Had an idea for a feature… when orchestrating, it’s all too easy to make orchestration blunders such as minor 9ths (yuck), or lower notes too closely spaced (muddiness) and other such orchestration rules.
It’d be neat to have a mode where such warnings would be coloured on the score, automating the manual analysis.
Wouldn’t that depend highly on style, genre, and personal preferences and proficiency? The things you mention aren’t orchestration errors in many cases. What’s wrong with a minor 9th? It’s a harmony that occurs in music. If it’s simply a wrong note, you should hear it on playback, and be able to correct it.
The muddiness of a low chord may be quite intentional in some circumstances, and also something you’d easily hear (and see) yourself.
Of course, there are lots of things you can do wrong in orchestration, but I’m afraid it’ll be very hard to teach a notation app what to avoid, and when.
Should Dorico tell us ‘this flute solo is too low to be heard in this texture’, or ‘this double stop won’t work in this tempo’, or ‘this note can’t be played pp by a non-professional player’, or ‘insufficient time for this instrument change’, etc.? How would Dorico even know that, or decide to point it out?
Ich verstehe Ihre Wünsche nicht ; tut mir leid.
All of the above information is, in some capacity, available to make those decisions. I am in no way implying Dorico should do it. I see this as an analysis which could be done either as a separate program that analyzes MIDI (perhaps with an audio rendering for spectral analysis) or, when (if) Dorico gets scripting / API stabilized, as a 3rd-party plugin.
I’m thinking of it like reverb profiles:
It’s actually a pretty cool thought experiment for a musically inclined developer (or a musician who develops software).
Score writing software such as Dorico primarily helps us to write music on the page properly, and secondarily renders the notation reasonably in audio.
In my view, assistance with composition, counterpoint and orchestration (in short: taste) is not in the domain of score writing software.
It seems like there are 3 basic questions here:
For 1), I think pretty clearly all sorts of options, including various performance styles, could be done.
The second question is sort of the tricky one. I’m all for using AI and proofreading algorithms to help me with busy work and to catch mistakes, but it seems to me that 90% (at least) of writing really is the process of writing. If haven’t experimented with various themes, harmony, orchestration, etc., to find out what works and what doesn’t, and just relied on software to do it, then I wouldn’t really understand what’s on the page, and wouldn’t understand how to go about revising it or getting the best performance out of it. A forklift is way better at lifting weights than I am, but I don’t take a forklift with me to the gym. The goal there isn’t to physically move the weight, the goal is improvement by the process. The compositional process itself is important too and you lose a lot by simply bypassing it or relying on software.
For 3), that probably depends on whether the developers think it’s worth the time cost and whether it would be marketable. New versions need marketable features to attract people to upgrade. None of the things the OP posted would really be that interesting to me personally (I like minor 9ths and know how to use or avoid muddiness), but perhaps Steinberg’s market research says that it would be a popular feature. Or not. It definitely seems to be straying a bit from a dedicated emphasis on notation, although we already have features like Generate Chord Symbols and Generate Notes from Chord Symbols. We’ll have to wait and see what D6 brings!
the whole point in having decent libraries for playback is that we can now hear for ourselves as we go along what works and what doesn’t from an orchestration and harmonic viewpoint. These often arcane rules were devised at a time when the technology we have today wasn’t available. Of course what works for virtual instruments won’t always sound quite the same as with a live orchestra for obvious reasons but nevertheless the point remains. For the most part I agree with @PjotrB here. And if this thread gets completely out of control like the recent one on parallel 5ths and 8ths which I assume the OP didn’t see, then heaven help us…
Honestly, each time I hear mockups I am a bit less sure what the actual point of playback libraries is…
I agree with @PjotrB, I don’t view these as blunders but as stylistic choices. I personally love minor 9ths, and lower-note tightly-spaced clusters have their place too for certain moods and atmospheres (love the low brass moment from Mahler’s Symphony 2)! As for anyone wishing to avoid these, I think out of your examples lower notes closely spaced can be pretty obvious both in playback, but also if you learn even the basics of orchestration and where instrument ranges lie for transposing instruments in particular.
perhaps you should then listen to decent mockups with decent libraries
I would think these rules weren‘t meant to be arcane, but rather helpful and practical … but I agree with your post.
This seems like a possibility that nicely addresses @FredGUnn’s points 1 and 3, above. (Which, as @dko22 reminds us, was also an idea offered in the parallel-checker thread.)
Personally I like minor 9ths! There could be a benefit in this, but probably as an external piece of software or AI model that could act as an orchestration advisor.
ref: @FredGUnn “Should the developers spend development time on it?”
