Almost exactly 14 years after I visited Auschwitz and Auschwitz-Birkenau, I visited Weimar-Buchenwald. A relative was there four years ago and coined this visit as »brutal erschütternd«. There are no other two words that can better express what I perceived during my visit. Incredible and brutally shocking.
What’s even more shocking is that contrast during the visit. I was there since I was visiting several pipe organ concerts, mainly Johann Sebastian Bach being played, one of the greatest composers ever. Furthermore, Weimar was the residence of Franz Liszt, Goethe and Schiller for several years. So, Weimar and the concentration camp Buchenwald are both nearby examples of what humans are capable of – extreme barbarism on the one, extreme cultural attainments on the other hand side.
So I wonder whether we can learn from Auschwitz-Birkenau and Weimar-Buchenwald for recent developments in 2022 and 2023? Yes, I’m pretty sure we can. Barbarism ends rather sooner than later. Culture luckily survives for a very long time.
Zwar war es zu Silvester recht mild geworden, in St. Stephan war es aber noch immer recht frisch. Einige einheimische Besucher hatten sich daher auch mit Decken auf den Weg gemacht.
Ich war schwer beeindruckt vom Klang der fast 400 Jahre alten größten Renaissance-Orgel der Welt selbst als auch davon, wie Olga Minkina sie gespielt hat. Die Auswahl der Werke (etliche davon erst 300 Jahre nach dem Instrument entstanden), die Registrierung (nichts zu dick aufgetragen, sehr transparent und abwechslungsreich), die Interpretation (viele Wiederholungen als Echo gespielt), da machte das Zuhören schlicht und ergreifend extrem viel Spaß. Hier das Programm:
Heinrich Scheidemann (1596-1663) – Magnificat III. Toni (4 Verse)
Pfarrer Jürgen Weinert – Gedanken zum Jahreswechsel
Dietrich Buxtehude (1637-1707) – Wie schön leuchtet der Morgenstern (BuxWV 223)
Hermann Schroeder (1904-1984) – 12 Choräle für die Weihnachtszeit
Auf, ihr Hirten
Singen wir mit Fröhlichkeit
Lasst uns das Kindlein wiegen
John Rutter (*1945) – Toccata in Seven
So ein beeindruckendes Konzert hatte ich am letzten Tag des Jahres jedenfalls nicht erwartet. Dafür allen Beteiligten ganz herzlich gedankt.
Unser heutiges Tonsystem, das eine Oktave in 12 Halbtonschritte unterteilt, ist nicht perfekt. Stattdessen wohnt ihm ein kleiner Fehler inne. Wir hätten gerne, dass jeweils 12 Quinten und 7 Oktaven aufeinandergeschichtet die selbe Frequenz bzw. den selben Ton ergeben. Das ist jedoch nicht der Fall. Es entsteht eine kleine Differenz, die sich speziell bei Instrumenten mit festgelegten Tonschritten (Bünde, Tasten etc.) unangenehm bemerkbar macht. Die Differenz lässt sich mathematisch ausdrücken:
Seit langer Zeit versuchen die Instrumentenbauer, das pythagoräische bzw. syntonische oder auch didymische Komma durch verschiedene Stimmungssysteme möglichst gut auszugleichen. An manchen historischen Instrumenten findet man gar Subsemitonien, das heißt geteilte Obertasten, also mehr als 12 Tasten pro Oktave, wie beispielsweise zu finden an der Tastatur der Waller Kirche oder der Rekonstruktion einer norddeutschen Barockorgel in Örgryte, Göteburg.
Beim Spielen historischer Stücke lohnt es sich, nach Möglichkeit eine zum Stück passende Stimmung zu verwenden. Um die Auswahl zu erleichtern, hier eine kleine Übersicht.
Die pythagoräische Stimmung bzw. quintreine Stimmung ist im Mittelalter (ca. 850 bis ca. 1550) auch auf Tasteninstrumenten gebräuchlich. Da 12 Quinten etwas größer sind als 7 Oktaven, besteht sie aus reinen Oktaven, 11 reinen Quinten und einer stark verkleinerten Wolfsquinte.
