Funny that I came across this right after a concert (Bottesini, Elgar, Hindemith, Monti, Saint-Saens, and Dvorak -- all 1st movements of one of their respective concertos, except for the Trauermusik -- if you're wondering).

So to start off, it's not really a valid criticism to complain about Hummingbird's alleged emphasis on absolute pitches as opposed to relative pitches. Honestly, I can read the notes without giving a crap about note names. I mean, are the circle-names that distracting? The visual position of pitches is still consonant with traditional notation, so it's still maintaining the relative spacing you would expect. Additionally, the new accidentals are at least still on the left where you'd expect them to be.

As far as why this is better than the "teaching notation" [1] of today, it should be obvious when you zoom out a bit: the current system simply doesn't visually scale. That example's quite difficult to read without looking at staff lines. Reading a letter name from inside a note head is impossible unless you make everything "Fisher-Price"-sized (i.e. unmanageably huge). I mean, good luck reading this [2] at any distance if it had note names inside. Sure, by the time you get to the point to be able to play that sort of piece you shouldn't need note names, but if we're discussing notation, Hummingbird is definitely better than "teaching notation" in that respect.

On the other hand, we have some new issues that are introduced:

(1) Lack of stems or beams. As others have noted, this really is difficult to use to scan pieces quickly before/during sight-reading. They've missed the point with proportional spacing, because there are pieces with really horrendous spacing that are still actually readable. However, getting rid of beams makes thinking about rhythmic patterns harder, because now you have to look in two places to find rhythm grouping (in their version of "beams") and rhythm (with that squiggly crap under the "note head").

(2) Not to mention, this results in another readability issue when it comes to notes shorter than a quaver (8th-note): scaling. I cannot tell the difference between 16th and 32nd notes when I zoom out even a little; the curves all smoosh together. And WTF is with those rests? They're completely unreadable at the same zoom level. Traditional notation, however, makes things obvious because the flags are a clear indicator of the length of the note. In fact, I can zoom out 8 times farther in traditional notation and see note lengths EASILY.

(3) This brings up the issue of lower visual information density. I'm not buying that this will help non-visual learners. I mean, where's the auditory or tactile feedback here? Not to mention, this hampers visual learners too, because traditional notation manages to make all the visual cues large yet it still takes the same amount of space. Hummingbird depends too much on minimalism, and turns fat, thick beams into anemic lines, and heavy, bold accidentals into awkward little markings. I'm not saying that this is an issue that is confined to Hummingbird; badly typeset traditional scores, particularly computer-generated scores, suffer from this as well. I'm saying that at its best, Hummingbird is on the same level with respect to clarity as incompetent traditional typesetting.

(4) Why did they get rid of dotted notes? I mean, functionally, a half-note tied to a quarter-note is the same as a dotted half, but notationally, they are used differently. The former would be used in simple duple to represent 1 beat + 1/2 beat, 2 beats + 1 beat, etc. and the latter would function better in simple triple or compound meters, where it might represent a whole measure or a whole beat, respectively. Either way, this type of rhythmic grouping is absent from Hummingbird. I mean, there's a reason why most typesetters don't go around using triple-dotted half notes every time there's a dotted half tied to a dotted eighth; it's less readable.

(5) The abstraction in the note names is weird. If you're going to abstract note names, why not just use position? Why do you have to learn a system using circle-symbols that no one uses anyway?

(6) Yeah, that key sig thing is really a non-starter. Although I do have memory lapses regarding the key signature at moments, it's still important, for the same reason that knowing harmonic and melodic progressions are important, namely that you know how the notes function in relation to one another in the piece. If you want to scan a piece of music and write a harmony for it, or do anything to it besides just reading it, the key sig is almost a necessity.

(7) This might become a non-issue in future versions/revisions, but what about other notation for ornamentations and other musical aspects in general? Slurs (look a bit like rests, and violates every sense of notational consistency as a result), certain chords (if you want to typeset two adjacent notes, normally you just shift one note head to one side; Hummingbird makes that impossible), multiple voicings (how do you mark separate voices if you can't use stem direction b/c there are no stems to work with?), trills (not the "tr" mark, but the wavy line that sometimes follows or replaces it; looks a bit too much like a bunch of 8th notes tied together), tremolos (if there are no flags, how do you write a tremolo?), Bartok pizzicati (which look remarkably similar to 16th rests and sharp symbols), tuplets (how do you express a tuplet if there's no stem? Then the tuplet bracket looks like a beam, and the tuplet number might as well be a fingering), acciaccature/appogiature (there is no concept of note size, so notating grace notes is impossible), glissandi (they look like beams -- is this getting old?), and articulation marks (many of which are too close for comfort to random symbols that I won't bother enumerating here).

I mean this huge post is just scratching the surface of the fundamental issues with Hummingbird. By all means, find a way to improve notation, but if the intent is make reading note names easier, don't sacrifice literally everything else. If I was given complete control of Hummingbird, revision 2 would be identical to traditional notation, except for the note heads. Yeah, there would still be scalability issues, but the change wouldn't castrate the notation as a whole, it would allow a very smooth transition to traditional notation, and at the very least those circle things are ignorable by professionals. The final issue, then, is how do you differentiate half notes and quarter notes or tremolos... Ah never mind, it's just a mess already. And don't forget, we have a better two-century old version of this proposal already [3], but some shapes still look a bit too similar to one another...

TL;DR Hummingbird is equally fast to read, but is less information-dense and scalable, and invalidates many aspects and necessary facets of current notation, which all combined introduces unnecessary headache for everyone who's not a newb. Can't comment on behalf of the noobs, because it's been a while, and you can't really imagine what it's like to not be able to read notation when you actually can.

Also what guy designed those samples with GRAY staff lines? They blend into the white background waaay too easily, and makes Hummingbird harder to read. I mean, that traditional F-C-F quarter note triad looks like an F-(B? or C? or D?)-F triad in Hummingbird.

[1] http://lilypond.org/doc/v2.14/Documentation/f8/lily-13872c43... [2] http://lilypond.org/doc/v2.12/Documentation/user/e3/lily-bfe... [3] https://en.wikipedia.org/wiki/Shape_note

user:	plus9z
created:	April 11, 2013
karma:	32