mind kept boggling at the implications XEN-FMTS2 had for composition and tuning exploration.  As I started playing with it, I was delighted – things were possible now that I’d wanted to do for years.  In the few months that I’ve been playing with it, I’ve barely scratched the surface.  This is a deep and powerful tool for sound exploration, one that is fairly easy to learn, and which is pretty light on the CPU, and even lighter on your wallet.  Highly recommended.  And while you’re at the website, be sure to check out their other two free softsynths, Ivor, a subtractive softsynth with the same MTS tuning accuracy, (Ivor is named after Ivor Darreg, 1917-1994, the microtonal music pioneer), and XenFont, an SF2 sound-font based synth which also uses MTS tuning files and has a wide range of modulation possibilities.

Download here: Windows only, 32-bit, VSTi, Free


You may also be interested in

Browse SB articles

SoundBytes mailing list


Welcome to SoundBytes Magazine, a free online magazine devoted to the subject of computer sound and music production.


If you share these interests, you’ve come to the right place for gear reviews, developer interviews, tips and techniques and other music related articles. But first and foremost, SoundBytes is about “gear” in the form of music and audio processing software. .


We hope you'll enjoy reading what you find here and visit this site on a regular basis.

Hit Counter provided by technology news