Die reine Stimmung entsteht in der zweiten Hälfte des 15. Jahrhunderts und findet sich in Renaissance sowie Barock. Sie nutzt für die Intervalle Frequenzverhältnisse kleiner ganzer Zahlen. Der Klang ist schwebungsfrei und nutzt neben den reinen Oktaven und Quinten der pythagoräischen Stimmung reine große Terzen. Um dies zu erreichen, werden allerdings zwei verschiedene Arten von Ganztönen in den Frequenzverhältnissen 9/8 und 10/9 eingeführt, die zusammen eine reine große Terz ergeben. Akkorde sind besonders rein, allerdings nicht in allen Tonarten.
Alternative Bezeichnungen sind natürliche oder harmonische Stimmung, englisch “Just Intonation”.
Die mitteltönige Stimmung war von der Renaissance bis ins 19. Jahrhundert (Blüte ca. 1550-1750) eine vor allem für Tasteninstrumente gebräuchliche Stimmung. Sie besteht aus gleichgroßen Ganztönen, die dem Mittel aus großem und kleinem Ganzton entsprechen, sowie aus vielen reinen großen Terzen. Es handelt sich um eine fast reine Stimmung, allerdings nicht für alle Tonarten.
Mit einer mitteltönigen Stimmung lassen sich beispielsweise Orgelwerke von Dietrich Buxtehude authentisch wiedergeben.
Die wohltemperierte Stimmung ist eine Erweiterung der Mitteltönigen Stimmung, die ab Ende des 15. Jahrhunderts entsteht. Der Quintenzirkel wird dabei ohne Wolfsquinte geschlossen, wobei die Quinten allerdings gezielt unterschiedlich sauber gestimmt werden. Es entsteht (im Unterschied beispielsweise zu Johann_Georg_Neidhardt 1706) auch bei dieser Stimmung eine deutlich ausgeprägte Tonartencharakteristik. Weit verbreitet ist sie etwa zwischen 1700 und 1870. Bekannt sind beispielsweise eine Reihe von Stimmungen von Andreas Werckmeister, von denen Werckmeister III am bekanntesten sein dürfte. Johann Philipp Kirnberger entwickelt ab 1766 mehrere Stimmungen für Tasteninstrumente, von denen vor allem Kirnberger III an Orgeln häufiger zu finden war.
Englisch: Well temperament.
Die meisten Instrumente werden heute in der gleichstufigen Stimmung gestimmt, die etwa seit Ende des 19. Jahrhunderts bekannt ist. Die Oktaven werden in zwölf gleich große Halbtonschritte unterteilt, die wiederum in Hunderstel unterteilt werden. Die Quinten sind nur geringfügig (1/12 Komma) verstimmt, was allerdings durch eine stärkere Unreinheit der Terzen (2/3 Komma) erkauft wird.
Alle Tonarten klingen gleichermaßen “unsauber” und können daher, auch innerhalb eines Werkes, vollkommen gleichberechtigt verwendet werden. Sie weisen aber eben auch keine individuellen Merkmale mehr auf.
Alternative Bezeichnungen sind gleichtemperierte oder gleichschwebende Stimmung, gleichschwebende Temperatur, englisch Equal Temperament oder 12-tone equal temperament.
Ages ago I tried to cope with the EG2 of the Kronos’ software implementation of the MS-20. Meanwhile there are hardware reissues of at least two manufacturers, KORG and Behringer. The behaviour of EG2 is consistent to the original machine, which makes some users scratch their heads.
Many synthesizers provide two ADSR envelopes, one for the VCA (volume contour), and one for the VCF (which most often is a low pass filter).
For the VCA envelope, you’ll not find a parameter “Envelope amount”. The envelope usually starts with the VCA fully closed (except there’s a parameter “Initial gain” and set to something different than 0). After a key press, the envelope will fully open the amplifier at the end of the attack time, then fall down to sustain level and slowly close the amp after the key was released.
The MS-20 features an additional parameter “Hold time”. It essentially prolongs the sustain level after the key was released, similarly to pressing down a hold pedal for a couple of seconds. I’ll leave this parameter alone for now, so what remains is an ordinary ADSR envelope.
