This is all sounding very exciting!
One question I have which rather shows up how ignorant I am of music theory.
I can compose and orchestrate quite well in terms of the music itself - but, embarrassingly, I simply can’t get my head round the rules of spellings/enharmonics etc.
My music tends to jump around harmonically quite a bit so I rarely have need of an initial key signature.
Yeah, I just don’t get it.
So if presenting my scores to an orchestra rather than just mocking up myself I always have to hire someone to do that for me.
Other aspects of producing the score/engraving I’m ok at. It’s just the damn spelling!
I always wonder - and I know this is lazy of me - but if there seems to be such a definite system of rules in place for spellings, why can’t a notation program analyse the general key at a given moment, and the harmonic context of that given moment, and then spell things correctly?
I’m sure that’s easier said than done, but any chance Dorico has the facility to help me out in this regard?
Thanks very much.
Perhaps Dorico will actually be able to help you in this area, jamieboo.
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 our own implementation of that algorithm (with a few modifications, dare we say improvements) is implemented in Dorico’s engine, though it is not yet exposed in the user interface of the software.
A possible workflow for you would be for you to compose and orchestrate your music as you normally do, then run the spelling algorithm over the music, which would result in the enharmonic spellings of the notes being adjusted where necessary to follow the traditional conventions for such things.
However, in order to verify that Dorico is doing something sensible and to limit possible embarrassment when giving your music to musicians, it’s worth spending a bit of time getting your head around the conventions of pitch spelling. It’s always best not to have to rely completely on the computer to save your blushes, as the people who write the software are human and therefore fallible too!
Brilliant, Daniel!
Thanks for such a helpful reply.
Eager to see how things develop.
This is fascinating. I looked up David Meredith and read some of his pdf on algorithm PS13. I feel sorry for the engineer who had to implement that in Dorico because it’s quite complicated. I certainly didn’t understand it all upon my brief perusal.
Correct me if I’m wrong but it seems that PS13 is applied to a stretch of EXISTING music to determine the proper spelling of the notes. It relies on the many relationships it determines between the notes. However a music notation program also needs to spell on the fly, as users input notes by midi. With Sib and Finale we are very used to them having set ideas for spelling during note entry, given the key. That is, we’re used to having to flip the enharmonic from time to time as needed. Am I right that Dorico will work like Sib and Finale for note entry? Or will it have the advanced PS13 intelligence on the fly, during note input? I suppose it could even back track and correct earlier spellings as a result of later input?
You’re right that PS13 really has to be run on a stretch of music in order to do its thing, so it would be ideally suited for e.g. selecting a passage of music and running a “fix enharmonics” command, or being employed automatically during MIDI import or real-time recording. As I understand it, it can’t really be used during step-time input as it needs both future context as well as past context to work out the correct spelling.
So for regular step-time input from a MIDI keyboard, you’ll suffer the slings and arrows of our default enharmonic spelling, with the option of fixing it up via PS13 afterwards, or fixing it up on a case-by-case basis by respelling as you go along. We also plan (though whether this makes it into v1.0 remains to be seen) to allow you to set up a pitch spelling table for a given key so that you can directly specify the default pitch spellings that should result when you hit a certain key on your MIDI keyboard.
Ooh, neat! Can this also be extended to forcing accidentals on certain notes when using mouse for note entry? For example, If I had a key signature of 3 flats but always want F# when I click on F? (This would obviously not be needed with a MIDI keyboard since I would hit F#.)
No, you wouldn’t automatically get a sharp when you click on, say, the F space in the treble clef: you’d still have to specify that you wanted a sharp. Spelling tables help with MIDI input, not with mouse input.
I hope this means that composers writing in (e.g.) C# minor won’t be able to get away with using Cnat instead of B#… 
[A particular bugbear of mine]
I made a test once, comparing Sibelius with my DAW (Logic), recording 12 simple chords to check if Sibelius got them right without editing, and it worked fine. Does Sibelius use that PS12 algorithm as well? In that case, I’m confident that Dorico will be fine as well.
