NOTATIO

MJCube wrote:Fascinating! This never occurred to me before, and the optical illusion issue is obvious on close inspection. I suppose it might be relatively simple for an application to align a single sharp in the Durand style, above. But what about chords, mixed accidentals, etc.? How can this be codified for more complex cases?

It the case of chords, it seems that the accidentals aren't shifted at all. I have only observed this with sharps on single notes, at least. If you by 'mixed accidentals' mean triple sharps or the obsolete practice of cancelling accidentals, I haven't seen these in Durand's scores at all, but I would think the accidental closest to the note would govern how they are treated.

In terms of how an application might deal with this if you in fact wanted such shifts on chords, however, SMuFL (the new standard for music font layout supported by Sternberg's upcoming application) would probably apply metadata to the relevant glyphs, indicating the position of the optical centre. From there, the application could potentially adjust the accidental vertically according to this metadata.