Beams and stems above or below notes

I think Dorico is sometimes making the wrong choice about whether a beam and its associated stems should go above or below the notes. The attachment provides an example where it seems Dorico is making its decision based on where the majority of the notes lie on the stave, but I think the decision should be determined by the note which is furthest away from the middle of the stave. I think the point of having the option of placing stems/beams above or below notes is to save vertical space. In this example the D is clearly the note furthest away from the middle of the stave and therefore it should determine that the stems/beam should go above the notes in order to minimize the total vertical space taken up by the stave and its associated notes and other marks.
Beamsandstems.dorico (840.9 KB)
I may be mistaken but I think this is pretty standard practice. Of course it’s not a huge deal as it’s very easy to flip the beams/stems to the other side, but it’s something quite basic, and I’m finding I’m really having to watch out for these. If Daniel and his team have strong views on such a contentious matter I’d appreciate at least an option for beams/stems to be treated in this way.

Gould says:

This one’s open to interpretation, isn’t it?

FWIW, Ross usually uses flat beams here with a group of 3 repeated notes:

That looks really poor IMO using flat beams with the way Dorico is stemming this by default:

This looks much better flipped:

I’m in agreement with @Jeremy_Aknai that Dorico’s default stem direction seems like the wrong side in this particular example.

1 Like

Erm… judging from those examples it appears that the wonderful Ms Gould entirely agrees with the point I made…

In the upper example the most extreme notes above or below the middle of the stave in each beamed group are equidistant, so rightly the stem direction is determined by the position of the majority. The lower example is more akin to my example. I’ve just tried and Dorico does get that one right, so I’m not sure what’s going on, because I’ve found many it gets wrong (IMHO of course).

The exception is when a minority of notes are much further from the middle stave-line than the majority.

It’s a case of what you define as “much further”. If you take your example and change the final note from a D to a middle C, Dorico automatically flips the beam/stems.

^^^ that’s how Jeremy’s example opens for me, just to make sure we are all seeing the same thing.

Isn’t Dorico doing this wrong either way you look at it? Gould pg 22, basic rule is “The outer notes of the group determine the beam direction … this is regardless of the direction of the majority of the notes.” Using that advice the D determines the direction and stems should go up. Using your quote from her above, the D is the “much further” minority exception so “it is better to have a majority of shorter stems in the ‘wrong’ direction” which means it’s ok for the C stems (the majority here) to be going the wrong direction. Dorico is doing it wrong whether you consider the D a “much further” exception or not.

The rule on pg 22 determines the beam direction, which is not the same thing as the stem direction.

Crap, I was skimming too fast. Stem rule is basically the same though. pg 24 “The note furthest from the centre of the stave … dictates the stem direction.” The D will still dictate the stem direction here, which should be up.

Read the small caps letters immediately above that rule. Are there an equal number of notes either side of the middle line?

Those are all equidistant though and that’s not the case with Jeremy’s example. Her examples have unequal numbers, but the notes furthest from the center line are all the same distance, hence another rule needed to determine direction. That’s not the case with Jeremy’s example as his are not equidistant from the center.

I’ve had some odd stem direction issues in a Dorico project I’ve been working on, similar to Jeremy’s illustration. They don’t seem solvable within the Engraving options. In 20+ years of using Finale, I’ve never had these problems. The Dorico developers need to take a look and maybe offer additional options here.


+1. Here’s the Finale default (with Patterson Beams)

This is something I remarked on a number of years ago. The example I posted at the time shows, in my opinion, a number of stem-direction anomalies. These were defended as being a result of choosing ‘Use default direction’ instead of ‘Determine by context’ in Engraving Options/Notes/Stems, yet I feel that the argument that these stem directions are ‘correct’ or ‘consistent with my settings’ isn’t evidenced by practical experience. The very first bar just looks wrong despite the defence that this is the result of my settings. Same with bars 59, 60 and 68.

The question of beam slants is another tricky issue and I remember discussing this with Daniel back in 2016. I still question the presence of some of the settings in Engraving Options/Beams/Slants/Advanced Options. Forcing horizontal beams for groups with any repeated notes shouldn’t, in my opinion, be a setting at all, let alone the factory default.

Other strange settings have to do with when to treat beams as concave. An inner note at the same position as an outer note should force a horizontal beam. The factory default here (i.e. the first setting), again, looks wrong to me.

The second setting here seems strange to me, as well:

In good, modern editions, I can’t remember ever seeing beam angles like the ones in the unselected settings above.


It’s not your settings, and it’s not the context, it’s just wrong. If I just enter that one bar alone outside of any other possible context I get the same result:

The Bb should determine the direction as it is furthest from the center. The fact that there are two repeated notes is irrelevant for determining stem direction here. In fact, Kurt Stone gives a very similar example on pg 50:

Dorico is also oddly inconsistent here. Dorico has the stem direction correct in bar 16 of your example, where basically the same rules would apply.

The factory default gives this:


so yes, it is dependent on your settings (or Vaughn’s).

If you switch the setting from Determine by context to Use default direction (and leave the setting underneath to Down Dorico sees middle line notes as always down. It’s no longer a wildcard that can switch dependent on context.

In this case, the two middle line notes would be downstem notes if unbeamed, and the Bb would be an upstem note if unbeamed. 2>1 so downstem wins. Yes, it’s wrong, but your settings are wrong for this situation.

Both settings are always going to give results that are sometimes wrong for some people, because the real world isn’t binary.

You could certainly argue that the algorithms need rethinking, or the available settings need more granularity, but the fact remains that the factory defaults get this particular example right (by which I mean the recent D-Bb-D, not the original post!). It strikes me that perhaps separate options are needed for whether notes are beamed or not, but maybe there’s some other gremlin in there that I’ve not thought through (and that alone wouldn’t fix the initial example in this thread).

@Vaughan_Schlepp I’m puzzled about the factory defaults you’ve mentioned. All the options you screenshotted - the highlighted blue ones - are the factory defaults.

Here are a few examples: then

The note with the * is the note that determines the stem direction in each. The note on the middle line has absolutely nothing to do with the stem direction in any of these examples. If Dorico is erroneously using the middle line setting in these cases, then that’s either a bug or just plain wrong. There is no support for using the note on the middle line in these examples as a determining factor in any notation guide I’ve looked through.

With “Default Direction” Dorico uses all three notes to determine the stem direction.
With “Default Direction” up, Dorico has no choice but to see the Ds as upstem notes.
With “Default Direction” down, Dorico has no choice but to see the Ds as downstem notes.

This is just maths. If you set Use default direction you’re tying Dorico’s arm behind its back. It just has to count the number of green arrows and red arrows. If red wins, the stems go up. If green wins, the stems go down.

So what settings would you suggest to get correct stem direction with notes on a middle line? This can’t be that hard as every other notation program can do it easily. I refuse to believe that Dorico simply can’t do this correctly, but maybe it can’t.

Finale default

MuseScore default

This isn’t really a complicated situation and there is no issue with correct stem directions in either of those programs. Also, the setting for Default stem direction for middle line really has to be Down. If it’s Up in the default then that’s a really archaic convention not supported by modern engraving.

Ross pg 83

Stone pg 49

Just leave the setting on the factory default, which is to determine by context.
It gets every single one of your examples correct.

Incidentally, Gould recommends determining by context. She’s a good 40 years later than Ross; I’m not sure when Stone dates from.