Stems are supposed to be 3.5 stave spaces tall. When the notehead is on the 2nd line outside the staff or beyond, the stem should extend to touch the center line. When the 3rd flag (or beam) is added to a note, the stem is lengthened.

That part I get.

  1. What I don't get is how to size all the stems attached to a beam. I've looked in a LOT of books and it's still unclear to me ... anybody know?
  2. Is there any standard for how much to tilt the beam (if it should tilt)?
  • 3
    I'm not at all an expert on typesetting, but I suspect that the answer will be that stem lengths are dictated by ease of reading and aesthetics.
    – Babu
    Commented Jul 17, 2012 at 4:00
  • Sometimes the noteheads are found on both sides of the beam. Are there rules for that one might wonder? Commented Jul 19, 2012 at 12:07

3 Answers 3


Convention differs. The 3.5 spaces thing isn't a fixed rule, either. Remember that notation is arbitrary and you can twist it to suit your needs. As Babu says that can include readability and aesthetics.

Personally, I just try to ensure the average is approximately 3.5 spaces. A nice trill might be 4-3-4-3-4-3-4-3, for example. Obviously in some cases you'll need to angle the beam to achieve that, and I'd basically follow luser droog's advice in that case (where possible). In some cases you might want to forgo the beam as well — if they're really hard to connect it might (might) be an indication that they're not really part of the same phrase.

  • 2
    I think the 3.5 space stem length is VERY standard, except where there are reasons to modify the stem per #flags, reaching middle line for ledger notes, etc. For hand written sheets, you can do what you want. But if it's published, it probably needs to meet all engraving standards. Which are PRETTY dang hard to work out :/ Commented Jul 17, 2012 at 4:56
  • @Matthew Notation is a complex and nuanced subject. It's arbitrary by definition, but there are entire books written on the subject, and I do recall my theory professor at one point stating that beaming rules are incredibly complicated--but these rules DO exist. It's just less relevant for musicians to learn nowadays with the advent of Sibelius and Lilypond, unless you're actually writing the software.
    – NReilingh
    Commented Jul 17, 2012 at 16:55
  • @NReilingh For sure. I view it like general typography -- definitely some science involved, but at the end of the day kerning and serifs and so on are all up to the font designer.
    – user28
    Commented Jul 17, 2012 at 17:24

OK, it's much simpler than I thought. I asked on pianoworld.com and you just find the shortest stem of all of them in the beam and shorten it down to 3 spaces if beam is within staff or 2.5 if it's beyond the staff. If the notes in the beam are only one space apart, you don't shorten it at all, of course. The outside 2 are only used to determine if the beam slants (in addition to all the other slant criteria). How much to slant is not well defined from what I've found (I've read several engraving books) other than that slants should be slight as I've mentioned.


The beam should tilt or not depending on the relative direction of the two outer notes.

And again with the stems, make the outer two stems normal length (or as close as you can manage to maintain the proper tilt) and let the inner ones fall where they may.

  • 2
    Per the book "Behind Bars" (which is GREAT), the beam should be level if a) outer 2 notes are the same b) a repeating pattern of notes is beamed (say, b d f b d f) c) an inner note has a shorter stem length than the outside 2 Also, beams should be .5 spaces thick with .25 spaces between. They should either sit or hang on a line or be split evenly by a line, but NEVER be centered on a space. This leads to adjusting the stem length by .25 space for level beams so this will occur. Now, when the beam tilts, it's recommended to keep it at 1 space on average. I'm not sure your way will work..:/ Commented Jul 17, 2012 at 5:04
  • Ok, actually, i think you're right. But since beams are pretty level even when tilted, the shorter of the 2 on the outside gets standard length of 3.5, or shortened to a min of 3 (2.5 if beam is off staff). So if slant is low to hi with stems up, leftmost note gets 3.5 or slightly shorter length, and all other stems get lengths to hit the beam slant. Ok, I think I get how it works now. Thanks for the help yall - I appreciate it. I'm beginning to REALLY regret adding standard notation to my sequencer - arrgh ! Commented Jul 17, 2012 at 5:25
  • 1
    I'm still trying to dig up the book I've got on this. I suppose I'll have to resort to the Amazon history; I'm just not finding it. But all of these rules carry an implicit "...unless it looks better to do it differently." Commented Jul 17, 2012 at 6:36
  • Hmmm, in my example, I think I chose the wrong note for the anchor stem. Left note lower than right note, stems up, beam more level than the noteheads - would make the right note the shorter "anchor" note. (I think?) Commented Jul 17, 2012 at 15:13
  • Yes. A sort of trapezoid. If the short stem is less than some threshold (maybe try 2 to start?) then stretch them all to bring it to the min. ... A couple threads I started on this in comp.lang.postscript are here and here (I hope the links work, googlegroups is a bit "transitional" right now). Commented Jul 17, 2012 at 15:44

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.