Simultaneous rubati -> temporal metrical accentuation

[pfefferniels]

If I see it correctly, it is not possible to have to different rubati affecting different metrical layers at the same time. Why?

To model this performance e.g., I want to have a slight rubato per half bar and a slight inegalité, i.e. a rubato per eighth notes. So roughly something like:

<rubato date="0.0" frameLength="1080.0" intensity="0.7" loop="true"/>
<rubato date="0.0" frameLength="360." intensity="0.56" loop="true"/> <!-- at the same time! -->

Or is this rather a question for meico?

[axelberndt]

You are right here, this is no meico question but an MPM one.

It is correct that MPM does allow multiple layers of the same feature in the same map. This would open the door to very ugly constructions that are usually unlikely to make much musical sense.

Regarding your case, let us think about how to achieve the effect you want with only one layer. In fact, from pure listening (it's about time that I get the audio analysis widget in MPM Toolbox done!), I perceive two very different rubati in the left and right hand running against each other. In that case, one rubato would be modelled locally on the part for the left hand, the other on the part for the right hand. Is this the effect you wish to achieve?

Or do you have a problem modelling one of those individual rubati? In that case, which one and can you provide a visualization of the timing you want to achieve? This would help me to better understand the case.

[pfefferniels]

No, I mean only the left hand rubato.

So in this kind of tempo architecture

the lowest layer represents the duration of a sixteenth note, the one above eighth notes etc. The fact, that the second half of the bar is played faster is something I would encode as a tempo change. Now what is happening in between should be a rubato – so a slight accelaration of the three eighth notes. But every eighth note in itself also has a inegalité (in the sixteenth notes). This pattern could potentially repeat through all the piece.

[pfefferniels]

P.S.: the graphic above is based on this recording, where I try to demonstrate what I mean. Emil Sauer is actually doing something more complex …

P.P.S. It might also be understood as a kind of "quantitative" metrical accentuation …

[axelberndt]

Following our oral discussion we came to the result that the described phenomenon seems to be closely related to the musical meter. All the rubato-like timing distortions align with barlines and beat positions and recur or change measure-wise. It might be regarded as a temporal counterpart to the metrical accentuation feature that is already part of MPM. Within the frame of one measure certain beats are expanded while others are compressed.

We might consider extending MPM's metrical accentuation model accordingly. For now, please collect and list literature on this phenomenon here so we get a better understanding of its inner workings.

[pfefferniels]

Since I just stumbled on it, Lars E. Laubhold describes precisely that phenomenon in "Arthur Friedheims Einspielung von Ludwig van Beethovens Diabelli-Variationen für das Philipps-Klavierrollensystem Duca […]", in: Aringer, Utz, Wozonig (Hgg.), Musik im Zusammenhang. Festschrift Peter Revers zum 65. Geburtstag, S. 661:
"Darüber hinaus wendet Friedheim die heute besonders im Cembalospiel gebräuchliche Technik der agogischen Schwerpunktsetzung durch Ausdehnung der takthierarchisch schweren Zeit an." see here

[pfefferniels]

Another paper describing this phenomenon can be found here (Robert Hill, "Carl Reinecke’s Performance of Mozart’s Larghetto and the Nineteenth-Century Practice of Quantitative Accentuation", in: About Bach, Urbana 2008, S. 171-180)