(Have searched the forum and the docs; may well have missed this - if so, my apologies.)
Can Dorico be made intelligently to respell enharmonically by algorithm; or is it all best done manually?
Take the case when writing in a ‘contrived’ key such as a whole tone; or completely atonally, say.
It is preferable to spell an interval initially entered as D♭to A as C# to A - if for no other reason than that most players will be more used to seeing C# and A on the staff together than D♭and A.
The place I could find which comes closest to having Dorico change such a D♭to C# appeared to be in Write > View > Notation Options > Accidentals > Transposition > ‘Prefer simpler accidentals’.
But that isn’t really what’s needed, I don’t think; I’m looking for a way for Doricoautomatically to ‘detect’ the more musically legible alternative… to prefer, in short, Major/minor intervals over Augmented/diminished ones.
Advice on where to find this, and/or how to achieve it consistently gratefully received, please .
I think as things stand there’s no option but to do things manually. The only caveat is that Dorico sometimes changes its mind about a previous note as you play in the next note(s) so don’t stop to correct each one as you’re inputting.
The option you’ve found in Notation Options relates to automated transpositions from Concert Pitch to Transposing, and is, as you’ve suggested, irrelevant.
the only time I make use of the algorithm when composing is when I pipe MIDI directly into Dorico from somewhere else, such as OpenMusic — and in those cases, I find that the algorithm very much makes the right calls as far as horizontal reading is concerned. It’s also the case with a MIDI keyboard, of course, but, when composing, (my) actions on the keyboard tend to be much more deliberate and it’s unlikely I would let it slip or forget to correct these instances at any point.
Because otherwise, when else would it really spring into action? I understand you (and many more) use Dorico and other softwares as a pedagogical tool, but, if the music’s coming from the composer’s hand, it’s more likely to be correct than anything, since the decision is taken from within the context and compositional framework and is (should be) inseparable from it. Even if the algorithm is refined to take vertical cues (and it has, as far as I recall), it won’t be able to “deduce” (I really dislike the term like this) the context that validates one choice over the other. Is it tonal music? Is it atonal — in the historically well-defined sense? Is it spectral? Is it something else?
I find these are matters of compositional framework and social convention. The tools merely try to keep up, and they are not always able to; they should not be setting the pace, I think. It is much more preferable to ask a fellow composer, a teacher, the players, other scores. I can assure you you won’t be bothering anyone!
I’m asking this in the context of primary composing (not importing or teaching). But I am very much less advanced than you
I rather like the idea of going through manually and respelling - say when I’ve used a ‘contrived key’ to begin with and either change my mind and/or remove it for Dorico to put in all the accidentals - again, for legibility and out of respect for the eventual players.
But in the same way that I learn from the many other things which Dorico ‘does for me’ (beaming, barring, ties in unfamiliar key signatures etc), I do like the idea of having immediate feedback from the experts in front of my eyes as I work.
I know, as well, that there are many circumstances when it really comes down to judgement and experience - when both the note(s) before and after a candidate for such respelling have a legitimate claim on it.
I think a big difficulty with the software telling you what’s “right” is that, with the exception of tuned percussionists, fretted instrumentalists and keyboard players, intonation is a relative thing. Real string/wind/brass players and singers tune G#s differently from A flats! Without going into too much detail, sharp keys ideally sound brighter than flat keys.
Just as with beaming, sometimes the software can’t read your mind, though it tries its best.
My second post was certainly in that vein, but an afterthought really.
I’m aware that (for example) Finale has the capability to be told to favour sharps, or favour sharps, and there’s certainly value in that.
Dorico DOES now have a Filter Sharp Notes function and a Filter Flat Notes function, both of which can be used in conjunction with the Respell commands - a slower method of achieving roughly the same result. As far as I’m aware the functions to filter double-flats and double-sharps don’t (yet) exist.
Yes. Certainly to favour either accidental regardless, and without due caution, would be asking for trouble.
Though something more subtle might be useful: to slightly favour either one when it was going to make the (melodic) interval more easily recognizable to the player.
Leaving aside wariness about B# and F♭anywhere (although I know there is a place for each of those), and at the same time working happily and judiciously with your point about Just Intonation, the kind of circumstance which I’m thinking of is where I have scored, say, a D♮and want to raise it when my KS (explicit or implied) is B♭Major or G minor.
In that case I think I’d like to have Dorico guide me towards E♭ and away from D# - particularly if it’s preceded by a C, say, and followed by a G because the ‘thirdness’ of those thirds is much more apparent.
I have a feeling that if you use L (lock rhythm) and then retype the note on a MIDI keyboard Dorico will normally give you the right note (Eb in your example). I wouldn’t want to bet on it, though.
Does it make a difference, really?
My supposition (and it really is just that; a guess) is that there’s some clever code that knows what key signature is printed and does its best to match that. If there isn’t a key signature then it’s less reliable.
Whether this is the case or not, there’s nothing we users can currently do to influence the software.
TLDR; Dorico’s good at this but it sometimes gets it wrong.
An academic researcher by the name of David Meredith developed an algorithm called PS13 that calculates the most appropriate pitch spellings for tonal music, even without reference to a key signature, and that algorithm (with a few modifications, dare we say improvements) > has been implemented into Dorico’s engine> , though it is not yet exposed in the user interface of the software.
Do we know how it is used; or is it simply the logic behind all/any/some respelling?
Using the transpose option in write mode and transposing a whole note up from D# minor gets me to E# minor. I have no idea how to convince Dorico of the advantages of F minor…
No worries. On first glance the system seems to be needlessly complicated, but composers sometimes do weird things and we do need there to be a way to get there: for example the slow movement of Delius’s posthumous Violin Sonata starts in D# minor and then has a section in D# major. I don’t advise it, but it’s totally legitimate…
I’ve learnt a lot by doing this manually recently; I started with a few obvious rules from page 85 of ‘Behind Bars’ and have almost doubled them to make a nice little list.
It’s also heartening that choosing spellings (especially in post-Tonal music) is as much an art as a science. It seems as though to have Dorico do it all for one would spoil the fun .
.