One of the drawbacks about starting the composing process in a program like Logic, is that once one opens the score editor in order to tweak something, the accidentals doesn’t only look plain wrong, but due to how the interpretation of accidentals etc is set up, one can’t really trust that all the played notes are shown (see image).
you’ll suffer the slings and arrows of our default enharmonic spelling, with the option of fixing it up via PS13 afterwards, or fixing it up on a case-by-case basis by respelling as you go along. We also plan (though whether this makes it into v1.0 remains to be seen) to allow you to set up a pitch spelling table for a given key so that you can directly specify the default pitch spellings that should result when you hit a certain key on your MIDI keyboard.
The problem with such a table is that if I’m in the key of C, and play an E major chord in root position just before an A minor, I want the G# to be shown as a G#, and not as an Ab, because the triad won’t have the shape of a triad with an Ab in it. In other situations, still in C major, one certainly wants that note (Ab/C#) to look like an Ab, which makes me wonder if it’s possible to do something in Dorico which ensures that the PS13 always is run (on the recorded MIDI Notes) once one stops a MIDI recording? It looks top me as if Sibelius is doing something like that, since it recognises chords and give them the right accidentals without any extra activity on the user’s side?
Well, I’m not sure every user of Sibelius would agree with your assessment that it always gives chords the right accidentals without any extra activity on the user’s side! But yes, as far as I know Sibelius does try to choose the simplest spelling for a given set of chords, with reference to the prevailing key signature.
Dorico will certainly do its best to produce a useable set of spellings given a combination of pitches and the prevailing key, but I think the chances of it being able to do so in such a way that produces the spelling the user has in his or her head 100% of the time are essentially nil!
I’ll add a question to this thread instead of opening a new topic.
In a 4/4 bar, the first chord is a D flat minor chord. From the third beat, the chord is C sharp minor, which in this case is correct because of how the harmony will proceed afterwards.
The tenor trombone has the third through the first three beats of the bar: F flat on the first two beats, e natural on the third beat and some movement on the fourth bar.
The F flat and the E natural should be tied …
When Dorico tries to help me, it creates a dotted half note F flat, instead of what I want: a half note F flat tied to a quarter note E natural.
Is there a way to tell Dorico to be just slightly less helpful in such cases?
You should be able to do this by using the Force Durations option (shortcut O, icon that looks like a G clamp) to type in each note, then outside of note input, select each note with Ctrl+click and hit T. It’s certainly possible to tie notes with different enharmonic spellings, and this should do it.
Sorry to nag, but:
• I activate Force duration
• I create an F flat
• followed by an E natural
• I select both notes with Cmd+shift
• hit T
… which sadly alters the E natural into the F flat from which it is tied over.
Possibly of interest: in this case, contrary to my previous example, the F flat is to the left and the E natural to the right of the bar line.
I’m obviously missing something.
This really does work OK for me. I know you can’t see what key I press, but I promise I press T when the tie appears!
There must be something I’m missing. I’ll keep working on it. Thanks anyway!
I wonder, does any notation program follow the simple rule of spelling accidentals in major-minor chromatic scales?
Major chromatic scales:
up - III, VI remain unchanged
down - I, V remain unganged
Minor:
up, down - I, V remain unchanged
I created a few scales just to test it. What kind of rule does Dorico use?
Hi,
In reference above to changing incorrect or unwanted enharmonic spellings, Daniel at Steinberg wrote:
“you’ll suffer the slings and arrows of our default enharmonic spelling, with the option of fixing it up via PS13 afterwards, or fixing it up on a case-by-case basis by respelling as you go along.”
If you might be able to help me out here, because as of yet I’ve been unable to figure out how to get Dorico to do this. I’m sure there is just something I’m missing, but I’m just trying to change a G# in a chord to what should be an A flat. How would I go about this. Nothing I have tried has managed it yet. Thanks!
Select the G#, then type Alt+=, which is the shortcut for Write > Respell > Respell Using Note Above.
Daniel, I’m not getting the same result as you when I try to tie a double sharp note with its enharmonic one. What is happening here?
Thanks!
Albert.
