Organ-grinding tweets for August 2017

Irregular rhythmic irregularity in southern German music

How to write pastiche inconsistent non-isochronous subdivision for mechanical organ. With metrical analysis of the notes inégales in a bit of the Furtwängler recording of Mozart’s Gran Partita.

I still haven’t managed to find a communicative functionary in the northwestern Czech Republic, but I have as a result got to know several nice southern Germans with interests in mechanical music and pre-Sarajevo central Europe. One of these has now put me in touch with a Swiss gentleman who was looking for someone to write a couple of one-off arrangements of Austrian favourites for his roll-controlled salon organ for Mrs’ birthday. The catch is that he wants the music to sound not as it might be written but as it is played, which involves rhythmic give-and-take far more complex than for example American swing.

Irregularity in Austrian waltzes

The best known example of Austrian-ish rubato is the Viennese waltz, which British classical musicians, with their honest Hanoverian (i.e. Prussian) rhythmic heritage, tend to perform by making their beat 2 eat into beat 1’s tail by roughly a semiquaver and with simple tempo changes. Danubians, however, (still) vary the amplitude of the beat-2 lurch from bar to bar, depending on factors such as where they are in the phrase and what’s happening in the melody (the bar subdivision of which may not coincide with the accompaniment), and tempo is up for grabs to a far greater degree.

I should say that waltzes like this aren’t designed for people with three legs of unequal length: as Harold and Meredith Sears observe, the Viennese waltz is fast, and dancers tend to either canter (foot down on beats 1 and 3) or hesitate (foot down on beat 1); while Franz Fuchs from Klosterneuburg on the Danube compares it to Alpine skiing and indulges in a bit of gratuitous abuse:

It has everything, except the Prussian goosestep.


Irregularity in other Austrian genres

This variation extends to other forms. Franz has written a simple guide with MIDI examples of how to take a written polka and turn it into music (version 4), as well as an excellent introduction to the the application of inequality in various genres, including the Viennese and Alpine waltzes, Ländler and Steirer, the polka and Boarisch, marches and mazurkas. But not yodelling, which is beyond irregularity:

But this irregularity is also irregular. As Franz wisely comments:

None of these indications is to be regarded as an absolute must. It always comes down to the nature of the melody. In folk music there are no exact rules. Time and again you will find situations or entire melodies where this information simply isn’t appropriate and where you’ll have to accentuate differently. Some waltzes are if anything fast Ländler, while in others some passages are played in the exact metre. You have to try it out for yourself with every piece. Above all, you have to listen and investigate how other musicians do it.

Here’s a nice polka from Die Hoameligen showing amongst other things that stuff gets tighter when there are semiquavers around:

Obligatory Ernst Mosch moment

This eccentricity also extends to a considerable degree into the literate tradition. While Brits, Yankees and Prussians play their marches and polkas more or less straight down the line, their southern German counterparts deviate considerably. Regular readers know that I’m just looking for an excuse to usher on Ernst Mosch and his Egerländer Musicians, here playing along with their gold disc at the 1965 Stuttgart Radio Exhibition:

… and for a curious contrast of absolutely no other interest, here they are again in 1981, celebrating their 25th anniversary with a piece of Sudeten swing (do check the rest of the recording, though):

Inequality in Viennese Mozart

It’s also found in performance of parts of the classical repertoire untouched by popular dances. The Furtwängler recording with the Vienna Phil in 1947 starts up the bar 2-3 vamp in the Adagio from Mozart’s K.361 Gran Partita like a steam train on a cold, wet morning, and shifts continually throughout:

In contrast here’s a recording by Mackerras and a group of London players which, while superior in many ways, slots into a very slightly lilting rhythm in bar 2 and more or less stays there:

Just for fun I’ve had a bash at calculating the ratio between semiquavers 1 and 2 for each quaver in bars 2-3 Furtwängler’s notes inégales. Here is the relevant line in the score:

… and here in Audacity format is the relevant clip from Furtwängler, slowed to roughly half tempo and containing my time event labels for semiquavers, which are available here as a text file. (Judging when a note happens is akin to determining whether one’s favourite uncle has entered the room on the basis of his stomach or his smile, so by all means suggest improvements. 1)

The following chart shows for this miniscule data set the irregularity of the irregular ratio between semiquaver 1 and 2 for the 16 quavers in question:

Grouping the quavers into 4s, you get the following mean semiquaver ratios, which might indicate that players tend to favour long-short feet at the beginning of a metrical unit and short-long at the end:

Austrian rhythm and notes inégales

I think that the irregular ratio and the use of short-long as well as long-short probably distinguish this practice from French notes inégales, which I understand to be invariably long-short. If the two are related at all then I imagine notes inégales to be a pedantic, lobotomised offshoot of some Austrian or Italian tradition, rather like the robotic valse-musette.

However I wonder whether all those treatises detailing performance practice actually tell the whole story. Perhaps Thurston Dart, who I seem to recall in the Interpretation of music preaching against mixing equal and non-equal notes, would have revised his opinion had he been as fond of mid-C20th German popular music as he was of American jazz. Perhaps there’s a Two Cultures thing going on: progressives play early music and conservatives play commercial folk, and ne’er the twain shall meet, till earth and sky stand presently at God’s great judgment seat, and that’ll be fun. Perhaps, perhaps, perhaps.

Loss of inequality in barrel organ music

I haven’t seen any evidence of inequality in serinette, flute clock and other barrels from the C18th Black Forest or other southern German foci of production, nor have I found great criticisms of a loss of rhythmic flexibility in historical Central European writing. Perhaps this is because the tunes being pinned were often religious or foreign or high art and thus exempt, or because pinners were working from paper scores and were unmusical or didn’t regard interpretation as within their contract scope, or because irregularity was a characteristic principally of marginal (bi-cultural) areas (think of the Oberkrainer, Zillertaler, Egerländer repertoires…) about which no one cared a great deal, or because irregularity was regarded as Luddite and mechanical constancy as the future, or because people were so delighted with such cheap and reliable results that the loss of eccentricity seemed a minor price to pay, or because grinders added variation themselves by modifying their crank speed. More perhapses. In a way the survival of such techniques is another echo of the Audi cliché of the southern German lands: Vorsprung durch Technik … und Tradition.

Down to work: my modus operandi

Based on my limited knowledge of the styles involved, and bearing in mind Mr Fuchs’ comments, I’m going to record simple mixes with Audacity, MIDIfy the results, import them into Sibelius and then fiddle around until I get some results, which, since they are going to Switzerland, are at least one transaction unlikely to be affected by Brexit.


  1. I hope you’ll also read W. Bas de Haas & Anja Volk, “Meter Detection in Symbolic Music Using Inner Metric Analysis” and report back