View Single Post
Old 03-19-2016, 08:04 PM   #74
ijijn
Human being with feelings
 
ijijn's Avatar
 
Join Date: Apr 2012
Location: Christchurch, New Zealand
Posts: 482
Default

Okay team, here's a quick update...

Regarding how things stand in the notation camp regarding flexibility and extensibility, the good news is that the per-voice/voiceGroup staff display option is basically doable without changing anything under the hood: it's essentially a visual re-shuffling of the data and could be added later when the time is ripe. Such a view would be necessary before I could personally consider using the notation editor in a big way, and I would be hard pressed to overstate how keen I am to do that.

The reason for my reluctance is because even a modestly busy and fairly well-behaved multi-channel track gives the misleading surface impression of a Xenakis masterwork, largely on account of the fairly wide note ranges and rhythmic independence: here's a sample track with 10 (primarily monophonic) channels on the go, and even after assigning voices it still looks rather scribbly (there are really only 2 voices per staff to go around in terms of the visuals)...



whereas being able to drill down into channel/voice mode and explode these lines into their own staves without having to resort to separate tracks () would be simply glorious.


The slightly unfortunate news is that the clef system and general staff-ness properties would most likely need a slight overhaul to facilitate various scenarios (I outlined some possible solutions in the discussion of voice groups) and some of the various notation text events would require a subtly different (or at least elaborated) internal structure to mirror their context: a ppp marking on a staff that represents voices 5-6 of channel 1, for instance, should reflect as much in its parameters, so that it doesn't bleed into any of the other data streams and muddy the waters.

A simple NOTA dynamic ppp <channel/s> <voice/s> ypos... would do it, or NOTE <channel> <pitch> dynamic ppp... for a note-snapped event, and something like -1 could work as a placeholder meaning "all" whenever there is a genuinely global event that needs it. Then it would just be a question of slotting those values into the mix based on context. Plus other largely cosmetic but very useful things such as creating and storing a staff name in the absence of using the track name, probably via a simple text box.

Bonus points are awarded for implementing linked and unlinked options within a group, which would immediately allow dynamic separation within the staff, so you could have pp for "voice 1" and mf for "voice 2", for example.


I'm still a little at a loss as to why, in the current implementation, voices are not being tapped for their incredible potential as sub-channel properties but rather as another sub-track paradigm, which is what channels are anyway, don't you think?

Separating lines visually by channel would have made just as much sense in the absence of voices, so voices don't actually add anything meaningful to the mix as it stands aside from specificity of up/down-ness in appearance. These channels could have spun out into more staves when their number increased past 2, 4, etc. with odd channels stem up and even channels stem down. By comparison, the use of voices in the current setup limits the options (only 2-3 rather than 16) more than it contributes to them.

I don't actually suggest that we use channels for this purpose, but I do suggest that we carefully consider the role of voices as yet another resource in the musical chain. There is a missed opportunity in the hierarchy here: if we think of voices within channels within tracks then our creative options are widened immensely and there is a less confusing sense of how everything fits together.

I'm holding out hope that this can be addressed at some point before the features are set in stone. It would be a terrible shame if such an amazingly powerful tool were not to see the light of day.
ijijn is offline   Reply With Quote