Hacker News new | past | comments | ask | show | jobs | submit login
How I’d redesign piano sheet music (medium.com)
193 points by cyanbane on July 4, 2015 | hide | past | web | favorite | 159 comments

What happens if you rotate that first image 90° counter-clockwise? http://i.imgur.com/i2V57On.png You get a traditional sheet music with simpler graphics.

What about the new features?

  - Easily printable: traditional scores already are
  - Screen and scroll-friendly: you can stack traditional score pages vertically
  - Chord names: already exist on traditional scores
  - Lyrics: same thing
  - Foot pedal arrows: same thing
The only novelty here is to match a piano's physical horizontal pitch range, which forces the time axis to be vertical instead.

I guess sheet music is like code: sometimes you just need to learn it the hard way.

I challenge "easily printable." This way of writing sheet music is vastly more space inefficient. Each beat requires an entire row of space, whereas a beat in sheet music might require 3 square inches at most. Having many pages of sheet music, like in coding, makes it much harder to understand the song as a whole and navigate repeats.

I do like the fact that — unlike with sheet music - the 12 semi-tones are spaced evenly, as they are in reality. This allows for fast visual recognition of chord types and intervals.

As long as you're playing tonal music, chord types and intervals will be strongly associated with the 7 note names, not just the 12 semitones. A perfect fourth will always be the same size on the staff, for example.

If it's "spelled" differently -- if an interval of that size is actually written as C-E#, for example -- then that's a sign that something weird is going on and you're not supposed to think of this interval as a fourth.

The fact that we write music using the 7-note diatonic scale is kind of like compression: it optimizes notation for the notes and intervals you're most likely to play. It's just not suited for atonal music, much like compression is not suited for random-looking data.

I know little music theory, but I've been playing instruments "by ear" for a long time and it made many things easier to me when I learned that if you're using diatonic scale, it's very unlikely that you'll use notes outside of the scale; e.g. if you're playing a song in C Major, most of the time you'll play C and D, not C# or Db, so if you need that note, it's easier to switch than having an almost unused line in the sheet.

This is exactly what standard sheet music does. A "C" chord can be C-E-G, which makes one shape on a piano keyboard. It can also be "inverted" to E-G-C or G-C-E. These form distinct shapes and it takes about a week to learn to spot the patterns and play them. Organists even do it with both hands while playing notes with their feet.

I remain firm in my belief that if you can touch type you can sight read music. I've seen five year olds learn to do it and I've seen 90 year old people learn to do it.

Actually, the semitones aren't spaced evenly here. The white notes on a piano keyboard are spaced evenly, and then the black notes are half-way between the white ones. So it's rather like conventional notation in that respect (the notes of the diatonic scale are equally spaced, as long as you're in the right key) but since it doesn't have a notion of key signature you can't arrange for that to remain true in (say) F minor.

This preference for some keys over others is pretty much built into the piano keyboard too, and the notation is specifically intended for keyboard music, so it's not obviously crazy. But it certainly doesn't have the property that the semitones are equally spaced.

The 90° isn't the big problem with traditional notation. It's that the same note in different octaves looks different. It's as if in programming you wrote a for loop differently for every level of nesting. The new notation draws against the same pattern of staff lines for each octave; plus, you don't have to remember sharps and flats from the key.

I've long wished for these two improvements, especially the first one; OTOH I'd keep the traditional whole note, half note, etc., for precision and concision -- the piano-roll style feels more like training wheels.

Regarding remembering flats and sharps... That may be easy for beginners, but it's actually not good. You do have to remember the sharps and flats because you have to know in which key you are playing to see where is the music going. And when you're playing you end up not needing the flats in each note because they're always there. They end up being redundant, so by removing them you gain simplicity and you can see more clearly the flats and sharps that actually matter and change how the music sounds.

It would be natural (ha) to add a mark next to a note that doesn't belong to the key, or color it, etc.

Notes in different octaves are different and hence should look different. Even if you restrict to keyboard instruments, I am not sure you describe an existing problem with classical notation.

For the non-keyboard instruments I know how to play, tones in different octaves having different identifications is a must unless you want to add extra parts to the notation.

I think the point is that the standard clefs would be better aligned if there were 6 lines instead of 5, so both bass and treble could be EGBDFA. It would put the low C one below the bass clef and the high C one above the treble clef, instead of two below and two above but only one in between.

Yes, that was my general thought -- (added:) but that still would have two octaves different, where a note between staff lines in one octave is on a staff line in the other.

Also when you go above or below the two clefs and draw ledger lines for those notes, you're losing the background pattern once again -- you start having to count notes. With this proposed notation you'd draw another background octave (or enough of it for the notes used).

Totally agree. And it's less compact, more verbose, "too graphical" I might say. It's fine to show how to play "Der Flohwalzer" or something, but would be horrible for reading complicated pieces or reasoning about several parties at once. Not to say it's nearly impossible to read the music note-by-note as you play, so you are always thinking in patterns which are hard to recognize when written like that.

So, no way it could replace traditional notation. It could be used to show how to play something to a newcomer, but it's not novel, then — I've seen visualizations with highlighted piano keys since like as long as I remember myself. And not that practical anyway, since it's not that hard to remember how to read musical notation and start using it right away (especially if there's no alteration symbols in the key signature and no somewhat obscure symbols like fermatas and stuff).

You (and the OP) fail to mention the main advantage that sheet music has had over these alternate forms (I know that in celtic music they have tab, which is like this but for mandolins, banjos, etc), and one of the reasons that it has survived this long despite being pain to use: portability.

It is extremely easy to port sheet music from one instrument to another, in most cases this doesn't require any change in written music itself (Or at least, if it does the change is usually very trivial). Conversely, this new format is limited to the piano. I doubt that it would be trivial to port this music to the violin, for example (assuming the violin has a similarily structured music format).

It's not like code, music sheets are unoptimized for learning. And at the end music sheets are just here for you to learn a music, they are nothing more.

Personally I'm very fond of that kind of music sheets: http://herbalcell.com/static/sheets/legend-of-zelda-twilight...

I've learned complex musics in weeks while it would have taken me month with a real music sheet.

> And at the end music sheets are just here for you to learn a music, they are nothing more.

Professional session musicians play directly from the sheet, and can promptly forget the piece afterwards. They do not intend to learn the music.

Interesting enough, a similar observation can be made about writing prose. A printed book isn't usually a tool to learn a novel by heart for recitation.

  Professional session musicians play directly from the
  sheet, and can promptly forget the piece afterwards. They
  do not intend to learn the music.
As someone who has had 'both sides' where classical and folk[1] music are concerned, I can say it is only[2] the professional classical session musicians that do this. From personal experience, I can say that most classical musicians miss out on one part of music learning entirely; that is, the 'learning things off by heart and playing it from memory' part. Whereas all the folk players I've met have not only the ability to efficiently sightread, but also to carry around the equivalent of several volumes of sheet music in their heads; ready to rattle off at a moment's notice. A few of the players I've met (myself included here) see the music as nothing more than a formality, and instead prefer to both learn by ear and play from memory.

Another thing that the classical teachings seem to ignore (or view with distain (People in my orchestra viewed it as 'hard work')) is cultivating the ability to hear a long piece of music and replicate it -- something that a significant minority of folk players also seem to be able to do.

[1]: celtic/irish folk, in my case

[2]: Of course, there are always exceptions to the rule

Thanks for the insight!

> Professional session musicians

I forgot to add I wasn't talking about professionals. I'm talking about people who want to play music as a hobby, for pleasure. Learning Solfege shouldn't be a hassle.

I'm a bassist, of the "don't quit your day job" variety. Owing to the accident of having a certain musical background, I'm a fluent sight-reader, and the groups that I play in require this skill. I frequently encounter a mixture of standard notation (SN) and chord symbols, plus the occasional Nashville number chart.

From what I can tell, people have explored different notation systems for a couple of reasons. The first is that SN is an entry barrier for beginners. The second is to express musical ideas that don't fit within the bounds of SN.

But in my view, the reason why SN remains in use, is that there's a symbiosis between composers who can write it, and musicians who can sight-read, i.e., perform it directly from the sheet. Tabulature, or other pictographic notations don't work because the composers don't intimately know all of the instruments that they're writing for (including the variety of tuning and fingering systems for each instrument), and nobody knows how to sight-read those notations.

