In transcribing a work for piano with two voices in each staff, I've come across a passage in which there are two voices in the LH staff for some duration, but in one measure there is only one voice played for 3/4 of the measure and then a 3/2 tuplet of 8th notes.

There are four ways I can notate this:

  1. three quarter rests on top of staff
  2. a half rest followed with a quarter rest (because when resting for part of a measure, it seems "normal" to place them in descending duration)
  3. a quarter rest followed with a half rest (to match up with the lower voice's tuplet and minim)
  4. nothing at all (lower voice's notes fill the measure)

Refer to the below image for examples of each of the four ways (respectively):
Notation examples (4) as described above

I'm a little unsure what the best way to represent this (I'm using LilyPond), so I'm asking: From a performer's, professional musician's, and/or composer's perspective, what is the best way to represent this in notation? "Best way" meaning least confusing, most common, and/or easiest to understand. Both voices have notes in the prior and subsequent measures.

  • 2
    I'm not sure whether this question comes down to personal preference or not. As it is meant to be played on the piano, the approach to the voices should be a bit different if you want to ensure clarity for a performer. From that point of view, I think there's an objective case to be made for the last option, because to a pianist it is far more readable than the other three. My piano brain would actually like to see no rests at all in any measures whenever at least one voice is playing, and I see nothing confusing about option four, it's just cleaned. But I think we need some backup for that. Apr 18, 2017 at 13:01
  • 1
    I want to clarify my "no rests when a voice is playing", I mean when a voice is playing on one of the staffs. So I only want to see a rest in the upper staff (for example) if my right hand is meant to be playing nothing at that time, not if only one voice has dropped out. Again, a real answer to this question would quote some well-regarded piano sheet music as "objective" support. Apr 18, 2017 at 13:04
  • I definitely prefer the first and last approaches to the second and third; not sure which I prefer over the other. Another point I'd suggest, however, would be that I'd nudge the first duple eighth note to the right just a smidge so that the note heads wouldn't form a straight line. Otherwise it's not visually obvious that the bottom three notes all join to the bottom stem.
    – supercat
    Nov 30, 2019 at 18:14

4 Answers 4


The last idea is not good, as it leaves the tenor voice with only the last crotchet taken care of. The third isn't so good, as the rests split the bar enevenly. Not a sin, but can be better. Second isn't bad, but would look better with the minim rest more to the right. I think the top one is easiest to read, given that for each triplet, there is a rest in the tenor part. Very clear.

  • 5
    Since this example doesn't contain anything that looks like "counterpoint", the last example omitting the rests is actually nearest to conventional notation. It would be better to have the stems up the first part of the bar, not down. To do that in Lilypond, look at the voiceOne, voiceTwo, etc, and oneVoice keywords in the documentation. If you really really want to include the rests, the "correct" way is example #2, breaking the rests at the half-bar, not #1.
    – user19146
    Apr 18, 2017 at 11:25
  • @alephzero - Don't know about Lilypond, but it looks like SATB to me, with stems as they are, first bar bass with stems down makes it easy to read and simple to work out voices. Your answer?
    – Tim
    Apr 18, 2017 at 11:58
  • @alephzero: Yes, I've become skilled with voiceOne, voiceTwo, as well as \stemUp, \stemDown, etc... But it's a hair-pulling experience: :-\ Anyway, this is indeed how I'm doing the above notation. Thanks!
    – pr1268
    Apr 18, 2017 at 12:48
  • I'm a little confused at whoever thought Tim's answer deserved a second downvote; surely one gets the point across?
    – Richard
    Apr 18, 2017 at 13:05
  • 1
    Tim, regarding your comment that it "looks like SATB", the asker explicitly specifies they are writing for piano, so it's definitely not SATB. As a pianist, I would find the rests distracting and confusing. If it were SATB, then I would agree having rests would make it clearer. Apr 18, 2017 at 18:02

In this excerpt there appears to be three distinct voices (I've added some colour for clarity):

  • The top voice (the melody) stays in the treble staff.
  • The middle voice (the harmony) starts off in the treble and moves to the bass.
  • The bottom voice (the bass line) stays in the bass staff.

At no point are any of the three voices resting, so it should look like option 4:

Three voices, with the middle voice crossing staves, without rests.  The melody stays in the treble staff.  The harmony is in the treble staff for the first three beats, then moves to the bass staff. The bass line stays in the bass staff.

Another way to represent this that separates the harmony from the melody visually, would be to move the middle voice to the bass staff for the whole measure:

Three voices, without rests. The melody stays in the treble staff.  The harmony stays in the bass staff.  The bass line stays in the bass staff.

However if I am misinterpreting, and there really are four voices in the piece, you will need rests in both staves: a crotchet rest at the end of the measure in the treble, and a minim rest followed by a crotchet rest for the start of the bass staff; it should look like this:

Four voices, with rests.  The melody stays in the treble staff.  The upper harmony stays the treble staff, resting for the last beat.  The lower harmony is in the bass staff, it rests for the first three beats. The bass line stays in the bass staff.

Notes (assuming four voices):

  • Options 1, 2, & 3 are not right, as you are missing a rest for the upper harmony.
  • Option 1 is also not right, as some of the rests can be merged.
  • Option 3 is not right, as the minim rest should not span across the centre of the measure.

It is fine to have no rests, as in the OP's fourth example. However, if there are to be no rests, it would be better to point the stems of the lower stave's first four notes in their usual direction (upwards), unless you think that there's a reason to the contrary which is compelling enough to warrant pointing the stems the other way.

If there are to be rests, then the OP's 1st and 3rd examples are wrong, but the 2nd example is fine. The first example is wrong because the first two rests should be combined into a minim/half rest (it is also correct to combine all three rests into one, and have just a single dotted minim/half rest). The third example is wrong because you shouldn't have a syncopated minim/half rest like that.

It is fine for the notes in a stave to be notated as if in one voice for part of a bar/measure, but in two voices for the rest. Here is an example from the finale of the first work I checked: Beethoven's Piano Sonata 1 in f minor.

Beethoven Piano Sonata 1 in f minor, finale, b.5

  • I agree with you and Todd Wilcox (in the comments on the question): there's no reason to have rests at all. It's not as though a voice has stopped: it's just traversing from treble to bass clef, and the rests make it look as though there's yet another, different voice, which is not the case. That said, I find the downward stems of the first four notes in the bass clef make it easier to separate this voice from the one approaching from above, but that's just my preference. Apr 27, 2018 at 19:47

In my experience piano music often leaves out rests in additional voicings unless they are necessary for clarity. Essentially, you don't have to add and/or remove voices only at measure changes. It is perfectly fine to split/collapse parts in the middle a measure, as in your 4th option (upper voice simply doesn't exist until beat 4). This also means that the part of the measure with only one voice should be notated as such (nothing special, just how you would notate it if there were no second voice at all, with the stems pointing up instead of down in your case).

I have performed choral music with voice splits mid-measure, but this indicates that both voices should sing unison until the split. If the voices are performed by separate musicians and only one of them should play you must include rests to avoid ambiguity. If one musician is performing all the voices (e.i. piano) there is no ambiguity, so it's never an issue.

Again, clarity matters! If it's at all unclear, err on the side of caution and notate the rests. However, in the example you posted, if you don't notate the rests, I highly doubt anyone would play it like this:

enter image description here

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.