Tranposing instruments and microtones

First of all, I’m beyond stoked on how well Dorico now does notating and playing back microtonal music–it’s absolutely game changing for me and my personal creativity and output have skyrocketed. A critical addition for me was getting Noteperformer–ALL the instruments respond to non. vib markings and it sounds fantastic–most of my day got sucked into writing for full 18th century orchestra, never trying to write for one before in my life, because it sounds fantastic. I took the exceptionally tedious time to enter every accidental in the Sagittal Athenian set defined in 1200 tone equal division fo the octave, which is nearly a hundred accidental types, but to my surprise my workflow is still plenty fast and playback seems entirely accurate.

A bug I just noticed is that Dorico handles microtonal accidentals in transposing scores incorrectly. I decided to check out what performance scores of my admittedly very unlikely to get performed extremely microtonal work would look like, and it definitely resulted in pretty crazy accidentals for the Bb trumpet and F horn. For example, the first note in the horn is an unaccented G, but in the transposing score it displays as a Ebb minus a syntonic comma for whatever reason, which is almost but not quite the D it should’ve displayed as. They were all accidentals as defined by

I was curious to see if it was simply shifting transposing instruments down a fixed interval in 12EDO, but it seems like transposing instruments do play accurately. I set up a score in 13EDO with a clarinet and piano and had them play in unison and they were spot on. In a short section it seemed like Dorico was generating sane accidentals for the clarinet part when given the relatively few options of 13EDO.

Welcome to the forum, Cody. We have discussed this issue at some length with a number of microtonalists over recent months. There doesn’t appear to be consensus within the community about what the appropriate accidental is. Many tonality systems do not have a consistent number of divisions between each diatonic step in the octave, which has the consequence that when a note is written for a transposing instrument, it may not be possible to write the note with the same accidental at the written pitch of the instrument. Dorico chooses the accidental that produces the note closest to the specified sounding pitch in the concert score.

It’s not really clear what else it could do, but we’re open to suggestions about what you would like it to do!

Thank you for the reply Daniel. It’s really clear that development is really reflecting user feedback in a direct way.

Yeah, that makes sense. In this case it’s coming up because I have such high precision accidentals that it’s registering the difference between a just major second and a tempered one, and some hilarious diminished third minus a comma is actually closer to ET than an acoustically simpler 9:8. I suppose ultimately you’d need to be able to define what number of divisions the instrument actually transposes at. I just tested writing a semi-complex line in 72EDO with a clarinet and everything transposes correctly because its major second is literally the same as 12EDO. Ultimately not a big deal–if, in a miracle, I actually have to have anyone read my scores on a transposing instrument, I’m sure I can take the time to edit the score by hand.

But seriously Dorico is absolutely amazing for microtonal composers and I deeply appreciate the time spent to get things working correctly. From a practical standpoint, the main thing that could improve workflow would be being able to define key shortcuts to the custom accidentals, although I’ve gotten pretty fast at note input regardless.

I certainly agree that being able to access your custom accidentals via shortcut is important. My idea is that we could have a popover for this, allowing you to type something that identifies the accidental – perhaps it could provide an auto-completer for the names of the accidentals you’ve defined in the tonality system, or allow you to type the number of divisions you want to raise or lower the written note by.

Is there any likelihood (in the future) that Dorico would support definite pitches for the Major scale and accidentals that rationally alter these, as in Ben Johnston’s JI notation?

Our general approach to this has been to say no, and that this kind of tuning system is better handled through the use of VST instruments that can use e.g. Scala files. This is why we designed Dorico’s tonality system features around EDO-based systems, rather than around scale-dependent tuning systems.

A suggestion for steveparker:

Dorico’s EDO based system isn’t perfect for realizing Johnston notation, but it can be pretty good. Last month I whipped up a 224EDO that did a very good job of apprixmating JI Mixed Sagittal Athenian notation, with the major scale note spacings being 38-38-17-38-38-38-17 making a pretty good approximation of the Pythagorean scale. Johnston’s 5-limit scale, if I do my math correctly for my particular setup wouldn’t be much different (and would of course include different glpyhs for Johnston accidentals). The 224EDO intervals should be something like 38-34-21-38-34-38-21.