Another issue with any method involving computer graphics, is that there are still a surprising number of composers who use pencil and paper, because notation software is so cumbersome.

In one band that I'm in, the composer brings new material to each rehearsal. It's all written out by hand.

I agree. The value of SN is that it's abstract. It assumes a certain structure (the one that western music uses) and describes melody in terms of that structure. It doesn't concern itself with the actual sounds involved (eg. frequencies and durations) or the mechanics of producing them (eg. tablature).

SN is difficult for beginners because music is difficult for beginners. Notations that focus on the mechanics of the instrument are easier to start with, because they don't require any theoretical background in music. That's fine, but they won't be able to replace SN.

I'm very interested in this sort of cognitive technology, but I have yet to see a proposal for improving on SN in its own terms. Is there some notation out there that is simpler than SN, but just as expressive? Something that could be approachable for beginners, but still useful for experts?

One thing I'm familiar with, thanks to my kids, is the Suzuki method of violin instruction. Kids start out just playing by rote memorization, and by ear, with no charts. There were books for the Suzuki repertoire, more for the use of adults, as the kids didn't read from them.

When I was growing up, Suzuki was controversial in the US because teachers were afraid that kids would lose the chance to learn how to read, and be musically crippled for life. I have to admit that I was among the skeptical, since I learned according to the European method.

The first note that I ever played, an open string, I read from a book that had that one big note sitting there in front of me. The entire emphasis was on "getting the cats out of the instrument" as it were, but yet it was all done with reading from the git-go.

What's happened since then in the US (don't know about elsewhere) is that reading is introduced gradually through a separate curriculum. It's been going on for long enough that Suzuki kids are now playing professionally in orchestras, and we've met one or two "superstar" concert violinists through master classes for the kids, who have mentioned their Suzuki background.

So apparently you can start out without reading, and live to tell about it.

Well we all learn to talk well before we learn to read and write, so I guess it's not too farfetched.

The learn by ear before reading is great, the Suzuki method for it is much less so. Suzuki learning by ear is analogous to learning a natural language by memorizing famous speeches rather than by actually using the language.

> In one band that I'm in, the composer brings new material to each rehearsal. It's all written out by hand.

It is way faster to write out by hand than using software. You use software at the end, when your ideas/tests settle and you want a nice copy. Before that, it is really not worth it.

That doesn't have to be the case. Eg lots of people can write prose faster on a keyboard than on paper. Computer entry of musical notation doesn't have to be slow and cumbersome.

It's hard to beat pen and paper, but not impossible in principle.

Yes, musical notation is hard. But no, this is not an improvement.

While a lot of the problems of it comes from its limitations at the origin (think medieval musicians writing music) it is a very flexible and interesting format

What are the problems I see with it:

- Apart from the center notes, it's hard to know which note is which. I know that the second line from the bottom is a G, beats me what's that thing 3 lines above the regular lines

- It's hard to capture what's happening. Chords on top help

- Having to identify notes in G clef, F clef (and C clef sometimes)

- It's an absolute mess when you have lots of simultaneous notes

I know there are a lot of historical, instrumental or music-theoretical reasons things are like that, but there's room for improvement

I just wouldn't think of twinkle twinkle little star when designing it, I would think of something more complex (and Let It Be, props to him, is not very complex but also not very simple)

(And that's not touching the issues with the piano, that's another load of items)

It's clearly stated that the solution is meant to be like guitar tabs for piano. It's a limited, instrument-centered approach and not intended as a complete or universal solution.

Yes, but since he frames it in the title as a "redesign," it's hard to get out of the frame of thinking about it as such. If he had titled it "A Simpler Notation for Piano Music" or something, his intent would've been clearer.

But then again, probably wouldn't have gotten to the front page without the word "redesign"


Why the ad hominem?

I understood the scope of the OP's suggestion. I just thought it could've been presented more clearly. The fact that tons of the comments here are directed at the idea of this notation as a replacement would seem to bear this out.

For my thoughts on the notation itself, see this post: https://news.ycombinator.com/item?id=9831337

The title is "How I'd redesign piano sheet music," then the second paragraph (after the TL;DR) mentions guitar tabs.

At this point, we should just all agree that article titles are at least slightly sensational. If no one on HN debated the clickbaitiness of a headline, comment volume would drop dramatically.

Edit: reordered the paragraphs.


We all know how frustrating HN threads can be. But please don't make them still worse by posting things like this.

Traditional sheet music is already like guitar tabs for piano, because pianos (unlike guitars) only have one physical place where you can play a given note (like middle C). The reason guitar tabs make some sense [0] for guitar music is that "middle C" on a guitar can be played on several different strings.

[0] I say "some sense," because tabs are still fairly limiting, particularly in depicting complex rhythms. Traditional sheet music is still very common (and, I contend, indispensable) for advanced guitar study.

> for guitar music is that "middle C" on a guitar can be played on several different strings.

True, and in the Piano you have to care about which finger to play it with (which is a similar problem)

That's true, but I find it's generally a much simpler problem on piano than guitar, and thus sight reading on the piano is much more tenable than on the guitar.

Does it matter on which string the note is played, except to not get yourself into an unplayable position?

What baddox says is true, but I have to add that there is a noticeable difference in timbre between the same note played at two different places. So if you find yourself with the choice, you have to use your own judgment to decide which fingering you think sounds better.

Yeah, it's mostly about playability. It's easy to get into a position where you're playing the notes in the current chord correctly, but there's no way to play the melody of shift elegantly to the next chord. Classical guitar sheet music contains a lot of hints for fingering. Fret bar indicators are quite common, and sometimes you'll see complete stacked fingering indicators. Some more difficult pieces will give you almost nothing, and it's up to your experience (and just brute force) to figure out the ideal fingerings.

This has some things in common with what we did in designing the notation for keyboards in Rock Band 3 (see https://images-na.ssl-images-amazon.com/images/G/01/videogam... for an example), where we had the same sort of concerns in mind.

This sort of "piano-roll" notation has the nice feature that the elements are easy in theory to parse: time goes along one axis, pitch goes along the other, and you just do what it tells you. As far as notation for Western music goes, it does have some disadvantages.

All twelve pitch classes are spaced equally, so the scalar structure of tonal music is harder to make out. I can tell just by glancing at a page of sheet music whether it is tonal or atonal, and I can't do that here. All the notes kind of look the same (though I'm sure this is true for someone who isn't fluent at Western notation trying to read sheet music!). One way we tried to ameliorate this in Rock Band was to color different groups of notes differently, so you had a lot of features to grab onto (e.g., the boundary between E and F is the boundary between blue and green).

Durations are completely visual, which is nice from a intuitive point of view but means that it's harder to parse the underlying pulse and rhythmic structure of the music. A grid might help here. (I was constantly insisting that the grid in Rock Band be made to be as helpful as possible.)

Anyway, it's a nice visualization of keyboard music, and I don't doubt that this is easier to understand for people who don't read music already. I wish he had chosen a less hyperbolic title, though.

It's not really new as it's used in most MIDI software to represent chords over time [0], or in many music videogames [1]. I think this style might work for very simple music sheets, but would become way too confusing and imprecise for complex ones.

[0] http://www.musicmasterworks.com/MMScShot.gif

[1] http://i.ytimg.com/vi/rNg50UGkXfw/maxresdefault.jpg

There seem to be lots of different visualization methods [1] and the creator of musanim has posted an interesting history of their use with MIDI [2].

[1]: https://www.youtube.com/watch?v=2hmwXThnqi0

[2]: http://www.musanim.com/mam/mamhist.htm

There were similar interfaces on the Atari ST in the mid 1980s, see The Music Studio and Cubase.

I think it was first made famous by Synthesia: http://www.synthesiagame.com/

No way. "Piano roll" UIs have been available inside of MIDI software for 20+ years.

Does the author actually play piano? I didn't see him explicitly say it. I play violin, and reading sheet music was never a limiting factor. It was not a challenge to learn, and you don't learn all the symbols in the beginning. As you get more proficient and read harder music, only then do you learn new symbols.

Maybe sheet music can be improved upon, but the real question is should it? Sheet music is easy enough to learn, and once you learn it, all music opens up to you.

