I can't understand the note duration notation very well.

In the image below, the highlighted A in the second measure seems to start half a beat between the first and the second beat. But shouldn't we wait for the bass A two beat duration to finish, and only then hit the second A? How do we know when the second A should start? In this case I know intuitively (knowing how the piece sounds) that it should start half a beat after the bass A. How would we write these notes if we wanted the second A to start only after the bass A finishes its duration?

Enter image description here

  • 57
    My eyes! MY EYES! (that's really a terribly written part) Commented Apr 6, 2021 at 17:46
  • 39
    I don't know where you got that copy of the score for that piece of music, but it's been done terribly. Find a different copy of the score. Commented Apr 6, 2021 at 17:46
  • 29
    I've seen rubbish, but this re-defines the term.
    – Tim
    Commented Apr 6, 2021 at 17:52
  • 10
    I think I can smell burning wood… that must be poor old M. Couperin spinning in his grave...
    – Tetsujin
    Commented Apr 6, 2021 at 18:07
  • 10
    Never had so much fun reading through the comments of a question as much as this one.
    – Clockwork
    Commented Apr 7, 2021 at 7:43

5 Answers 5


I can't understand very well note duration notation.

No wonder. The music you're trying to read is objectively incorrect in several ways. I won't list them all, but as you've noticed the vertical alignment is out of order. Furthermore, the rhythmic notation does not comply with the meter. Follow the advice in the comments to get a different copy.

I would use the original edition, which you can find (among several others) at IMSLP. This piece is on page 6 of that edition.

You may prefer a modern edition, but you would be well advised to invest in a professionally edited and engraved one.

Edit: I see that you've indicated in the comments that this is a guitar arrangement. If you click "arrangements and transcriptions" on the IMSLP page, you'll see that there is an arrangement there of this piece for unaccompanied guitar.

  • 15
    "Incorrect" is a compliment... I've never seen such a terribly written part. It's missing a rest in the first bar, the rhythm is wrong, the part divisions are again wrong (and incoherent), there are rests that don't make sense, collapsing stems, overlapping alterations, it's missing the key notation, there are "orphan" cuts, written ornaments with slurs (?!?)... I believe the "arranger" just wrote the tabs and used some program automation to write the notes, with some terrible attempt to minimal (and horrifying) adjustments. Commented Apr 6, 2021 at 18:02
  • 4
    @musicamente I do not disagree. I chose the phrase "objectively incorrect" to indicate that the errors are not errors of style that have evolved over the centuries (for example, some old scores will place a whole note in the center of the measure rather than aligning it with the first note in another staff that has smaller note values). Such errors are more subjective. This is, as you say, way beyond that.
    – phoog
    Commented Apr 6, 2021 at 18:12
  • 3
    Thanks a lot, the IMSLP link is very nice, I'll delete the classclef one asap, learning to read notation with stuff like that only makes things worse...
    – George
    Commented Apr 6, 2021 at 19:05
  • 1
    @phoog Suggested revision: both the vertical alignment and rhythmic notation seem to be correct. The problem is the total absence of rests and anything to clarify which voice is which.
    – Aaron
    Commented Apr 7, 2021 at 14:25
  • 1
    @Aaron agree. Actually the note lengths are “correct” in the sense of how long they actually sound on the guitar. You need to read each voice as pertaining to one string, see my answer. Commented Apr 8, 2021 at 10:39

The score sheet you quote here is literally garbage. Most likely it is the result of importing a foreign file into a music program that follows different conventions (like exporting MusicXML in one program and reading it with another program that is not prepared for parsing the output of the other program) or trying to combine different voices by program in a manner not supported by the program and then saving the resulting mess instead of trying to figure out what went wrong and fix it.

Or it is the result of an attempted conversion of MIDI into a score sheet that has gone horribly wrong.

Or some music program has nominal export for the input format of a different program, but this functionality was never tested or used with music of the given structure/complexity.

One can speculate a lot about the exact reason, but the result is the same: this is garbage, and so much so that divining the intent has become impossible. You need to try finding a different source. Or, in case this file was given as a file purportedly in a format for your music program, see if the same source does not provide a ready-made PDF file for your use that is less likely to be garbled like this one is.


This writing is really very odd. But you can derive the correct rhythm if you imagine the eighth notes as off beat (counting “and”) ... there’s no logic in the notation of the half notes. Imagine you have a 4 voice part written on 4 different layers and channels. But it will still be difficult to derive a reasonable rhythm. It will be much easier to compare this miswriting with another arrangement like e.g. the following:

enter image description here

  • Wow, when you see it done correctly and compare how much is wrong with the OP's version, it's pretty amazing. It was in the wrong key, transposed a 3rd down (thus requiring too many ledger lines to be easily readable), wrong time signature, doubling the note-lengths to compensate, Commented Apr 7, 2021 at 19:52
  • 1
    @DarrelHoffman indeed a lot of things are wrong with it, but your points are not among them. –Yes, it's transposed to a different key, but that's hardly wrong seeing as the piece was originally for a different instrument and tuning standard anyway. It's not “too many ledger lines”, this is the standard style for classical guitar. ... Commented Apr 8, 2021 at 10:05
  • ... Yes, it's in ⁴₄ instead of ²₂, but that's just a matter of emphasis. The note-lengths are not doubled, they're just written in the actual time each string is sounding (the notes on different strings blend into each other), which is kind of correct in a performance-technical sense, but of course neither necessary nor helpful not properly written. Commented Apr 8, 2021 at 10:09
  • And incidentally, this score is pretty bad too – the overhanging notes across the bar lines are almost certainly not as intended by the composer. Commented Apr 8, 2021 at 10:43
  • I didn’t choose this notation because of its correct transcription, but to demonstrate the distribution of the eighth notes and to make the rhythmic pattern clear. That’s why I wrote: e.g. the following ... Commented Apr 8, 2021 at 13:09

I fully agree with all the comments remarking that the typesetting is a complete abomination here, however it does make “sense” in that the note values do express how long each note sounds on the guitar. Essentially, what happened here is that each string has its own voice, and the notes blend into each other. I give it five points for this detail attention... and minus 5000 for jamming this almost completely unnecessary information into the score without making it clear how it works!

What's missing completely are the rests that most of the voices have, if we're going to give a complete one to each string. Here's a “correct” rendition of this idea:

Bogus rendition with completely written string sounding lengths

Of course that's even more unreadable than the version you used. And it's completely unnecessary too to be so explicit about the length each string sounds, because it's generally obvious from the harmonic progression. The standard way to express “leave strings ringing” in guitar music is to simply add some slur-ties, use only as many voices as needed, and write those fully with all notes cut to the length and rest-filling needed to make the rhythm clear:

Proper version with slurs for legato

  • One possible explanation: OP's source must have went through MIDI at some point. Commented Apr 9, 2021 at 22:19

Observe that there are two voices, one with stems up and one with stems down. This is common in classical guitar transcriptions (but poorly done in your version). When reading such a score, remember that the up-stem (treble) voice is independent of the down-stem (bass) voice. So the A in the second measure starts on the first beat of that measure.

  • There are actually at least three voices. For example, the half-notes in the second measure comprise two voices on their own.
    – Aaron
    Commented Apr 7, 2021 at 17:14
  • In fact there are multiple voices with stems up in the first measure (simultaneous E's). So you can't just use the stems to determine which voice is which. "So the A in the second measure starts on the first beat of that measure." You might think so, but in the original, that note starts a half-beat later.
    – LarsH
    Commented Apr 9, 2021 at 0:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.