EG2 is always coupled to the VCA and behaves exactly as described above. It is not possible to decouple this connection and use EG2 for the filters exclusively, e.g. by patching EG1 to the “Initial Gain” jack. This behaviour may cause the first confusion for many users.
Now for the filters. Either of them provides a knob “EG2/Ext” which allows to control the amount of envelope to be applied to them.
ADSR envelopes applied to a filter by an amount knob usually behave very similar to the envelope applied to the VCA, except for two little differences. The envelope does not open and close the filter completely. Instead, it affects the filter cutoff only by the amount as set by some controls like the ones depicted above. Usually it does so starting at the currently set cutoff frequency as the baseline. Any sustain level greater than 0 thus shifts the cutoff frequency by a certain amount as long as a key is pressed.
The MS-20 behaves differently. From the original manual:
The EG-2 output is internally patched to the VCA so that changes in volume over time will occur to sounds according to the EG-2 control settings. In addition, another EG-2 output is sent to the Filter Modulation controls (EG-2/EXT) so as to modulate the VCF's cut-off frequency. In this mode the filter's "steady state" is the Sustain Level. The filter's cut-off frequency will start below this level initially, will rise above the Sustain level during Attack cycle, will fall to the Sustain level, and then will fall below to the initial level again following trigger release.
This special behaviour still was easy to adopt. What makes it difficult to cope with is the fact, that EG2 still is coupled to the VCA. So here’s a little résumé for all of us who approach the MS-20 with the mental model of our usual comfort zone:
The parameter “Hold Time” prolongs the sustain after the key was released, similar as pressing a hold pedal for a couple of seconds. It does not act as a delay before the attack phase. I recommend to just leave it alone (at least as long you are still not familiar with the machine).
EG2 always controls the amplifier. There is no means to get rid of this connection.
In case that Attack, Decay and Sustain are all set to their minimum, not much will be audible, since the amplifier opens and closes immediately. This is the behaviour to expect from an amp envelope.
A second output of EG2 is routed to the filter’s cutoffs by default, unless a patch cable is plugged in via the patch panel. This second output behaves differently than the amp output, since the sustain level will never be added to the set cutoff frequency.
Raising the sustain level will not raise the filter’s cutoff as one might expect. Instead, it moves the envelope below the cutoff frequency by the amount of the sustain level. This is only audible during the attack and release phases. In case both are set to their leftmost position, it is not audible.
Raising the sustain level reduces the headroom for the decay, just as it does on a common ADSR envelope.
Reducing the sustain level does not lower the cutoff. However, since it closes the amplifier, the sound will loose volume.
To circumvent the issues, EG1 could be used for the cutoff instead. As soon it gets patched via the patch panel, EG2 gets decoupled from the filter. However, EG1 is quite limited, since it is missing decay and sustain.
A further solution is to use an envelope generator of another synth. Maybe some eurorack modules, or another synth wich has CV outputs from its envelopes. So does the Pro-1. I connect both to the MIDI keyboard, turn down the volume of the Pro-1 and use its filter envelope to control the cutoff of the low pass filter, and its amp envelope to control the high pass filter. This gives me much better control over the filter of the K-2 than the device itself provides. Problem solved :) .
On Tuesday Chick Corea, the musician who probably coined my musical taste most, died at 79. I first heard him jamming with his band at Colos-Saal, and the last time with Bobby McFerrin in Baden-Baden.
I have no clue how often I listened to Romantic Warrior, especially the synth solo of the »Duel of the Jester and the Tyrant«, during my youth. I also was deeply impressed by later 70s titles, e.g. »Falling Alice« of the album The Mad Hatter, where Chick excellently combined traditional and electronic instruments. Not to mention all his other work.
With Chick, an extremely influential and unique musician left the planet.