I'm a violist, and I used to teach violin and piano. With stringed instruments, as you say, just making a sound and holding the instrument right are so difficult for beginners, reading music pales in comparison. Piano is unique-ish as an instrument that you can sit down at and start making a sound essentially right away. I found that for my beginner piano students, learning to read was probably the biggest challenge starting out. However, not learning to read is even worse, as then you won't be able to learn music going forward.

And, all that being said, learning to read music is really not very difficult. There seem to be a lot of complaints from the HN crowd about music notation, but learning to read music is wayyy easier than learning a new programming language. Seriously, just get over it and learn the system that everybody else uses, and you will be able to play any piece in just about any repertoire on the planet.

In my view the hard thing is sight-reading. With programming, if it takes one person twice as long as another, to read or write a program, at worst their productivity will be a bit lower. Or maybe higher. ;-) If you can't read music at tempo, then there are certain kinds of work that you can't do at all.

Granted, sight-reading skill is a matter of degree. I can read a jazz band gig, but would struggle with complex modern orchestral music. In musical styles that I'm familiar with, I can often make a good enough guess about what's next, that I only have to focus my attention on the dangerous bits. My attention often wanders away from the page.

I used to think that sight-reading was something that had to be learned, starting at an early age. I don't know if I still believe that, but I know that most adults who attempt to learn, find it to be prohibitive. This is one reason why it would be great to find an alternative. Even a piano with keys that light up.

As someone with an occasional passing interest in playing the piano, Synthesia has been great for me when it comes to learning songs. https://www.youtube.com/watch?v=WolqGAgiolM

Of course, it is not exactly a substitute for sheet music, but it is eminently readable to the novice.

I play both piano and trumpet and I think two hands make reading music at least a little bit more difficult. I've never struggled with reading trumpet music, but lots of piano sheet music still gives a hard time at tempo.

But there's no easy solution to that. You can create notation that makes things easier on certain situations (e.g., it's not unusual to see jazz music sheets with the melody and chords in american notation, such as http://www.saxuet.qc.ca/TheSaxyPage/Realbook%20C/Fly_me_to_t...).

However, you can't create such a notation for every situation because, some times, piano pieces are hard and notation is not going to make things easier. There's hardly a way to clearly see the notes you have to play and their timing when there're multiple notes in different melodic lines. with different rythms: it's just hard and it's not because of the notation.

I have the feeling that it's badly conveyed by """musical pedagogy""". All the theory and notation isn't the essence of music but that's the main communication channel they make student interact with and through.

I wouldn't say it's not challenging, but it's out of place, thus tedious. As other people said, it's an abstraction for people doing the art, IMVHO it's not important unless you are doing the art.

Indeed, and there are entire musical genres where there is little or no pedagogy, and very little sheet music. You'd be hard pressed to find sheet music for the typical rock 'n' roll song. Tunes -- often quite elaborate -- are developed or learned by ear, and played from memory.

But it's not just for pedagogy. It's a practical tool for musicians, because it lets you throw people together to make coherent music quickly. This could be just playing string quartets in somebody's living room, or running a large performing ensemble. The size of one band that I'm in virtually ensures that there will be at least one new member or substitute at every performance. The ability of players who can read the stuff influences what can be composed.

And a lot of this music isn't even commercially viable for the people who play it (such as in my case), but there's just the enjoyment of exploring the world of written music.

Indeed. Basically the only somewhat difficult part is remembering the key signatures without having to count which notes are flat/sharp.

Computer scientists of all people should understand why standard notation is used. It's a high-level symbolic language that allows for music to be ported across instruments and for the communication between musicians. Low-level markups such as tablature are akin to assembly and machine code. They cut out the work of the compiler at the expense of portability and higher level reasoning (such as key signatures). In this analogy, the musician is a dynamic compiler of musical notation into the physical actuations required by the instrument. Tablatures remove most of the work of compilation by depicting a more direct relationship to the physical playing of the instrument, and for this reason they are less cognitively demanding. But if one wants to be a musician, rather than a player of an instrument, I recommend sticking to a symbolic language such as standard notation. It's hard to imagine a composer writing a symphony in dozens of unique machine codes.

I really love this. When I was young, I was really excited to learn to play the piano. The biggest problem I had was being able to read the sheet music. It was incredibly difficult for me to read and understand the notes for both hands at the same time. A big part of this was that when individual notes were off the normal staff (e.g., C is below the staff), you either had to have an innate sense of how far from the staff it was and what note that corresponded to, or you had to stop playing and count to see which note it was. This seems to solve all of those problems.

So does a deck of flash cards. If you learned multiplication tables you can learn to read music. And if you can play video games you can play piano.

And it's NEVER too late to learn.

I suppose one could also say "if you can learn to sight-read a novel you can learn to sight-read music" and it might be true but I think there is a certain natural ability prerequisite. I played in school concert bands for four years and could never manage to sight-read anything very complicated. Practicing was a chore and when we got new music I always had to sit and pick through it very slowly measure-by-measure until I figured out how it was supposed to sound. Once I knew that, I could "read" the music but I could never play a new piece on first sight. I knew all the notation but I could not look at a new piece of music and "hear" it in my head.. It felt more like trying to read a book letter by letter. I never was able to really see "words" and "sentences". By the time I quit I really just hated everything about it.

I also had tremendous difficulty learning basic addition, subtraction, and multiplication facts compared to most of my friends. And I don't play video games.

I do enjoy listening to music quite a bit, but I think I am a person who doesn't have an ability to play it.

There are two factors - an individual difference and a general matter.

In general, musical sight reading requires specific practice most people are not exposed to. Even if you play instruments fairly well, sight-reading practice is not something you implicitly pick up. I play piano for hobby over 30 years, took lessons time to time, but it's only recently I consciously started practicing sight reading and the effect is remarkable. It's a very specific exercise, different from just practicing a piece.

And there is the individual difference. My son is very visually-oriented, that he can make sense complicated figure at one look, but he's having hard time reading long sentences. Some people may just not good at read music.

I chose "flash cards and multiplication tables" as an example for a reason and I find your response fascinating. Thank you.

I will state that group instruction via band class at certain ages is perhaps the worst possible way to develop a love for something ("practicing was a chore...")

I too started note by note, then measure by measure. I felt like a total idiot! Gradually it became "phrase by phrase" and then turned into this surreal feeling where my eyes wander a few bars ahead and somehow my hands catch up in time. I wish everyone could experience that. I was just SO clumsy at first and I started SO late. I'm SURE you could do it given practice.

Are you dyslexic as well, or does it only happen with music?

Not in the usual sense. If there is such a think as numeric dyslexia I might have that to some degree. I frequently transpose digits. I have trained myself to be very careful whenever I have to transcribe a number, fill out forms, or even dial a phone number. I find it helps to look at numbers in groups (pairs or triplets) rather than individually. I don't recall every trying that with music...

But I don't have any difficulty with reading text, and enjoyed reading a lot as a kid so in that sense I am not dyslexic.

My dad decided to learn in his late 70s. Now in his early 90s, he's not good, but he gets s lot of pleasure from it.

I use such music sheets when they exist: http://herbalcell.com/static/sheets/legend-of-zelda-twilight...

it already simplifies reading A LOT.

tl;dr don't cripple yourself with substandard notation when it takes 2 weeks to learn it.

You shouldn't be counting.

Any decent pedagogical training is going to introduce the lower and higher notes one at a time. And 'decent' can include self learning. Have some patience and don't try to jump into advanced things right off. You will just learn bad habits.

Say you know by sight all the notes g below middle c. The next exercise should introduce the f below middle c. When you see it it will be the one and only note you haven't trained on, and before you know it it will be trained into your muscle memory. Soon you will just see all the notes and know what it is. No counting required.

We are talking a couple of weeks here. A note a day, say, will get you pretty far - scores almost always change clefs before going 14 semitones above/below the staff.

Some people will do anything to avoid learning in a disciplined manner, and then spend years never advancing or fighting their bad technique. Take a bit of time, and the world of music is opened up to you.

Analogy - imagine somebody asks you to teach them how to pitch (baseball). You ask to see their current throw and it is some weird, lurchy, shot put type of throw. You show them a standard over shoulder release. They say no, they want to keep their current style, and maybe, just maybe, over many months, first remove some part of the weird lurch. In six months, then maybe they'll raise their hand a few inches. After that is working over several months, then maybe they'll start moving their hand behind their shoulder just a bit. Why, in just 10 years they'll be able to throw a ball!

