Just in case, here are two files. The first uses the Aria Player which is set to quarter comma mean-tone with D being the reference note. If you don’t have Aria Player installed, it’ll probably revert to equal temperament. The other file uses NotePerformer but I haven’t figured out how to tweak the tuning.
Silver Swan Aria Player.dorico (426.6 KB) Silver Swan NotePerformer.dorico (422.6 KB)
I fear Dorico delegates the choice of the scale to the sound player. And while many of them can select a scale other than Equal Tuning, I can’t see something like this in NotePerformer.
Paolo
That’s what I was afraid of.
While in the Edit Tonality System dialog, you have to think in cents. So, while the divisions between A and B is ‘2’ in the 12-EDO Equal Tuning, it may be ‘200’ in a 1200-EDO Meantone system. All of them adjusted according to the intervals in the scale.
Then, when editing the accidentals, you can change the pitch delta of the sharp and flat to values around ‘±100’, depending on how many cents they will change the pitch in a meantone system (or any other alternative tuning).
I wonder if you can have flats and sharps changing a note of a fixed amount of cents, independent of the scale degree. Can they?
Another way could be using a 31-EDO division, and use smaller numbers.
Paolo
(First, I say this because I struggle with this kind of realism)
Weirdly I wonder if the problem with a realistic orchestral sound is that the vendors treat it as an orchestral sound… What I mean is, most synthesizers have sophisticated matrices, where all manner of performance aspects can be linked with what is effectively a function or formula, which saves a lot of note by note editing. Its not about the effort to me as much as speed - I’m willing to do a lot of work if in the end, I can meet both quality and delivery schedules.
In the meantime, I tried to create a 1/4 comma Meantone scale, starting from D. Some small adjustment was needed due to rounding errors, so anybody else can retune it by taste.
Since you can’t edit a scale, and then indicate a root note, it is a matter of compensating for each grade of the scale. Having the root and the scale separated would be very useful.
Paolo
Meantone tuning is only intended for keyboards, a vocal (or brass) ensemble performing Gibbons should aim for pure intonation, with its consequent opportunities for pitch slip.
By the way, I am not convinced that the G sharp in the top part should not be G natural.
Like the Tannhäuser Overture, this realisation also seems to use circular breathing throughout!
David
1 Like
Vaughan, I took the freedom to play your wonderful Gibbons with VSL’s Dimension Brass, and experiment my meantone scale. Is it working?
EDIT: After comparing various files, I think it is not working. The first example is with the meantone scale in Dorico:
Gibbons, The Silver Swan, VSL VI, Meantone (apparently not working)
The second one with the meantone scale set in the Vienna Instruments player:
Gibbons, The Silver Swan, VSL VI, Meantone set in the VI player
(EDIT: With the first Dorico/VSL example I felt I could hear the different tuning, but I’m not even sure Dorico can transmit it to the VSL player. But now I’m nearly sure it is not working. And I left the pitch at A=440, since I’m not sure of the recommended tuning for Gibbons.)
EDIT: Added two tests with NotePerformer, with my meantone scale and equal tuning. Here, I’m sure I can hear a difference.
Gibbons, The Silver Swan, NP, Meantone
Gibbons, The Silver Swan, NP, Equal Tuning
Paolo
1 Like
Thank you for your efforts, Paulo! I also feel the tuning to be different but it’s hard to pinpoint. Some of the thirds don’t seem to be as pure as what I’m used to hearing, but the overall effect is very good. I also miss the characteristically mean-tone sound of the augmented chord at 30" and 53". There’s nothing like two stacked pure thirds and the ensuing severely detuned diminished fourth!
Thank you also for the table above and for the library file. I’m afraid I don’t know how to import it. I’ll also do some more experimentation.
There is a dedicated command in the panel:
I’ll try to analyze one of the instrumental parts to see what is happening to the tuning.
Paolo
I’ve generated the Horn 2 track alone with MT and ET tuning. Then, I’ve inverted the phase of one of them, and summed them.
Unless there is some other phenomenon happening (lack of sync in the audio programs), the volume of them playing together is not kept to zero, but heavily reduced. My hint is that the two files are similar (hence the volume reduced), but not identical (therefore, they are not cancelling).
EDIT: The VSL test includes computer-generated reverb. This is not always the same, therefore producing different audio files.
Paolo
Above, I’ve added two tests with NotePerformer on Gibbons, and one with meantone set in the VSL player. Custom tuning seems to work with NP, probably not with VSL.
Paolo
Glad to see more people concerned with virtual instruments not tuning together! 
Even if Dorico doesn’t yet support that kind of automatic tuning that you’re referring to, I must say that the intonation system feature can actually provide a not completely terrible way of achieving it.
I recently took a Palestrina counterpoint class where the final assignment was to compose a motet. In order to produce a pure tuned audio file for my perfect pitch teacher (just for our own enjoyment, not required for an approved grade), I input the music into a custom key signature featuring 5-limit tuning and hid the accidentals before creating the final PDF.
Here is the result using the wagner tuba sound from NotePerformer:
A blessing and a curse with NotePerformer somewhat present here is the pitch humanization, that is sometimes throwing all the hard work down the drain, so I had to render the file a couple of times before getting an acceptable enough result.
Btw, the level of nerdiness of this discussion is just divine. 
2 Likes
You’re right, meantone is only necessary for instruments where the tuning has to be determined in advance and can’t be altered on the fly. But, as with so many things, tuning keyboard instruments to something besides equal temperament can have an important added value, if you want to call it that. If not all tonalities sound the same, then when the music modulates to a different key, it really has an affect on the mood other than just pitch. This can be a very expressive musical device when used consciously. There may be many reasons why a composer might choose a particular tonality for a piece, but the fact remains that composers coupled tonalities to specific character associations. This was even codified in the Baroque but even Plato writes about it. Equal temperament removes these differences.
BTW, if you’re referring to the G#s at the beginning of bars 10 and 17 (or in the sound files at 30" and 53"), there’s no way that could be a half tone lower. Try it out on one of the files!
Very nice, Frigolito, and congratulations on writing a motet a la Palestrina! To bad we have to work with invisible accidentals in NotePerformer, though.
Isn’t this fun? And as Frigolito wrote, delightfully nerdy!
Unfortunately it doesn’t seem to work in NotePerformer, although your version using the Vienna Instruments Player does work.
Absolutely. One important step in playback realism would be to permit an alternate tuning system for keyboard instruments relative to the rest of the ensemble. In baroque music, you might have singers and instrumentalists making pure thirds whilst the continuo player (harpsichord or organ) will be simultaneously contributing a tempered scale which will naturally result in an ‘affection’ on the overall sound.
Yes! That’s exactly what happens in real life. The problem is, how would you achieve this using a notation program or a DAW? Even if the program could accept different tunings for different instruments, you’d still have to make separate, dedicated continuo realisations for each type of keyboard instrument, as there’s no way a computer can approach what a good continuo player does, at least, not at the moment.
Well, considering the advancements in AI technology, I don’t think we are that far away. I’m especially thinking of the OpenAI Jukebox project. This is for instance an attempt where the AI is trying to continue the beginning of Mozart’s A la turca in the style of Chopin:
Of course, this OpenAI Jukebox project includes a whole range of insanely impressive (and slightly creepy) outcomes which have aspect that could be more or less directly transferred to virtual instruments technology.
1 Like
Creepy is right. This is fascinating and surprisingly legit (first half anyway) even if their piano sound vst is crap.