Some other microtonalists, from what I have seen, have gone with much larger EDOs like 1200 and then just picked out the divisions they’re interested in. Whatever your resolution, I think it’s very useful to be able to define each written scale degree with a different number of steps like I have done with my 224EDO. With sufficient resolution (whatever that is for your purposes) you can get the similar comma motion quirks that you get in proper JI. For example, in my 224EDO, the syntonic comma is 4 steps, which is the reason why you see that 4-step difference between my suggestion for Johnston’s 5-limit scale and the Pythagorean that I used.

I’ve used scala files in Finale with accidentals defining pitch bends. It’s painstaking and accidentals won’t stack (so I had hundreds of compound accidentals defined), and almost impossible with more than one voice on a stave. It is philosophically horrible to approximate this stuff with an ET and is a shame that it’s not planned for Dorico, which is probably the ideal app for it. I realise JI is a minority occupation…

I’m really learning a lot from using the Helmholtz-Ellis Just Intonation accidentals, with a font and example score from It’s based on a 12000 EDO, so the resolution is pretty great. I’ve made probably around 150 compound accidentals, since they don’t stack; but you can specify the exact pitch deviation for each accidental, as well as tweaking their spacing etc. When you click on different notes in your score, the corresponding accidental lights up in the right-hand panel, so you can tell at a glance if one note is higher than another.

Daniel: The popover idea sounds solid. I was thinking of getting an external 10 key and setting it up for the most commonly used of my accidentals if shortcuts could be definable.

steveparker: To do Johnston notation, define the EDO as 224EDO (smallest that is likely to be perceived as just), 1200 or 12000(what I’ve done, with pure Sagittal). You’ll have to do some math for 224, otherwise with 1200 or 12000 EDO you’re just giving values in cents or 10ths of a cent. From there you can define the steps between nominals as successive steps in cents of the just major scale. You can combine multiple symbols in a single accidental to get compound accidentals as used by Johnston and define them as offsets in cents/fractions of a cent. You will have to spend more time setting up various accidental combinations, but it’s not so bad. I just set up the entire Athenian set in Sagittal and it’s just under a hundred accidentals and covers most possibilities. If you need an accidental rarely that exceeds whatever base set of accidentals you need, it’s probably rare enough that you can just add it as needed.

More or less my process in Finale with Kontakt… I was hoping that Dorico would (eventually) do this differently, by allowing a defined tuning for a just major scale, then allowing accidentals to be cumulative, and apply separately to different notes in the same voice. Otherwise I’m recreating what is already done in Finale.

I appreciate all replies though…!

The cumulative won’t work in Dorco, but set up as suggested it plays a just (to within 1/10th cent) major scale on the naturals. There is a practical limitation on resolution in any context and I’m not fretting that my JI is approximated by 12000EDO because there is no possible chance under the most extreme circumstances that I could perceive the difference.

I’m starting to think about transposing instruments as well, and looking back through this thread, I see this suggestion:


I’m wondering what the lack of consensus amongst the microtonal community is, specific to Dorico’s setup. It makes sense to me to take Cody’s suggestion as a pretty good one. Doing a quick think through what I see, changing the ‘Root’ of a custom key signature from an option of A-G to the specific number of divisions should actually simplify the system in some sense, because A-G are already specified numbers of divisions per the Divisions setup. An alternate idea, for convenience for us users, is to perhaps see that ‘root’ selection split into two: the ‘white’ note of A-G and then an additional deviation of divisions - but I think this is really just cosmetic, as these two selections would just be immediately added together in the algorithm anyway. If I’m in the right direction of my deconstruction of how this works, at least. This should get rid of the idea of having Dorico pick a ‘closest’ note as previously mentioned, which I figure is particularly problematic in the case of ties.

Also: the ‘Custom Key Signature’ column should probably auto-sort the key signatures by the root like the ‘Accidentals’ column does.