It's crazy. If you want to pitch a baseball, just learn the movement that is required. If you want to play piano, learn how to hold your hands, and learn to read the music on sight. If you want to play guitar, learn the correct way to hold the strings with the left hand, and learn the proper plucking/fingering of the strings with the right hand. Etc. It's a few weeks of boredom, followed by a life time of being able to play.

actually I find it very zen, and love to go back to the beginning exercises, seeking absolute, unthinking perfection, letting each note ring for several seconds. Learned that from reading about Horowitz, and it works. But it is a bit much to expect 'zen' from a beginner that just wants to play some Billy Joel tune. To them I say Billy Joel did this to get the skill to play his songs, and you are probably not more a natural genius than he is, so you probably can't skip over what he had to do.

What you say is not wrong but it's just not that simple for some people. In my case I did start with the basics, learned to play all the notes and could tell you any fact that the written music represented. The note, whether it was sharp or flat, the fingering to play it on my horn, the meter, the tempo, whether to play staccato or legato, fortissimo or piano, etc. but after four years as a kid I was never able to put it all together at tempo on sight. Once I had learned the piece yes I could use the written music as a reference and play it. But never on first sight. And I really don't think it had anything to do with the notation.

The traditional notation system has lasted so long for a reason. It's super difficult to figure out how long notes are meant to last with this notation system. It also takes up a lot more space than normal music notation. Moreover, it looks really difficult to write because of how note durations are represented.

"[The] target user is probably learning easy, contemporary songs, and already knows what they sound like before they learn them."

Leaving out exact note durations is intentional, and probably the main reason why it's so easy to read. The shaded tails from notes just have to evoke what's already in your head.

The problem there is that 90% of people have the lengths of notes in their head wrong. Sure, if you ask them to hum a song, they'll produce something that sounds rhythmically passable, but it's a far cry from being correct.

But think of the alpha channels!

Just wanted to leave this here, for people who get off on alternate music notation...

An amazing collection of functional/beautiful/arty alt music notation can be found in "Notations 21", edited by Theresa Sauer. One score per page, all wildly different in their approach to encoding music. Just wow.

This video has some images from the book:


A bit of a click-bait title, don't you think? Maybe "How I'd redesign piano sheet music for beginners."

This is basically just tablature for piano. Great for simple pieces, unweidly for anything more. Can you imagine notating Fur Elise with this? That's a fairly cmomon beginner/intermediate piece. Or etudes? Or any excercise that develops techincal proficiency?

Maybe it has a place, but like tab or lead sheets it is not a replacement for sheet music.

I play guitar and find having both sheet music and tablature together useful. Sheet Music has an incomparable amount of timing information and other nuances, and tablature has the fingering positions you can sight read effortlessly once you've got the piece down.

I am on board with the goals of the article, and I've actually been working on an iOS app to pick away at this problem[1] (albeit in a different way — mine is more fluid and is intended to be used interactively, not to be printed). Sheet music is a great format to represent classical music, but the fact that it doesn't allow you to easily notate one of the most common elements of modern music — syncopation — points to the fact that it's out of date for today's needs by about a century. Modern music does not fit into a rigid meter. It goes off-beat; features arbitrary note lengths and overlapping meters; pitch bends without a second thought; ties performance and production together with the notation. If you want to write down popular music on staff paper, you'll be using a lot of awkward dotted notes and rests.

In my app, I figured I'd try to distill written music down to its basic elements — pitch and time — and allow users to draw on notes arbitrarily, with the equal temperament pitch grid as a guide and pitch/time snapping available as an option. (As an aside: the pitch axis is basically a logarithmic graph of tone frequnency — cents from A440. Because of this, I can swap out the equal temperament scale with basically any arbitrary scale. I've implemented *.scl file import and have been playing around with odd non-equal tunings from huygens-fokker.org's archive[2], though I don't know if this will end up in the final user-facing product.)

[1]: Early demo video: https://www.youtube.com/watch?v=Ra8OvnoxKQw

[2]: http://www.huygens-fokker.org/docs/scales.zip

This idea isn't entirely new (is anything?); for an earlier version, check out Klavarskribo: https://en.wikipedia.org/wiki/Klavarskribo. It didn't really take hold, because it's much more inefficient than normal notation.

I spent a long time Googling before finding this, only to see that it had already been posted. When I was in college I went on a tour of the Netherlands with my school's choir. We sang in a number of Protestant and Reformed churches and I remembered one of the music students pointed out a hymnal, still used by a congregation in Friesland, that used this particular notation system.

Someone else suggested that this was a notation with which they were familiar, from an old hymnal of the (U.S.) denomination affiliated with the college I attended. I had assumed it to be quite an old system as the hymnal appeared quite antiquated. Still, it doesn't surprise me that it (only?) dates to the 1930s, nor that there would still be churches using it. Many old Reformed churches have short institutional memories but nonetheless cling to assumed traditions. That, and the Frisians are, stereotypically, an old, stubborn people.

My grandfather was an expert piano player who used klavarscribo. All klavarscribo players I know have a very distinctive style/intonation, maybe because they were all either taught by my grandfather or from his generation or something. I'm not sure about any inefficiencies, my family said you needed to know classical script to get into music academies, and the academics wouldn't switch so they were locked out, but obviously my family is rather biased..

Yep. They guy reinvented Klavarscribo.

The first issue that I see with this notation is that you loose information compared to standard notation. For example, a half note contains the information of how long it will last along with the note itself. This information is given to you when your eye hits the note at which point you can stop looking at it. In this notation, you either have to continue seeing the note for the length or memorize the lengths of different notes which would require you to "zoom out" much farther.

IMO, an experienced sight reader would have more difficulty sight reading more complicated music using this notation.

This is intended as a 'piano tab'. If you can sight read, you already have music. Those of us who don't know how to read music would benefit.

Sheet music evolved when it would be difficult to produce a different representation for each instrument. Now this can be automated it makes a lot of sense to produce specialised formats, such as tab for guitar or this for keyboard. As a novice keyboard player I'd certainly prefer to use this than sheet music. Having said that there are advantages in sheet music that a professional musician would find lacking here. There's no rhythmic grouping (the beams provide this cue) and I'm sure there are many others that I'm unaware of.

Yeah, this only makes sense if you already know the song. trying to infer the correct timing spatially would be difficult.

Here are some shortcomings I thought of immediately.

- Traditional notation is instrument-agnostic.

- It may be easily printable, but a staff is easily writable. With just a pen and a ruler, I can produce a neat staff in ~10 seconds. The ruler is optional if I don't care about prettiness. With this, you more or less have to print it: getting the spacing right by hand would be quite difficult.

- I have mixed feelings about the the traditional representation of key. Missing a sharp or flat in an unfamiliar key is probably my most common error when reading sheet music. OTOH, The traditional notation tells me which sharps and flats to use. With this system, I have to already know the structure of C# major if I want to noodle around (always). This would seem to be more hostile to beginners.

- The precision issue has already been brought up many times, but I would add that all the lost information is an archival disaster. It might seem superfluous since we have recordings, but given the rate of change in technology, access to data in old formats is by no means a certainty. Since this notation basically requires you to have heard the song in order to play it, it would become incomprehensible pretty quickly.

It seems to me that most of the complexity of musical notation reflects the "essential" complexity of actually playing the music. For example, consider key: even at the level of simple rock songs, a basic grasp of "key" is absolutely essential, and if you're playing a piano, this means you have to know which sharps and flats to use. If this seems pedantic, consider that the main difficulty of learning to play a piece is dexterity. It's going to take a bit of effort just to make your fingers hit the damn keys. Since this effort has to take place anyway, the little bit of extra effort to learn notation doesn't seem like much to ask. Besides, the amount of notation required to play pop songs is relatively small - that rendition of "Let it Be" contains about 10 symbols, most of which are related.

This isn't to say that traditional notation is perfect, or even good. It is rather baroque, and I'm open to a "redesign," as long as it really solves the problem. This redesign isn't useless, but its scope of usefulness seems so limited that I'm not really sure it's worth it.

This isn't redesigning piano sheet music, this is just creating a piano tablature. That's great and all, but this has existed for a long time, it's similar to how Synthesia[1] represents piano notes. The fact that the timing can't be easily read from it means that like guitar tabs, it can't be a replacement for traditional sheet music. There's a good reason sheet music has been around for so long, it makes sense and is versatile, it doesn't pertain to any particular instrument, and I don't think it's going away any time soon.

