[MEI-L] Measure 2 of Erlkonig; expanded notation

[Roland, Perry D. (pdr4h)]

Although I didn't see it before, I believe all the components of a render-able and playable solution are already present, no additions or transformations required. 

<bTrem form="meas" dur="4" num="3" unitdur="8" num.visible="false">
  <chord stem.dir="up" stem.mod="1slash" dur="4" dots="1">
    <note pname="g" oct="3"/>
    <note pname="g" oct="4"/>
  </chord>
</bTrem>

@form establishes this tremolo as measured; that is, individual notes should be audible, it's not a "buzz";
@dur captures the duration of the tremolo -- the chord/event inside should be aligned using this value, not chord/@dur;
@num records the number of notes in a sounded rendition;
@unitdur (currently called "measperf", but renamed in v. 4.0) is the written/base duration of the individual notes;
@num.visible says that the value of @num should not be displayed.

The attributes on <chord> should be used to render the chord visually; that is, with a dotted quarter-note duration and an upward-pointing stem with a slash.  These attributes should remain on the <chord> to differentiate them from potential attributes with the same name; i.e., @dur and @dots, on the <bTrem>.  Not to put too fine a point on it, but @dur (pluse @dots when necessary) on <bTrem> encodes the logical duration of the entire tremolo, @dur on <chord> captures the written duration of the chord.

These attributes should also remain on the chord because a naive encoder (OMR software or student) can only see what's on the paper.  In other words, the <bTrem> element represents an additional level of interpretation that's not present in the actual notation.  The stem modification is a clue that it should be present and could signal some kind of default interpretation; that is, the <bTrem> and default values for its attributes could be inferred based on its presence. Because it would be an automated interpretation, however, it may require human intervention to make sure it works properly.

--
p.