CAIN++ is a user oscillator to emulate plucked string instruments using the Karplus/Strong approach. Since the parameters are not absolutely intuitive, here are some hints by the author for a couple of patches:
C++lavinet Shape: Max Shift Shape: Max Pluck: 1 S Volume: 1 S Color: 1 Frequenzy: 20 Waveform: 1
C++Guitar Shape: Max Shift Shape: Max Pluck: 43 S Volume: 1 S Color: 1 Frequenzy: 20 Waveform: 1
Soft C++ Bass Shape: Max Shift Shape: Max Pluck: 98 S Volume: 32 S Color: 100 Frequenzy: 41 Waveform: 1
A couple of days ago, I had the occasion to directly compare both synths. Since there are more soundsets available for the xd than the prologue, I used Sunny Synths prologue panel, which can import minilogue xd paches. Those files can be extracted from the librarian provided by Korg, or extracted from an xd soundbank using loguetools.
The minilogue xd often gets perceived as a synth with a very interesting sound (which I’d like to confirm). Thanks to the direct comparison of both machines, I can confirm that it’s not the filter that makes both synths sound different. Actually, both are very close, and the filters seem to be identical. Many patches I listened to sound identical on both machines. Though not all.
The secret of the xd are minor, but essential changes in its modulation capabilities. The LFO got a 1shot mode, essentially providing a third envelope generator. Additionally, sync and ring modulation are not XOR as on the prologue, but can be used simultaneously. So consider routing a 1shot sawtooth LFO to the pitch of oscillator 2, syncing it to oscillator 1 while enabling ring and cross modulation. This subtle design change provides a lot of interesting transients. However, this only affects a couple of presets which actually make good use of those features. Most presets sound almost identical on either machine, including lead and bass sounds.
What I noticed was that the oscillators sound slightly different. The prologue’s seems to provide less partials while being more uneasy. The latter issue may be beneficial for pad and string sounds.
So which one is the better machine? Well, they’re almost identical. As a keyboardist, I prefer the prologue due to its better keybed. I still wish, though, it borrowed the modulation design and the joystick of the xd.
Due to a high pressure system from west and a low pressure system from the east, the upper rhine plane got some snow flakes during the last night. Since the temperature was above 0°C all day, the snow was melting quite quickly and is almost completely gone. A nice intermezzo anyway.
The history of Sequential’s (or temporarily Dave Smith Instruments’) legendary Prophet synthesizers started over 4 decades ago. For details, head over to Youtube.
The REV2 features a fully analogue signal path, though its oscillators are not voltage, but digitally controlled (DCOs instead of VCOs). Compared to its predecessor 08, it is being said that the keybed was improved (Fatar TP/9S). The power supply is now internal instead of a wall wart. USB-MIDI was added as well as eight banks of preset memory (instead of two). The REV2 can load presets from the 08, though they might not sound absolutely identical due to subtle changes in its design. All waveforms now obey to wave shaping, and a sub oscillator was added as well as a digital effects unit. The modulation matrix now provides eight (instead of four) entries. The REV2 actually appears as a polished 08 and is, despite its complexity, very accessible due to its great user interface.
The machine is available in eight and sixteen voice flavours. Eight voice users can purchase an extension board, which is not available at a bargain, though. 16 voices are recommendable, since sounds with notable release times are one of the strenghts of the REV2. During startup, the Rev2 displays the amount of voices available. Since the board can be easily removed, this message is of special interest for s/h buyers.
Unlike the Deepmind, the REV2 does not depend on its digital FX engine. For home or stage use, it’s a welcome addition anyway. Studio users may use an already present external processor.
The 1024 presets on the unit cannot be categorized or searched by certain criteria, and unlike the Korg logue synthesizers, no official sound librarian is provided. Recently, Knobkraft Orm, an open source tool, appeared, which lets the user load presets in SysEx format and easily browse and audition its content.
One of the strenghts of the machine is its massive modulation capability. It provides four LFOs with five waveforms each, which can be routed directly to several destinations. The top right hand section provides a third envelope generator, so modulating some oscillators pitch does not limit the automation of the filter’s cutoff contour. The top left hand section provides a modulation matrix with eight general purpose entries and five additional entries. The latter ones grant direct access to controllers like the modulation wheel, aftertouch, breath controller, foot controller or velocity. Evolving pads and living strings are one of its natural domains.
The oscillator section provides sync, but no ring– or cross-modulation. However, the top right hand corner of the filter section provides an interesting audio mod pot, a feature seldom to find.