[1] http://www.synthesiagame.com

I've been playing the piano since I was a kid and found it very difficult to learn to read sheet music. There was a period for maybe 3 months where I was fluent in reading after 3-4 years of study but then took some time off and lost the ability.

I think very visually and find watching someone play a song on youtube infinitely easier than trying to learn with the equivalent sheet music. I really feel like this comes down to a left brain/right brain thing but I'm not sure how true that generalization is.

Are more visual approaches like youtube tutorials or this redesigned sheet music the way of the future? I wonder if it might it be detrimental to take this approach from the start?

By pure chance I had just sat down at a piano, trying to figure out my old music sheets that I had managed to play, with lots of stress and never well in any sense of the word, 10 years ago. (My piano lessons got me to "Fröhlicher Landmann", before my teacher and I gave up.) Now I struggled with the most basic lessons and contemplated about the inefficiencies of the notation.

I gave up, somewhat frustrated, and then saw this post. Here is my experience:

- The top to bottom timescale is easy to get use to and makes sense, given that you need a wide space just to fit all the keys. I placed my iPad on the piano in horizontal mode and could well imagine an app that scrolls the sheet for me, for time progression. Edit: Sorry, I was so excited about the article, I did not even see that it goes on after the notes, where the author talks about this.

- With the standard sheet music, because decoding was so inefficient, I would first figure out the keys and then try to memorize the finger pattern as quickly as possible, so that I don't have to decode the notes again. With the new notation I got lazy and kept my eyes on the sheet for the whole time, which actually slowed me down. I had to remind myself to memorize more. Just something to get used to, I guess.

- The notation does not give an exact rhythm, which I actually loved. If you've ever heard the song, you quickly figure out what the rhythm is supposed to be. The minimal rhythm notation is great, because it's not in the way of the notes.

I had a lot of fun learning the start of this song, and I would love to see more. I always wanted to learn some Beatles songs, and the sheet music that my mother bought so many years ago was always too complex to be fun. With this notation, songs like that are a lot more approachable. I don't think it makes sense for classical music, where the timing is not as obvious, but for pop and folk songs, or playing along to Studentenlieder ("im schwarzen Walfisch zu Askalon", anyone?), it seems like a great choice.

One minor sugestion to the original author: It would be great if the grey lines that correspond to the black keys would be bolder and darker. I had trouble seeing them, but I don't have the best eyesight...

I love the idea of trying to come up with a better way of notating music to make it more accessible. That's a great idea.

However, I'm skeptical that this is it. First, looking at the 2 notations of "Let it Be," we can see that the new notation does not convey the same data, or even a simplified version of it. It's a different version of it. I find that odd.

Second, I don't think learning music notation is very difficult. I learned to read treble and bass clef when I was 5 years old. It was much harder trying to make my fingers press the right keys at the right time than memorizing which circle on which line matched with which key or note.

Looking at this, I'm reminded of Laban notation (https://en.wikipedia.org/wiki/Laban_Movement_Analysis) and Benesh Notation (https://en.wikipedia.org/wiki/Benesh_Movement_Notation) for notating dance. Both are very complicated and essentially never used by the people who actually do the dancing. Furthermore, I'm told that Benesh, for example, allows a notator to add written (human-language) notes below the staff to convey additional information. In other words, they know it doesn't convey everything it's supposed to.

So what do dancers do? They learn from video. They watch the masters who went before them directly performing the movement. Frankly, I think it would be great to see something similar for music. Having a video of someone playing a keyboard from above, possibly with a graphical overlay showing the notes pressed would be way better (in my opinion). (Which isn't to say it's without issues - far more bandwidth, video codecs, playback hardware, etc.)

I have always wondered how dance/ballet is recorded, and I'm not surprised it comes down to video. But music is more limited data I think, an opinion supported by the fact that sheet music exists and sheet dance is still questionable.

There's a passage in a piano piece I'm having trouble with and I tried watching videos of pros on youtube to figure out how to play it. It was frustrating and almost useless. The video goes by too fast. Granted, I was looking for technical details, not notes, but I think notes would be almost as hard.

E.g. Can you tell what notes are being played in the video below?


Here's another fun one to try:


You can slow down youtube videos in the gear menu and it still can't be seen. I suppose you would need 60fps to see that.

I think this a cool idea. I think piano needs its own version of the guitar tab to be more approachable. For someone who knows the basics, just seeing the chords goes a long way. But this gives better detail.

One thing it glosses over is how the format responds to greater range. It acknowledges the fact that only 3 of piano's ~7 octaves are shown, but it doesn't address how much difficult it would be to read if there were notes in both the C2 octave and C6 octave at the same time. It gets much more difficult to line up the timing in that case because your eyes would be scanning frantically every note.

In traditional sheet music, notes played at the same time are never more than an inch or two apart. Will this format support sheet music's equivalent of 8va to keep notes properly within visual range?

> I think piano needs its own version of the guitar tab to be more approachable

But then you end up with the same problem you get on guitar: beginners use tab as a crutch and don't develop skills as readily as they should once they're over the beginner hump.

This is the hardest part of learning and teaching music. On one hand, it's hard to stay motivated when you're mangling Twinkle Twinkle Little Star, but almost any shortcut you take in this stage will retard your growth later...

The best way to do that stackable staves: see the last two lines here: http://i.imgur.com/tqZRXzt.jpg

Ledger lines preserve the melodic contour but it's harder to read the notes. 8va's stay within the regular five lines you know how to read, but fragment the melodic curve. This way has the advantages of both the above and the disadvantage of neither.

(The lack of key sig in that image is a limitation of the software it was printed out from, I think; it should start in E and change around a bit afterwards. The "3"s are that software's way of printing tuplets.)

That's cool, I hadn't played any piece that uses the stacked staff (to my recollection anyway; it's been a good 15 years since I was actively training). It does work better for when the melody progresses into that range, true, but I think it's a "right tool for the job" situation. Often 8va's are used for when you want to raise a single note or chord one time, and that would be much more readable than a separate staff or ledger lines. It seems like it's a random access vs sequential read type thing. 8va is better if you're jumping to a far octave with no intermediate context; stacked staves would be better for when the melody incrementally takes you there.

I'd love to see what the equivalents of both approaches would be in this notation.

A single staff can switch between different clefs, if you ever want to redfine what lines mean what notes. Then you can switch back to the normal treble clef.

That double treble clef is called the super-treble clef and I've only ever seen it in the appendix of a book called the science of sound.

I don't want to diminish the author's effort here, but this system feels like it would have difficulty beyond basic cases. And, in the most basic cases, it doesn't offer much more than simply calling out the chords over the lyrics.

It does lead me to think that a simple piano roll layout, like the ones found in Logic and other digital audio workstations, would be useful for those of us who are terrible at sight reading music.

The only people who waste time trying to reinvent the wheel are those who never use a wheel.

This is silly. Standard notation isn't absolutely perfect but it's pretty darn good. If a symphony can come out of dozens of musicians consistently by using standard notation I think we are OK.

I think the problem often has to do with not wanting to put in the time to master not only the notation and the instrument but also the hours and hours of work required to crate the synaptic connections necessary to read and play without thinking.

It's like complaining about vim because it is hard to learn. The problem isn't vim, the problem is the person who doesn't want to do the work.

Reminds me a bit about how Mark Twain would redesign English spelling. Once you've read sheet music for a while and played it you don't really see it any more, you hear it in your head and you play it. I'm sure there are many different ways you could write it but they all suffer from the same challenge, which is is first training your brain to read it.

Given the amount of information on a page though, you would have to at least match that, otherwise you spend to much time flipping pages and not as much time playing.

The article says that the notation is for beginners.

I am surprised all the JavaScript programmers on HN are so opposed to an accessible notation for beginners.

I realize you're trolling but you might consider that thought a bit more carefully. In particular 'notation for beginners' is simply one instance of "<x> for beginners" which, doesn't scale to "non beginners". I have had a lot of UI's foisted upon me with the admonition "Its really easy for beginners to come up to speed!" and often it is, but as they get up to speed their ability becomes hampered by the design which doesn't support "non beginners".

