I found a few web sites that describe how to write music with the correct rhythmic notation. If I understand the gist, it's that if a note crosses a beat boundary, the note should be broken up into tied notes so that there is a note on the beat.

Let me begin with this sample:

Applying the rule, I get this:

Am I understanding the rule correctly? And yes, the last double-dotted quarter note might as well be a half-note—less clutter and no significant sound difference.

Here's another sample:

Corrected to:

There must be a caveat about notes that start on a beat and extend all the way to the next beat. For example, if I have a whole note in one of the measures, I'm not going to change it to four tied quarter notes.

I will continue to research this topic, but if anyone can express this as a series of simple rules, I would love it. All assistance greatly appreciated!

You need to be a member of Composers' Forum to add comments!

Join Composers' Forum

Email me when people reply –


  • I think that the general problem/solution is a matter of arithmetic and rationality, with certain allowances made for conventions that may seem irrational at first sight.

    What is of the outmost importance is to make the part as easy as possible to read and to that end not only rhythmical divisions should be considered but a few things on the side, like placement and readability of other signs except from notes. That really takes us outside your main question into the art of music engraving, but there is no harm to take it into account some times.


    In your first example I find the corrections … correct! The second version certainly makes the part more readable and I totally agree that the last double doted crotchet should be written as a minim as the semiquaver rest will not be missed by most ears. (Unless of course the composer insists and he happens to be a total serialist in time values). :-)


    In your second example I note that you avoid giving the time signature.

    I recognize this type of rhythm instantly as it happens to be a pan-Hellenic type of dance, ie 7/8 in one of its two possible rhythmic groupings: either 3+2+2 or 2+2+3 (the remaining 2+3+2 division is only theoretical as if it is played more than once, meaning for a few continues bars, it is falling under one of the main ones given).

    From this perspective I find your second example as falling under the first grouping division 3+2+2 and the notated passage  completely right in the upper staff and without any need for correction therefore the corrected version is wrong according to this grouping observations I provided.


    Having talked of groupings and conventions, my opinion is that the first should be made clear as per composer's  intentions whereas the second should be studied and be familiar with.

    I don’t know which score writer you use, I use Sibelius 6. I hit letter "T" for time signature and in the screen that comes up I have all possible choices for numerator and denominator. After I set this as I like I have a further choice to define up beats if any and a further clickable box "beam and rest groups" giving me a new screen where I can chose my sub-divisions and groupings according to the logic of my chosen time signature from quavers (eighths) to demi-semi quavers (32nds).


    Here the groupings are of absolute importance and should be defined according to one's purposes because the programme defaults to the most usual. (mostly western European rhythms).

    I give an example of this importance bellow of how a piece should be written correctly in order to make clear its rhythmical stresses and how it is notated wrongly if I allow Sibelius to default.

    The piece is in 9/8 and Sibelius defaults to the usual 3+3+3 western European sub-division,

    whereas the actual rhythmic thesis and arsis of the song can only be realized in the 2+2+2+3 sub-division. (btw, my use of the terms "thesis" and "arsis" are as per original Greek meaning of the words, therefore the other way round of how are known in the rest of the world).


    For some seemingly wrong conventions perhaps in another post.

    Please provide the address of the site you mentioned.


  • Wow! Thanks for the info!

    Sorry about the missing meter for the second example. It's 4/4.

    I use MuseScore 2. For any meter, you can define the groupings however you want. I believe you can create a 2+2+2+3 grouping for a 9/8 meter and then create a different grouping also using 9/8 and you can apply whichever you like.

    As for the site, I found more than one using a google search:

    Still looking...

    The story behind the second example: I figured if Beethoven could write a symphonic work based around a 2-measure rhythm (5th symphony, first movement, of course), I would give it a go myself. I laid out various patterns pretty much by putting down notes at random and kept the pair you see above. The entire piece uses this pattern for most everything (OK, I loosen up a bit in the middle). I know nothing about pan-Hellenic dances.

    You can hear the piece at https://soundcloud.com/freixas/a-walk-in-the-park. Should I have written it as 7/8? I don't see how that would work as each measure has an extra eighth note.Neither 2+2+3 or 3+2+2 fits. Do you still believe it is correctly notated in the first version?

  • Sorry, I did not make myself clear regarding your second example.

    What I meant is the grouping you give in the upper staff leads immediately my eye to read a 7/8 grouping (3+2+2) which I would play automatically with correct stresses at quavers 1, 4,6.

    I am aware that your bar adds up to eight quavers, while Im talking of seven (of a rhythm very familiar to me in its groupings, disregarding the extra quaver of your example - my mistake this, you mean exactly what you write).

    So here we have your own effort to invent different groupings and stresses towards gathering rhythmic material for your piece and in that sense both groupings can be legitimate and correct imo. I would still notate the last two tied quavers as a crotchet, unless you have other reasons that I'm not aware of.


    I had a quick look on the first link you provide and I found the first wrong/right example given:


    That is ok the way it is explained as far as strait rhythms are notated.

    But it brings me to reconsider some conventions of which I talked earlier.

    The first bar is defined as "wrong" in that example, but I take it and I repeat it several times with stresses according to the given grouping of 3+2+3 (at quavers 1, 4, 6) and I find nothing wrong with it whether my time signature is 4/4 or 8/8. I have seen both these time signatures used for this grouping for instance in Latin American Rumbas. If we take out the first 5 quavers then the pattern becomes a continus 3+3+2 grouping. I post a PDF of a Cuban dance which I hope will demonstrate the point.