There’s a lot more to be said, since the Prophet Rev2 is quite deep. Just search the web for more details about this outstanding synth.
After the minilogue (2016) and the monologue (2017), Korg released the prologue in 2018. Finally keyboardists get something similar to the minilogue with full sized keys (available in flavours of 8 voices and 49 keys respectively 16 voices and 61 keys). What remained consistent were the two analogue oscillators and their waveshaping capabilities.
The filter design differs significantly. Additionally, the prologue introduces a third (fully digital) oscillator, called Multi Engine. It provides either a noise generator, a VPM/FM oscillator, or 16 oscillator slots the user can populate via USB. As with the minilogue xd, such oscillators are available e.g. by hammondeggs or Sound Mangling (Tim Shoebridge) or can be created by the user using the publicly available logue SDK.
The full size keybed is convenient to play, though it lacks aftertouch capabilities. Like the xd, it allows to control several of its parameters via aftertouch using external keyboards, though. Surprisingly it shares a further limitation with the xd, the complete absence of velocity control except for the two envelope generators. At least it provides a sustain level for its 2nd envelope generator, which sets it apart from the xd.
With the xd, it shares its 500 program slots, of which 200 are prepopulated by factory presets. Unlike the minilogue, neither the prologue nor the xd do benefit of the additional soundsets.
One review I read was like “The prologue sounds great, though it is not very deep”. What it does, it does extremely well. What it lacks is modulation capabilities. One LFO only, two standard envelopes, limited velocity support and the lack of a modulation matrix. Some may tend to go for a Sequential Prophet 08 or REV2. However, since it sounds differently, it may not be a real option.
The current lineup of Korg’s logue synthesizers consists of the minilogue (2016), the monologue (2017), the prologue (2018) and the minilogue xd (2019). Slowly and steadily they conquer the terrain of analogue synthesis.
The minilogue xd shares the form factor with its predecessor minilogue, though featurewise it is more a shrinked prologue than an enhanced minilogue. Compared to the latter, additions include a stereo output, an improved effects section, the inclusion of a third digital oscillator (equal to the prologue), a changed filter design (much closer to the prologue than to the original minilogue), and a joystick controller.
As with the prologue, the digital Multi Engine provides a noise generator and a VPM/FM oscillator. Additionally there are 16 slots for third party oscillators, with the Waves oscillator preinstalled. Further oscillators are provided e.g. by hammondeggs or Sound Mangling (Tim Shoebridge) or can be created by the user using the publicly available logue SDK.
The device gets shipped with 200 of the 500 memory slots populated with presets. Korg didn’t port the additonal soundsets of the original minilogue to the xd. An open source converter exists with loguetools, but since the machines differ, the results are not perfect and require manual adjustments.
For keyboardist, there are drawbacks. In case the mini keys are to small, get the module version and connect it to a full size keyboard. A painful omission is the lack of sustain and release for the envelope generator. The latter actually exists, though it is not exposed to the user interface. Instead, the decay time is used as release time. The former, however, is definitely missing. This is painful, since it is essential for dynamic sounds like Brass emulations. A mitigation was to control the cutoff frequency by velocity. Unfortunately this cannot be done with the minilogue. The xd’s keybed does not support aftertouch. Interestingly a recent update allows it to control several parameters, including the filter’s cutoff. But velocity control is still missing. Keyboardists may be served better with the prologue.
Users who want to use it in conjunction with a digital audio workstation (here’s a matching video of Korg) may either prefer the original minilogue (since threre are much more downloadable presets available). The xd may be a better choice in case one wants to use it for bass sounds (due to the changed filter) or the Multi Engine is of interest.
In 2016, Korg released the Korg minilogue. Despite its size (and mini keybed), it is an synthesizer with excellent audio qualities. It still is sold for about 70 € less than its successor, the minilogue xd, which appeared in 2019 (after the release of the flagship prologue in 2018). Surprisingly, it provides features the xd lacks, e.g. the amplitude of the LFO can be controlled by the envelope generator, and noteworthy, the EG is of usual ADSR type, while the xd only provides attack and decay, no sustain or release.