It isn't surprising, the first time anyone tries to master something a bit more complicated their initial thought is "this is soo confusing! Why doesn't someone design something easier to learn!" and for the motivated ones they may follow through.

The problem is then that they have mastered their "easy" thing and now their trying to do more complex things, or in this case describe more complex music, and wham, suddenly they just can't seem to make it work. But had they actually mastered music notation first they would have mastered a system that has successfully expressed a wide variety of music, styles, and levels of complexity. So they wouldn't end up stopping in the middle of their "progression" to go back and learn a new system because their old system pooped out on them.

very nice! my 2c for improving notation are more evolutionary. i see two main issues with sight reading:

1. treble/bass clefs are not consistent: one is EGBDF, the other GBDFA. eg instead of the third line always being a B, it's a B in one & a D in the other!

2. having to remember sharps & flats. ie, despite the note being F, you have to remember that it's not really an F, but in this case an F# or Fb.

the proposal:

1. make the clefs consistent: make them both 6 lines, and both EGDBFA. this means adding an E to bass at the bottom, and an A to the treble on the top. now a student has to remember only one format and note positions are consistent. the third note is always a B regardless of the treble or bass clef.

2. to resolve flat/sharp issue, especially for complex keys with many sharps/flats, use the note's head shape to distinguish. ie, instead of just an oval head to represent a note's pitch, use for example a round note head for a normal G, a square note head for G# and a triangle note head for Gb. that way no matter what the key, your eye can easy and immediately recognize that that G is not in fact a G, but a G# or Gb. after all we already use note shape (solid/hollowness) to indicate duration. why not go a step further and make realtime parsing easier by making the notehead shape indicate sharp/flatness?

The more sharps or flats the key has, the more the need for putting them in the key signature, not less. Otherwise there'd be sharps and flats everywhere. And you wouldn't be able to determine the key by looking at one bar.

Experienced musicians don't look at a key signature and say "okay, so I need to perform so-and-so particular substitutions", they look at a key signature and say "okay, so it's in the key of X".

agreed, especially for experienced musicians. the point here is that if the notehead carries its sharp/flatness in the shape, just like a notehead carries its duration (quarter vs half), it'll become 2nd nature. if a piece is in F# major, then pretty much most of the keys have an implied sharp... A is not A. it's A#. this way a student doesnt have to constantly remember the key sig & what notes are raised or lowered. sort of like python's explicit > implicit.

as far as note being able to determine the key, that'll still be specified as now - this is an additional visual hint.

Oh, and there's already a notation within the standard system for your first proposal: a treble clef with "8vb" written below it.

Someone else mentioned adding another line atop the treble clef and then using the 6-line staff for both clefs. I like this idea because it maintains the one-line gap between the staves and also somewhat reduces the need for ledger lines, which drive me a bit mad.

hmm... i'm confused... 8vb is for lowering an octave. the proposal here is to make both clefs have EGBDFA. ie, the clefs are off by a third - this slides by a 3rd to make them have the same arrangement. ah, actually i think i got what you're saying: ie, just use 8vb and use treble for both. sure that'll work too, tho the A below middle C is kind of popular :), otherwise it'll have to be constantly dotted lined in.

Compared to normal sheet music, this system mostly deletes cues. In normal sheet music, you have both the shape of the note and its rough position horizontally on the page as cues about the rhythm. Measure lines are removed in this system, which is a pure loss of readability, like removing periods and commas from English and replacing them with extra spaces (while saying you are trying to aid beginners, who may be still learning what a sentence is!). In normal sheet music, a note is either on a pitch line, or between two lines, where as in this new notation, a note may be on the left half or the right half of a space between two lines. It's just visually more difficult to parse.

The octaves are seemingly delimited by solid black lines, but these lines actually represent the note C#, counterintuitively. My feedback there is that you've been staring at this too long if you think it immediately resembles a piano keyboard. Make it look more like a piano keyboard if you want it to be more readable than sheet music for beginners.

Make the example of your notation match the example of traditional notation. There should be three notes in the right hand at the beginning of Let It Be.

It looks like a keyboard doesn't it

Not really.

Yeah, I couldn't see it until the graphic of it next to the piano.

The graphic overlaid on the piano keys was much easier to read - as easy as guitar tabs. Why didn't he just do it all that way.

Having learnt standard notation one can easily read music composed for any instrument, and even compositions from a different musical culture (e.g. makam), whereas with this system one's limited to western-music piano pieces. Besides, standard notation is not hard to learn, in fact it's only as hard as learning the alphabet, which kids do in the school, and I'm yet to see someone coming up saying, hey this alphabet is too hard, better kids explain themselves with pictogrammes and emoji, which convey the ideas more directly and easier to recognise.

Is this really necessary?

When I took piano lessons as a kid, I never had a hard time reading the music. It looks really complicated knowing nothing and jumping straight to Beethoven, but realistically, you'll learn to read sheet music a lot faster than you'll learn to actually play it.

I only took lessons for a few years, but by the end I was able to read sheet music that I had no hope of actually playing.

And then there's hundreds of years of music using the existing notation, so anybody serious about music will inevitably end up learning the old notation anyway.

Disclosure - this is a local startup and though I don't know the founder (or founders, I don't even know) I did fall in love with this app at an open house a few years ago. I am always biased in favour of entrepreneurs, but I double down on local entrepreneurs.

Disclosure 2 - I'm generally a horrible musician, but when I was very young, my Grandma Yvette taught me to play the fiddle before the Suzuki program tried to save me and show me the violin. French Canadian fiddle music is all about patterns and so it's possible that that shaped my brain to look at music and its theory as a set of patterns.

However, those aside, I wonder about a tool like Musix (or Musix Pro) for notation. Within about five minutes of playing with the app, I felt like I had absorbed more real, practical music theory than in my years of formal instruction. Entire melodies started to make intuitive sense to me.

I wonder if a tool like Musix could be useful as a notation tool?? It still makes timing difficult to express and it would likely completely break with more complex pieces so I can't picture it seeing much uptake amongst highly experienced musicians, but I feel that if I had been exposed to an isomorphic notation at a young age, I'd be a significantly better musician today.

Anyone with more experience care to weigh in??

Back when Electronic Arts made tools for electronic artists, there was a great program called Deluxe Music Construction Set. I've heard similar versions of your story from people who experimented with that software. Clicking notes on a staff, seeing things light up on a piano keyboard. Some people prefer to learn like this, and some people did.

Professional notation software is an utter shit show. Microsoft Surface 3 launched with an interesting touch screen notation program in their promos. Stuff is out there, though notation isn't as popular (or nearly as well done, or as immediately rewarding) as some of the other apps like Ableton Live.

I want to say that it's really exciting to see something TAB-like for piano. Something to ease the learning curve into instruments other than guitar I think are in dire need. I hope more instruments can get this kind of easy-to-learn-notation.

I hope this doesn't entirely replace staff notation. The reason for this is because of its connection to music theory. Once I got to a certain level with the violin, I didn't have to think about how notes mapped to the violin physically, but was able to think more in terms of music theory (or at least it was rather easy to switch between the two). This opens up an entire world if you start tinkering with improvisation and gives you a common language with other musicians.

I think staff notation is also quite amazing because it represents music in such a way that it can be easily understood by musicians of many, many instruments at a glance. Imagine writing a symphony of however many instruments in different kinds of notation, or being the conductor who has to read it all simultaneously...

All that being said, I could totally imagine a world where a huge percentage of musicians (especially amateurs) learn only these easily readable musical languages, while people who want to dive deeper learn staff.

I like playing the piano because the staves correspond more or less directly to the physical layout of the keyboard. If you flip a score 90 degrees, you end up with this, with the notes laid out over the keys. Compared to almost all other instruments pianos are ideally suited to sheet music. Manuscript notation is also (necessarily) very information dense.

It's an interesting approach; there are apps that will do something similar to this. They're very much like guitar hero. Some of the newer electric pianos will even illuminate the keys in the correct order.

I think one of the reasons that guitarists use tablature is that using sheet music notation for the guitar is more difficult to learn (at least to me). I've struggled with the guitar because I haven't put in the time to get my head around translating between manuscript and which strings/frets I need to be using. Violinists don't seem to have the same problem, but perhaps that's because most violinists are classically trained whereas most guitarists aren't?

guitar is also difficult because the same note can be played in as many as 5 different places. sometimes its hard to tell which string to actually play it on, since a choice has fingering and possibly picking ramifications later.