Third party/User scripting might be best, perhaps something like (on Mac for example) Keyboard Maestro where you make your own script for whatever purpose you have with various options already provided, select which instruments/singers, range, dynamics, what to look for, intervals, which section or all, etc. then search and mark.
e.g. for pitch, aurally you hear a something like a minor 9th which was not intended, it could check for notation and pitch bend together throughout the selected measures.
Then those who wish can create scripts and share them, similar to Expression maps (and there might be less comment on its purpose, although sometimes it is nice to know why just out of interest.)
I could imagine grading student scores/notations, with some script automations might be useful in a small but useful way for teachers for example, running it through some scripts they have created specifically, related to what they are assessing.
of course you’re right – I really just meant they might seem arcane on the surface until you understand the motivation behind them.
Wow, I hadn’t expected such enthusiasm for discussing this topic.
I absolutely agree with everyone saying that computers and AI should not be a crutch for practice and ear development and personal taste development.
But that wasn’t what I was intending with the original post. I was more thinking of such a tool like MS word spelling and grammar checker. Something that can highlight things you might have done by accident, and suggest fixes, or if you did them deliberately you can click ignore and accept as is. A musical proof reader if you will.
For instance, yes minor 9ths can be great! John Williams octatonic music is covered in “wrong note” basses and other pitch collection chords that create minor ninths. But also, minor ninths can be accidentally introduced inadvertantly by certain voicings of simple chords, rather than deliberately. And yes, good ears will pick that up, but why not have a helpful tool that is even faster and warns whilst writing not just on playback.
In my vision of a tool it’d have lots of configurability as well, to enable/disable validations or groups of validations, so it could be tailored to individual taste. After all, who wants to mimic crusty old courtiers in wigs from Vienna, eh?! We are all individuals!
And yes, seems I’ve missed any drama on a discussion about counterpoint and four point writing tooling. Something I also think would be useful, but absolutely not as a replacement for doing the learning and personal taste development, but more just as yet another tool in our toolbag.
And to those who say, “no. Tis a step too far! All this technology is ruining modern music!”. That seems a strange argument, why use Dorico at all, just use quill and parchment and your imagination, otherwise you’re cheating, and contributing to the downfall of music. You cannot simultaneously be a user and a decrier of technology, willing to set limits on the use of tech by others, to impose your own ideals upon others. I’d much rather this discussion be focussed on what features could be useful, not quasi-religious beliefs on why tech is Satan incarnate, but only when it comes to writing assistance.
I suspect many of us are wary rather than enthusiastic. My position is simple - the Development Team already has a huge to-do list of the replacing the quill and paper kind.
Mark_Johnson: I enthusiastically and wholeheartedly endorse your post. In other words, I “second the motion!”
From an educational point of view a sophisticated , style and era adapted, harmonic rules based filtering tool visualizing where the defined rules are broken in a larger score as an add on to Dorico would be interesting.
There are articles, presentations at dedicated symposia, Youtube videos and probably even books written about how classical and modern composers of renown break established rules of their era and create great new directions.
Finding such very interesting exceptions is a rather
tedious work which would be greatly facilitated by such a filter tool. The analysis could then focus on why such exceptions work and makes a piece more interesting.
While I understand the analogy to written language, I feel like music scores work both horizontally and vertically in ways that words and language do not. In other words, grammar and spelling can easily be checked and refined horizontally based on current (yet always evolving) language rules. Comparing to a database of rules is likely straightforward, spelling especially where the rules around that are generally more set in stone and agreed upon (of course, there will always be exceptions and slang).
Whereas music has far fewer rules in this regard, way more exceptions, operating on both horizontal and vertical planes of time. So while I don’t think it’s impossible, I do think it’s exponentially more complicated than spell checking some sentences, and always open to way more complex interpretations, styles, time periods, tastes, and contexts - which could ultimately result in a lot of “false positives” with such filtering (a suspension as one basic example).
I’m certainly of the camp which believes the best way to truly learn these things is to learn by ear and also visually recognizing them by experience. One, of course, short-changes and confuses this process with something that automates all that for them, which for me isn’t an issue of finger pointing and screaming “that’s cheating!” but more that, one is cheating themselves – and that’s their loss.
If we do return to the analogy of language (or basic math for good measure), we have found that people who rely heavily on checkers and calculators lack the fundamental abilities to execute these tasks without those, which again, I feel like that’s their loss more than anything.
To each their own! I think if a third party developer created scripts that could achieve this, I won’t be a Luddite over here with a picket sign. More power to anyone who wants that. But I would rather the small development team focus on many more pressing matters 