The minilogue’s filter sounds completely different compared to the xd. Especially the resonance lets one create nice sweeps the xd cannot do this way, while it notably thins out the lower frequencies, which the xd doesn’t.
The device gets shipped with 100 presets populating its 200 memory slots. Korg provides seven further soundpacks as free downloads. Further artists have created soundpacks, and posted demonstrations of the presets on Youtube. One of them is Adam Borseti, who showcases some of his presets in this video. They are currently available via Reverb.com.
The minilogue, the minilogue xd and the prologue are quite similar and provide excellent analogue sound. In case the third user programmable oscillator of the xd was of little interest, I’d still go for the original minilogue.
The Prophet Rev2 features an undocumented special menu. Besides other things, it allows to copy user banks to preset banks (and the other way around). A very handy feature in case one wants to host other than the factory patches in the “ROM” preset banks.
To call the “debug” menu, just press the buttons down, up and global simultaneously.
Obviously, I could have figured this out in a couple of seconds by reading some man pages. But searching the web was even more convenient. So I leave it up to you to figure out what exactly the above commands are doing. Alternatively, just use soxi :) .
The Prophet always is loading a certain patch when booting. Unlike other instruments, it does neither remember the last used preset nor start with the sound of the very first memory slot. Instead, the user can actually set up the sound to load the following way:
Select the preset the Prophet shall load at startup.
Modulation is a key ingredient to synthesizer sounds. Modulation adds movement to the sounds, making them less static. Common modulation techniques include envelopes to shape the volume and partials over time, low frequency oscillators (LFO) to periodically alter the pitch of oscillators (vibrato) or the volume of the sound (tremolo), or velocity to add volume or brightness to notes played harder than others. Advanced synthesizers provide a modulation matrix which allows to route modulation sources (like the ones mentioned above) to various destinations.
As a rule of thumb, the more modulation capabilities exist, the more different sounds can be produced by a synthesizer and the more complicated it is to understand and use. As a consequence, some synthesizers are designed focussing on the quality of the basic building blocks (oscillators, filters) and provide little modulation capabilities (e.g. one LFO only synths like the Korg prologue and minilogue xd or the DSI/Sequential OB-6, for which Steve Hunt offers a “Low Frequency Expander for OB-6” breakout box).
When it comes to LFOs, it’s not only the amount that matters, but also the waveforms provided. The Korg prologue and minilogue xd’s LFO just provide three (saw, triangle, square). More advanced LFOs may also provide Random/S&H waveforms. A waveform seldom to find is Sample & Glide, which provides a less stepped random waveform. It can be found, for example, in the Korg Z1 (four LFOS) or emulated in the John Bown Solaris (four LFOS) by routing the S&H waveform through a lag processor (actually a low pass filter). This waveform allows for very subtle, unobtrusive modulations.
The abovementioned Korgs just provide one LFO with three wavefroms only. Additionally, the single LFO can only be routed to one destination at a time (pitch, shape, cutoff). This design is pretty straightforward, and the panel controls give the musician immediate access to the modulation. While I’d like to congratulate the designers for resisting the temptatoin to add more, I perceive the design as very limiting.
At the other end, there are synths (like the Access Virus or the Sequential Prophet 08) which provide both multiple LFOs and a modulation matrix. One can create a lot of very interesting sounds with them, but without reading (and understanding) the modulation matrix menu entries it is difficult to understand how a particular sound was designed. But anyway, the sounds of analog machines with modulation matrix are much sought after, as the still popular Oberheim Matrix-1000 proves.
The original Waldorf Microwave I, known for its extremely punchy envelopes, provides two special parameters for the LFOs. The first one is “Symmetry”, which has a special effect on the sine wave.
The second one is “Humanize”, which adds random variations to its rate. The Microwave also is known for its extremely flexible (and even loopable) third “Wave envelope”.
Admittedly, I’m a big fan of all kinds of weird modulation capabilities and can spend hours exploring what sounds can be achieved using them. On the other hand, machines limited in modulation are interesting as well, as I pay more attention on the details of the limited capabilities, which fosters creativity, as Kebu states.