How many guitarists (and I'm thinking classical or, perhaps, jazz guitarists) can just sit down and sight-read a piece? I suspect part of the craft is learning a piece, and learning the most efficient fingerings/patterns for the piece.

I've been in guitar orchestras, what we do is annotate our sheet music with optimal finger positions as we work through the piece. Especially for long pieces you can't necessarily keep in your head, this helps with sigh-reading tremendously since the positions are written in the sheet music.

Some sheet music for guitar also has note-level annotations for string number and finger positions, which does the above for the guitarist.

sight reading charts is a must for journeymen pros in jazz, though some of the greats didn't read.

Not bad, kind of guitar-hero like. I like that the key and key changes could be very clear. It's also nice that new musicians wouldn't have to memorize certain keys, since an F# when playing in G major could just be printed on the line.

It is also much easier to see intervals. A minor 6th for example would always be a certain number of lines/spaces away, no matter the key. Octaves are obviously also easy to recognize.

I think a couple things are confusing and not as clear as they need to be past the beginner level. For example dotted notes. Counting something like dotted 8th notes or recognizing a dotted note would be hard. You can see a similar confusing rhythm in Linus and Lucy, where the 8th note at the end of each measure is held into beginning of the next measure.

It's pretty neat and a unique project. Pretty cool and I could actually see it being used in for things like Elementary School Music class.

Vertical orientation of sheet music looks very much like a music Tracker. [1]

I've recently been learning a tracker program called Renoise [2], and the vertical orientation is a very nice change compared to other DAWs I've used (Reaper, GarageBand, Logic Studio).

Reading sheet music has always been a huge chore for me, so I generally seek out guitar tabs instead when it's possible. Vertical sheet music without the staves would be amazing for people who know music theory but never had the need to learn sheet music like myself.

[1] https://en.wikipedia.org/wiki/Music_tracker

[2] https://upload.wikimedia.org/wikipedia/commons/d/da/Renoise_...

This is a neat idea, and I can see a place for a different notation for all instruments to make music easier to read for them. However, of course, this means that music for one instrument is not readable for another, this is no universal system. If perhaps this were also tied to a universal ur-system which ties together the formats for all instruments in a way whereby someone who knows one format can read other instruments' musical notation, either by reading an abstracted central version, or have an automagic system for transcribing the musics to their familiar notation, then this might be the sort of thing to be useful. However, for general musical use, a universal system is likely to be nevessary for musicians, because it makes it much easier to read anything that's lying around.

The best use I can imagine for this is as a progressive tool to learn traditional sheet music. You start with this (simple stuff, maybe even add the finger-numbers onto the dots so one knows which finger hits which key) and make it harder step by step. Introduce pedaling, remove the numbers on the dots or replace them with their note (first by letter, later by graphical representations on traditional sheets), before you gradually transform it into traditional sheet music. Last step would probably the switch from horizontal to vertical but at this point you will have the notes down and it should make it easier to transfer from notes to keys.

Interesting. Reminds me of the Synthesia game [1], which is like a Guitar Hero for the piano but where you can actually learn to play the real instrument, and not some plastic controller.

I still prefer the Hummingbird notation [2] though. Instead of this, it improves the classical musical notation without creating an entirely new notation. It's relatively easy to go from hummingbird to classical notation and vice-versa.

[1] http://www.synthesiagame.com/

[2] http://www.hummingbirdnotation.com/

This is a very interesting subject. At least for me, as it draws on two of my passions: HCI/UX/usability and, well, music.

I think this kind of approach is defined essentially by a closer mapping between the sheet and the physical world (which is, according to usability tenets, something good). So, under this "definition", we can consider other approaches which are very popular and indeed work: guitar tabs and "Guitar Hero"-like games.

In the wise words of Thom Yorke, "anyone can play guitar", and I think this is - nowadays - mostly due to guitar tabs. Tabs make it very easy to step into guitar playing. So, I think, it's not a surprise that these approaches are so popular, as they have a easier learning curve for beginners, which is in itself also not a surprise if we consider the usability principle of close mapping with the real world. Traditional music notation just has more layers of translation, so to speak.

So, what are the drawbacks of this? I can think of three, which might explain why these approaches aren't promptly adopted by traditional music education:

1) As this is, essentially, a closer mapping to the physical world, said mapping has to be "done" to a specific instrument. An alternative notation such as this one or guitar tabs wouldn't be very helpful for, e.g., wind instruments. So, with this we would create specific "languages". Arguably, having musicians understand an universal language is a good thing - I'm no piano player, but I can read a piano sheet;

2) Repertoire, generally. The amount of music written in traditional notation is just huge, the inertia to start writing stuff with a different notation would be a problem;

3) Dynamics: I think these approaches overlook dynamics. It's not only about what and how long you play, it also about how you play it. For wind instruments, e.g., the dynamics can get quite complex - if a pianist has to read 5 notes, a flute player has to read one but with 5 dynamics annotations. This is, however, somewhat addressed in guitar tabs, as there are quite excellent tabs with not only dynamics annotations but also instrument-specific techniques. Maybe, in this approach, circle sizing could be used to communicate velocity?

How I'd redesign a piano, is with the black keys equal in color and shape to the white keys. That way, transposing a song by a semitone is equivalent to shifting your hands by one key. Much more natural.

I think they call an instrument with this property "chromatic", but I'm not sure.

Actually, perhaps somebody can explain why the black keys were invented. Because their existence makes little sense to me. For example, a guitar is a chromatic instrument, and can be played perfectly fine without any difference between "black" and "white" notes.

You have to ask yourself what should feel more natural: transposing songs or actually playing them. I don't know much about why the piano was designed like this, but I can see your idea wouldn't be useful:

- Black keys are a reference. Either by touching them or by looking at them, you can always know where you are. Not so easy with all the keys being the same. - The size of the keyboard is more or less adapted to our hand size. The distance between whole tones is similar to the distance between our fingers. A fifth is as wide as my hand, and I can play an octave without much effort. Making all the notes the same shape would change that and those common intervals would be more difficult to play. - Our hand is not flat. The fact that some notes are higher than others makes playing more comfortable. - Range. Same-shaped notes would occupy more space (unless you make them smaller than white notes, which would be uncomfortable) so either you get way bigger pianos (as if they were small) or pianos with less range.

There are probably more reasons why the black keys are there, these are just top off my head.

Thanks for your explanation.

I think there is another advantage (to merely having the white keys) besides just transposing being easier. And this is that the same intervals will always have the same spacing on the keyboard. I guess this should be my main argument for this type of layout being more natural.

Yes, they would be the same spacing, but that spacing would be extremely uncomfortable and lots of pieces would become impossible to play. Making the black keys the same size as the white keys, a 8th would become a 13th. I don't have small hands and I can barely reach a 11th interval. Basically, a lot of simultaneous notes could no longer be played. A solution would be to make the keys smaller, but then you wouldn't be able to play semitones because your fingers wouldn't fit comfortably. To fix that, you could raise some keys, so basically you would end up with the current setup.

Good points. But I really don't understand why the keys need to be that big. People are used to typing on smartphones, where the sensitive area of a key is like 30% of the width of a finger. Of course, making them that small would be a little extreme, but I think piano keys could be much slimmer.

Have you seen anyone try to type on a smartphone? Slow, and riddled with typos. Keep piano keys the width they are, unless you want slow, error prone pianists.

Besides, we already have the ideal chromatic keyboard layout:


Typing is 1-0, you either press the key or not, there is no need for as much precision.

This is absolutely true. I find it quite difficult to play music in C without looking at my hands if I'm moving around a lot - if I'm in a key with a couple sharps or flats, it's much easier for me.

Such an instrument exists! http://www.rogerlinndesign.com/linnstrument.html

My biggest gripe with this notation is that it assumes you're playing in a key, so the lines go <D E F# G...> for the key of D. That causes the spacing for a D chord and an E chord to look different, even though they're composed of the same intervals.

This stuff gets reinvented every few years.

It never sticks, because there's literally an entire industry devoted to traditional notation and traditional keyboard design.

Meanwhile in pop and electronica everyone started using button grids to trigger loops and samples, and the notation thing was almost completely bypassed.

This is very, very limited a anything busier than a few block chords - which includes a lot of music - will be too dense for it.