    I've heard your piece in sound cloud but rather quickly  and I did not concentrate on anything like thematic development, harmony, orchestration, etc, but I like its mood and find that the rhythmic content with its syncopations etc is a strong unifying element. I will have  another more careful listening. But I definitely thing that it is well defined in 4/4 rhythmically with clear pulse and down and up beats although the syncopations present in various instruments are very enhancing and exciting. That is your own rhythmic invention and you should keep it as is. Nothing like 7/8 there would have done so well.

    I also like the idea of the different movements given as a continus file, but I think, although nicely orchestrated it could benefit from more adventurous harmony, it is too sonorous to my ear, but then again, it's a walk in the park, so clarity and light and well being may be what you are after in which case, it does convey the feeling imo.

    I enjoyed the final climax very much and I thing the percussion ending is quite surprising as unexpected.

    Why don’t you post a PDF to the forum? I think people will like it!

    PS I forgot to mention: For the PDF guitar piece attached either 4/4 or 8/8 would do, but 8/8 is more correct in my opinion. This is what I mean by conventionally notating in some cases.


  • Thanks again, Socrates.

    I wanted to review the comments my friend made when he first saw the score for A Walk in the Park. In his first email, he said

    Quick glance at the score suggests that the transcription may be missing an eight-note beat in each 4/4 bar…  I’m counting 3 ½ beats in each 4 beat bar.

    After I noted that all the beats were there, he said:

    I’m good now on the 3 ½ beat confusion;  what threw me is the auto-insertion of a full-note rest on the upbeat, intellectually correct but unusual (as the full-note rest is commonly presented on the beat.   Same with the quarter note;  throws me to see the quarter note presented on the upbeat and not the downbeat.   Technically correct, but does stretch the brain (and the toe-tapping) a bit.

    He has a background as a trombone player in an orchestra (apparently hot stuff at one point). If his immediate reaction was that there was a 1/2 beat missing, then I would think I should pay attention to his concern.

    I'm meeting with a conductor next Tuesday to talk about getting this piece performed. I suspect I will come back from that discussion with a number of edits to make. Right now, whatever he says is the right way will be the right way. :-)

    I posted the link to A Walk in the Park mainly so you could hear the rhythmic pattern. I appreciate your comments on it and I may post it more formally to the group at some point. The piece is completely, absolutely diatonic, which explains the lack of an "adventurous" harmony. I've listened to it a bazillion times and I like the sound.

    I was aware of the limitation, though, so the next piece I worked on was not diatonic. It has not yet been "released", but you can hear it in the thread https://composersforum.ning.com/forum/topics/wind-quintet-review-cri.... This is the piece whose first measures are featured as the first example above.

    Wind Quintet: One - Review, Critique, Suggestions Wanted
    Here's my first attempt at putting up a work in progress—one not released in its final form to my adoring public. :-) Originally, this piece was writ…
  • If you use time signatures that say what note gets the beat, then you need to notate measures to show those beats. Musicians count rhythms by subdividing the beat, not transcribing the errors of the composer.
  • Hi, Rodney. I understand your first sentence, which implies that this is the correct notation (it's in 4/4)

    8607780852?profile=originalI'm not sure why musicians are transcribing anything, so I'm not sure I understand your second sentence.

    I have a background as a software engineer, so I like to have a simple set of rules that unambiguously tells me how to correctly notate any passage. So here's what I came up with:

    1. If a note or rest does not start on a beat and crosses the beat, break up the note/rest into two tied notes (no ties needed for rests). The first note/rest is the length up to the beat and the second is the remaining amount.
    2. If the note/rest starts on the beat, leave it as is, regardless of whether it crosses a beat boundary or not.

    The time signature by itself is insufficient to determine where the beats are, although there is good agreement for the most commonly used time signatures. The composer should always know, though.

    These two rules seems to cover all the cases I saw while flipping through some edited piano books. I did spot one case that might have just been missed by the editor, in which a 4/4 measure ends with a quarter note and then an eighth.

    I'm not really sure that rule #2 is correct.

    Finally, I might need to add another rule stating that the last note/rest in a measure need not be split even if it crosses a beat boundary. I think I saw some examples of that but right now the books I looked at aren't at hand.

  • Musicians should never have to transcribe, changing the notes of the composer to something readable. It's not their job, it's ours.
  • Btw, a time signature is more than enough to tell you where the beats are, and it even tells you what beats to stress accents.
  • Lol, sorry, eating while replying... In your last example I would've notated the second measure differently, because the three 8th notes beamed together remind me too much of triplets. I would just use a single 8th note on beat 2 and not beamed it.
  • Notation software generally ties eighth notes together 4+4 rather than 2+2+2+2 in 4/4 signature. Personally I find this extremely stupid, since it serves no real purpose and messes up reading as soon as any real rhythm appears.

    The rules given by Antonio are more or less how it works. You can add the exception of basic syncopation (eighth - quarter - eighth) notated as is, without the tie, provided it all fits within a "big" (half-note) beat. Basically can have it on 1 or on 3 in a 4/4 measure. This is a common thing and doesn't throw classically trained readers off.

This reply was deleted.