Werner, Early 19th century engravings are hardly references to go by (18th century manuscripts can also look nice but are for a variety of good reasons quite far from current engraving best-practices), and unless we're not looking at the same pictures, the conclusions one draws from the Rachmaninoff examples are mostly *1) in all of them, no beam within the stave crosses more than 1 staff space* (or, if they start outside the stave, they do not cross beyond the first line); so that is still shallow, in some cases shallower than the default angle Lilypond gives, and also matches the Gould recommendation, and there is noticeable stem lengthening/shortening to accomodate this.16th beams even being allowed to cross a staff space or not is a matter of house style - something which one would eventually hope to have available as an option but which is separate from the beam end placement issue. As a test I picked a few short beams as examples, Lilypond does not engrave either quite as in your source. The first two nearly match - 1/2 shallower space for the first one, 1/4 steeper for the second one - the edition you give, in any case they don't strike me as problematic [the second one could be shallower given how close the notes end up being, but that will clearly depend on personal preference]; but the last one (full beat 16ths) is clearly too steep: [image: image.png] *2)* the only steeper beams are entirely outside the stave, or only within it at one end (so considerations about beam angle are not hindered by beam end placement) *3) in all cases, beams ending within the stave follow the placement rules* that have been mentioned (i.e. none ends in the middle of a staff space), even if this leads to a beam angle which does not quite follow the notes. "[indicating] the direction of voices" has little to do with beam end placement.
How steep beams are by default is something which can be adjusted by Beam.damping (fully customisable by-interval defaults, as I mentioned in the thread about beam shallowness last week, would be a nice feature, but are probably not something which should be put at the top of the priority list at this stage). In any case I don't think beams being steep or not should have an impact on them being placed properly. Cheers, Aleksa Am Mo., 22. Dez. 2025 um 01:18 Uhr schrieb Werner LEMBERG <[email protected]>: > > > Stems being longer to accomodate correct beam placement is in fact > > the traditional engraving practice - explicitly mentioned by Gould > > and others. > > No, it is not. This is a modern development that happened mostly > after 1945. In early typsetting done on pewter plates (i.e., the > early 19th century) you will find *much* steeper beams, and even the > 'classical' engravings around 1900 use steeper beams. > > I like to use the first version of Rachmaninoff's second piano sonata > (published in 1914) as a good example for fine typography; attached > are some examples of beams (published by Gutheil) that you consider as > 'problematic', but which were standard then – see the attached images. > Obviously it was more important for the engraver to properly indicate > the direction of voices than to take care of crossing staff lines. > Actually, being a pianist, I agree with this assessment; it helps a > lot while reading such complicated music. Additionally, it allows for > more compact vertical typesetting. > > Neither Ross nor Gould apparently take this classical time into > consideration for their recommendations. > > > Werner >
