Transpose(P8) turns D#5 to E-6, e.g. in toSounding(piccoloPart))

[gregchapman-dev]

music21 version

9.2.0b2

Problem summary
When I do toSounding(piccoloPart), a transpose(P8) happens, and everything looks good except that D#5 gets transposed to E-6 instead of to D#6.

Steps to reproduce
I can reproduce this easily, and I will be working on it myself. I'm just checking with you (@mscuthbert and @jacobtylerwalls) to see if this is something you already know about, and can point me in the right direction.

[gregchapman-dev]

Edited because I had the to/from reversed.

[mscuthbert]

So weird! What is the .transposition on the Piccolo Instrument object (is it an interval.Interval P8 or something like interval.ChromaticInterval 12)? How was the Piccolo instrument created?

The only two file formats whose music21 converters seem to have the possibility of creating incorrect transpositions that I can see are MEI and Noteworthy Composer -- is your file from one of them?

[gregchapman-dev]

The piccolo inst.transposition is definitely P8 (not just 12 semitones). I don't think it's related to an importer (I parsed a MusicXML file and then called toSounding on the resulting score).

Bummer. I was hoping it was some edge case you knew about. Ok, I will just debug it through.

[mscuthbert]

I think that the MusicXML file is likely corrupted in the instrument definition (or it's a problem in our XML parsing), because this works fine:

>>> s = stream.Stream()
>>> picc = instrument.Piccolo()
>>> picc.transposition
<music21.interval.Interval P8>
>>> s.atSoundingPitch = False
>>> s.insert(0, picc)
>>> n = note.Note('D#5')
>>> s.append(n)
>>> s.toSoundingPitch()[-1].nameWithOctave
'D#6'

[gregchapman-dev]

Thanks! That helps me know where to look.

[mscuthbert]

Any progress on this?

[gregchapman-dev]

No, sorry, I've been distracted by other work.

[mscuthbert]

no problem -- just wanted to make sure there wasn't something I did holding it up -- I'll be here when it's needed!

[gregchapman-dev]

OK, tracked this through a bit. The instrument transposition is P8, but implicitDiatonic is True, so that probably is why the transposition was chromatic. Here's the instrument definition in the imported MusicXML file:

    <score-part id="P1">
      <part-name>Piccolo</part-name>
      <part-abbreviation>Picc.</part-abbreviation>
      <group>score</group>
      <score-instrument id="P1-I18">
        <instrument-name>Picc. (V2k)</instrument-name>
      </score-instrument>
      <midi-instrument id="P1-I18">
        <midi-channel>1</midi-channel>
        <midi-program>73</midi-program>
        <volume>80</volume>
        <pan>0</pan>
      </midi-instrument>
    </score-part>

I stepped through xmlToM21.py for this instrument, and it was imported with transposition.implicitDiatonic == False (as I would expect). An instrument was created from the midi-program and then replaced with one created from the instrument-name. Both of these had transposition.implicitDiatonic == False. So somebody (between import and the call to toSounding) must have modified or replaced the instrument.

I'll keep looking.

[gregchapman-dev]

Huh. instrument.transposition.implicitDiatonic is True at end of import (but it's False at end of xmlToM21.parseXmlScorePart). That narrows down the problem a LOT.

[gregchapman-dev]

Got it. It happens when parsing the attributes of the first measure in the Piccolo part:

      <attributes>
        <divisions>8</divisions>
        <key>
          <fifths>0</fifths>
          <mode>none</mode>
        </key>
        <time>
          <beats>3</beats>
          <beat-type>4</beat-type>
        </time>
        <clef>
          <sign>G</sign>
          <line>2</line>
        </clef>
        <staff-details>
          <staff-size scaling="108">108</staff-size>
        </staff-details>
        <transpose>
          <diatonic>0</diatonic>
          <chromatic>0</chromatic>
          <octave-change>1</octave-change>
        </transpose>
      </attributes>

The measureParser.transposition that is computed from this has transposition.implicitDiatonic == True, and then later this transposition overrides the good one in the Piccolo instrument.

@mscuthbert Do you have any idea if (1) this <transpose> attribute is wrong, (2) the code that parses it (xmlTransposeToInterval) is wrong, or (3) the code that does the override (updateTransposition) should choose not to override in this situation?

[gregchapman-dev]

Here's the full MusicXML file (zipped, so GitHub will let me upload it here):

ActorPreludeSample.musicxml.zip

[gregchapman-dev]

My opinion: I think xmlTransposeToInterval should be able to create a (non-implicit) diatonic interval from that <transpose> element. The code seems to inappropriately ignore the difference between 0 and None.