Notation evolved in an ad hoc way, so there's certainly a lot wrong with it. But it's good as an efficient representation of music. Most of the examples here would fit on a single line of notation - so this model is around four to eight times less dense.

Try setting something simple like Bach's Prelude No 1 from the WTC and you'll see how inefficient it is.

1. It would make the equivalent keys even further apart and many songs harder, if not impossible to play.

2. The 2-3 key structure makes it obvious where each note is, it would be extremely difficult to find out where the notes are if all keys looked the same.

3. How often do you need to transpose music?

I don't understand why insist in "little dots" notation in music if you plan to redesign how it's written. There are common patterns in music that standard notation does not take into account, e.g. a chord is much easier to read and recognize by its name rather than figuring out the dots. There are many more rules that would allow such groupings of notes in a way that's much faster and easier perform. Why not take that approach and make it as easy as reading texts?

You lose a lot of information that way, e.g. voicing/inversions. Also, reading text names get a lot more cumbersome when you need stuff like "Ab half dimished +9".

I like the idea but one big flaw is eye-movement.

With traditional sheet music, you can see all the notes you need to play clearly in one place.

With this one, if your left hand is playing in lower octaves, and right hand playing in higher octaves, your eyes will have to move around a lot just to know which keys to push. This is in addition to moving your eyes around the actual keyboard to make sure you're hitting the right keys.

However, I do like the simplicity of using this system to learn a fairly basic song.

So traditional Sheet Music and hence, notation is akin to C or Perl's verbose and/or messy syntax...the problem is, it's not just about the Notes with Music: there's a lot more in terms of dynamics and the like going on which notation has evolved over thousands of years to support.

Interesting read though. For something not so 'serious' (e.g. not a Concert piece), it could definitely make the correlation easier.

The funny thing about talking of tabs for piano, is that sheet music is already that. By representing music on a diatonic staff, with a key signature, it's shown in a perfect format to translate it to keys on the piano, at the expense of some other kinds of instruments, vocals, and general understanding of the music, where it would be more helpful to see the music on a chromatic staff.

Just call it piano tab. It's fine for chords and pop songs. There's no way it's going to work for complicated music. I love the project though. It's a fantastic job, I hate to be negative. You did a ton of thinking on this.

For pop music, fake books are easier. That's how you play pop music.

Those are great, I could start reading those immediately for piano. I think it would be more enticing for newcomers to learn this "Guitar Hero sheet music" than traditional sheet music.

Just imagine if Guitar Hero had been done with traditional sheet music. I don't think it ever would have taken off.

An interesting fact is that Jean Jaques Rousseau also came up with another method, arguably similar to this one. details at http://normanschmidt.net/rousseaumusicpad/

This looks pretty much like piano roll to me, used in just about every midi software.

I'd use it though, even though I can read music I like to use tabs for guitar. This would be nice for me since I don't play piano, but would like to play a simple tune once in a while.

Isn't C the first key of each group? If so, his black lines appear to be on the wrong side.

I really like this. I used to play piano, and I just hated reading sheet music. This seems to much more simple.

I don't think it'd work well in paper/book form, but on a scrolling screen I think it'd be a great learning aid.

this is the second redesign of sheet music I've seen in the past few years. I applaud every attempt, because it forces us to reconsider what we already know. however, as an amateur musician for the past 30 years, it's worth considering how great of an interface "normal" sheet music actually is. The same clefs and staffs are used for every instrument, so you don't have the compatibility issues that this piano sheet music would have. The only sheet music I personally can't sight read is for the cello, which uses a different clef.

Those old saloon barrell piano didn't exactly sound like Chopin so I'm guessing this is a lossy format.

That's because it was interpreted by a machine instead of a human. All notation is lossy when compared to human expression.

Although may be easier for moons, it takes up too much space to be practical for pros.

How is it anything else than sheet music turned by 90°?

You are a genius! Thank you!

Just a few observations in no particular order from my point of view (classical violinist and pianist--performing professionally for 28 years, and a music theorist studying the structure, composition, and history of the above for about 20 years.)

1. This, and all other similar modern attempts, is interesting in that it has so much in common with the very earliest forms of music notation. We have as a basic, "this is me telling you how this is supposed to sound" example as far back as the 7th century, in which you had written words with markings above, below, and around that are meant to guide chant practitioners about where to go next to shape the musical phrase. This is also the origin of modern punctuation in written language.

These attempts are remarkably similar in my mind to the early attempt to guide people. They strike me as memory aids more than anything. As in, I've heard this before, I know how it is supposed to go, I just need a little help here.

When I say origin above, I'm talking about the western tradition of music. The eastern tradition of music evolved very differently (both in terms of music notation and the notion of punctuation), and you can trace early concepts of music notation back a couple of thousand years b.c. But I'm not talking about that.

2. The history of music notation roughly follows the history of the study of music theory in a cycle: is the point to understand and describe what happened in the past, or is the point to prescribe what should happen in the future? People have been fighting this battle for centuries (again, there's a parallel here between language and music), and we continually go back and forth.

Notation methods carry that same problem with them. What are you trying to do? What problem are you trying to solve? I'm not going to say that modern standard music notation is perfect, but it's a good balance of both prescription and description.

3. Speaking of balance, there's another element that's been mentioned already: creation vs. performance. A good method of notation has to be easy to write, not just easy to read. It has to be specific enough that composers who want to care about details can clearly define them, and I think this method is sorely lacking in that regard.

4. Re: ledger lines. These are a modern, notational "convenience" that was brought about by printing technology making it reasonable. And because performers hated music without them. Music before the 18th century did something that is much more reasonable in my mind, but in practice is difficult to manage. The clef symbols used to be moveable. They sort of still are. The difference between alto and tenor clef is that the c is on a different line.

In pre-Baroque music, there were only 4 staves instead of 5 and the clefs moved all over the place to accommodate the fact that there were no leger lines. The clefs would switch to a different line in the middle of a phrase or just change completely from a c-clef to a g or f clef.

This kind of shifting around is, needless to say, extremely taxing on the performer. But very convenient for the writer and publisher. Leger-lines were originally an acquiescence to the needs of the player, believe it or not. But you know, give someone an inch, and they'll write Strauss.

5. These kinds of visual aids have existed forever in music. But they are only useful for the performers. Writers of music have to consider vertical relationships between notes. I mean, they don't have to, but we're stuck in a musical rut where we've collectively decided that the successful music is going to be based on some permutation of I-V-I. And that's even more true of modern i.e., popular non-classical music since about 1940) than it is of modern classical.

You need to understand the conventions, and visually seeing "difficult" intervals is key to helping you write music that conforms to those norms.

In my opinion, modern standard notation is almost as good as a syntax highlighter in terms of seeing parallel 4ths, 5ths, octaves. It's not quite as good, but it's close.

The proposed redesign is . . . not so great at that.

6. Portability has been addressed by others, but I haven't seen the scalability argument yet. Let's say this works well for one person playing alone. What happens when you try to get 4 people to play together? What about 100? What about 1,000? Someone has to look at all of that and figure out what is supposed to be happening and then lead the group and know when things are going wrong. (I've conducted concerts with 8 harpsichords playing at a time. There is no way this style of notation would work.)

7. I just don't think this is viable as a redesign of piano music. A learning aid? Maybe.

fascinating. nice work.

EUREKA! 1.)learn from what U know - Vim, Zshell, Python and engineering 2.)more MATURE so learn the ORGAN not a piano! 3.)go jazz, play from radio, never bought a book on circle of 5ths 4.)some are slow, tool over 6-12 months to coordinate left and right hands. playing off syncopate all the time. 5.)CLICK! read the book. Eureka! suddenly some of it works. 6.)ORGAN is more expensive but MUCH easier can hold down key, change the instrument,etc.

7.)have not looked at sight read, since either jazz or direct composition. I USED the simple method. tap the beat with feet. use basic drumstick to sound melody with pots whistle the 2nd part, etc. 3 piece band?

8.)hobby only, not stress tested like real code

9.) the auto transfer from guitar chord trellis to ORGAN can be useful.

QUESTION: better to go deep first? and be a genius on organ for Tchaikovsky? or just go radio plus jazz and make it up HACKER STYLE as U go?

Rather than just addressing sheet music, the piano keyboard itself could be improved. See the Jankó keyboard for an example:


Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact