Let's talk notation, hacker style!

Ask general music theory or songwriting questions, get feedback!

Moderators: MattKingUSA, khz

S1gmoid
Established Member
Posts: 23
Joined: Thu Jan 23, 2020 8:25 am

Let's talk notation, hacker style!

Post by S1gmoid »

Hey. :) As a hacker and software expert, when I really understood traditional musical notation, and how the grand staff works, I realized that we've been stuck with centuries of accumulation of technical debt. It's like as if all database work in the world would have to happen in DBase II because it's sacred, because Mozart used it. :P And of course DBase II is completely useless for modern workflows, so there's all these crazy hacks and "improvements" that are now an order of magnitude larger in volume than the underlying legacy tech. :P

Of course, I do think it's important to be musically literate, and to be able to read and write traditional notation to communicate with others. However, I was hoping to find some alternative so I don't actually have to use it when I'm notating something - it'd be like having to use katakana to write your next novel in English. ("I-TSU-WA-yu-DZU-BII-RA-I-KU-HA-U-i-N-GU-TSU-YUU-ZU-KA-TA-KA-NA-TSUU-RA-I-TO-YUU-RU-NE-KU-SU-TO-NO-BE-RU-I-N-E-N-GU-RI-tsu-SHU" :? )

So my purpose with this topic is to kick off some discussion on alternative notation systems, and what people who do songwriting would prefer. I have some ideas what I'd like, but that's just me.

On traditional notation

So let's start with the analysis of the status quo: why I don't like staff. Because it was never meant to do what we made it do. It was originally invented to notate specifically the C-Major scale (and consequently the A-minor scale, as well as the other church modes of the C-Major scale.) This was fine, for a while. For a while, C was determined by whatever the organ of that specific church was tuned to - and when singing alone, it was whatever you decided it was. It was a movable diatonic scale, and people played and sang the different modes of that diatonic scale. It was pretty fine indeed.

But then it stopped being fine, when people started to compose and play in different keys. I can see how originally, the "distant keys" warranted their afterthought status, given their lack of tonal clarity, but with equal temperament, all that stuff went away too.

And well, I guess the biggest testament to how broken this thing is, is that most musicians can't read it! I mean, seriously, how is it that there are people who earn their living through making music, and yet refuse to learn the "accepted" written language of music? The people who reliably read staff in today's world of music are classical musicians, jazz scholars, and producers. Most rock songbooks don't have melody lines, possibly because a good percentage of rockers wouldn't read them anyway - they just sing from memory.

As for rhythm notation, I'm surprisingly okay with it. Of course, it would be nice to have something that handles tuples in a more "native" way, as opposed to the dirty afterthought way we notate them, but I think it's actually superior to many of the proposed alternatives.

So I went and looked around for alternatives to use. There were a few things I liked, and a few I hated with a passion. :D

Klavarskribo :cry:

Oh dear, this thing. It is probably the alternative notation with the biggest media presence. It's on the old-ish side, and seems to have a cult following. I personally dislike it to the point of considering it inferior to traditional staff. Fully positional coding of the 12-note scale is probably a horrible idea. Humans are not that great at deciphering positional encoding at all. Our limit is at about 7 positions (see short term memory, 7 plusminus 2 items, etc.) 12 positions just flows into a big graphic mess. Meh.

Dodeka :roll:

Okay so apparently some people thought that Klavar wasn't machine-like enough, and replaced symbolic rhythm coding with, guess what, positional time coding. Look at a Dodeka sheet. It's literally just a piano roll. These people "invented" the piano roll on paper, and put a trademark on it. Rotfl. Okay I might be a little too sarcastic here, and I know some contemporary musicians live and breathe the piano roll in their sequencer or DAW, but I'd rather put sharps and flats in front of every single note than having to play from one of these. Seriously, I challenge anyone to reliably distinguish a three-eights (quarter dot) note from a fourth note in Dodeka (in music you have to sight-read and play at first sight, not in music you already know).

We're not machines. Machines are great at moving along a punchcard or a datastream at a steady pace, and so positional encoding is natural for a machine. For us, it's not. We can decode symbols in moments. Positions we need to actually count and compare. Psychology has catalogued several bookfuls of distortions in our visual perception that make the decoding of lengths and distances completely unreliable.

Hummingbird :?

Another alternative notation with a media presence is Hummingbird. It looks cute, with which ends the list of things I like about it. They didn't change the logic of the staff at all, Hummingbird is still a notation for C-Major and A-Minor, with added hacks for other keys. It has a bunch of mnemonics which make it easier to read, I guess, but it's still the exact same thing with a reskin. It's like a font for dyslexics, rather than a new, more logical alphabet.

Pitch bracket notation :lol:

"What IS this f****** thing?!" is what I found myself saying, impersonating the "Boston Fish Guy", as I first saw this... "notation" system. It is for the world of music what the Mindfuck language is for the world of computer programming. I seriously can't think of this as anything other than a geeky joke.

Shape notes :idea:

The shape note system has a history. It adds explicit scale degree information to the pitch in a traditional staff, while leaving everything else untouched. It was, and is used mainly in church choir songbooks in the Anglo-Saxon world. Now its long history of adoption shows that it is really useful, but again, the fundamental issue with the staff has not been addressed.

However, I see a potential use for it in songwriting: By doing away with the staff, and using note head shapes to notate scale degrees in a single line, while using traditional rhythm notation, you get a way to write diatonic music quickly and conveniently. It even allows you to write chords quickly unambiguously. Using sharps and flats would allow for the notation of chromatic notes in diatonic-based music, and of scales like the melodic minor.

There's a big but(t) here though: the accepted head shapes are not very natural for handwriting. I suppose most shape note score (ie. in church songbooks) is typeset and printed, rather than written. Designing a variant with more hand-writable (but still recognizable) shapes would make this really, really useful indeed!

TwinNote TD :o :wink:

Aw finally! A chromatic notation I actually like! Here's why I like it (much better than most alternatives, including Clairnote, which was developed by the same guy). It's chromatic without being a space hog. It combines symbolic and positional coding efficiently, with 2 symbols over 6 positions in an octave. It uses traditional rhythmic notation unchanged, which I believe to be superior than any of the proposed positionally coded alternatives. It's readable and writable for humans.

Personally, I prefer TwinNote TD over regular TwinNote (Regular TwinNote uses filled and empty note heads to add redundant pitch information, rather than rhythmic information, while TD is fully consistent with traditional rhythmic notation.)

Currently this is the system I see myself using. It's not perfect, but it's good enough. It's what I'd call a Minimal Viable Solution for my musical notation problem.

Going further

There are various scientific and artistic notations developed to describe atonal, microtonal and non-Western music. These are either extremely abstract and ambiguious, or are primarily for recording and scholarly analysis rather than communication between musicians. Still, I wonder what could be improved in terms of notation...

Like, marking intervals semantically, allowing for playing them pure if possible... (Or, allowing for marking an interval intentionally impure, such as in the distant keys of Bach's Wohltemperiertes Kalvier.) I've been thinking of a potential notation that notates intervals rather than notes - it could be a fully symbolic coding of music with minimal positional component, which should make it very intuitive for humans to read. Or, another thing that would be nice is a way to notate polyrhythms as first class citizens.

So far I haven't really seen solutions proposed to such goals... I wonder if any of you have ever considered something similar. And do you use an alternate notation system for your personal needs? Which one do you like most?

merlyn
Established Member
Posts: 594
Joined: Thu Oct 11, 2018 4:13 pm

Re: Let's talk notation, hacker style!

Post by merlyn »

I appreciate the heads up on these alternative systems.

You could apply the same arguments to written English. Like bow, bow, and bough. Nobody would suggest scrapping English but it can improve over time. I read notation every day and I think it's pretty good. :D

Like the design of a violin or a hammer some things hit on a workable format and stay there.

I would be interested to hear examples of how these alternative systems make reading easier in practice.

S1gmoid
Established Member
Posts: 23
Joined: Thu Jan 23, 2020 8:25 am

Re: Let's talk notation, hacker style!

Post by S1gmoid »

What I find the main advantage of alternative systems is isomorphism: that the same interval always looks the same. And it's not even the reading part that I'm particularly interested in, but rather the writing part: being able to notate music without losing too much focus to the technicalities of the staff. (I find it funnily appropriate that they call it 'engraving'. That word carries overtones of a slow, deliberate and technical activity, which I think writing traditional score really is like for most people.)

User avatar
milo
Established Member
Posts: 421
Joined: Wed Sep 06, 2017 2:55 am
Location: Southern Utah, USA
Contact:

Re: Let's talk notation, hacker style!

Post by milo »

Does tablature count? I can learn a song from guitar tablature 100x faster than from standard notation, because tablature mimics the way I think about how the music is played. Guitar music is about where I put my fingers on the fretboard, not about what the names of the notes are.

What do you think of Piano tablature? https://www.wikihow.com/Read-Piano-Tabs

I realize you were probably looking for something more abstract and generalizable, but maybe the answer is that there is a different preferred notation for every kind of instrument. Or for every kind of musician. Being the Linux geek that I am, I'm happy to have so much choice!

S1gmoid
Established Member
Posts: 23
Joined: Thu Jan 23, 2020 8:25 am

Re: Let's talk notation, hacker style!

Post by S1gmoid »

I think it counts, if it's what you're most comfortable with. I prefer a representation that tries to capture the "soul" of the music for general use, but I do see the point of tabs. It's quick and dirty for both notating and playing by sight.

Openmastering
Established Member
Posts: 36
Joined: Thu Dec 05, 2019 10:27 am
Location: Tübingen, Germany
Contact:

Re: Let's talk notation, hacker style!

Post by Openmastering »

Hey there,

First of all thanks a lot for talking about this.
It just opened a new world of possibilities.

I'm a classically trained violinist, so sight reading music notation isn't a problem for me.

I see your point about trying to improve upon the traditional system. I really like the 6+6 paradigm because of the consistency between intervals. The clairnote system fits me better because it's hard for me to switch to the triangles in the twinNote system. And yeah, I like the TD version better because I don't have to rethink the entire music.

As far as I know, once we've learnt to read, we recognise words as a pattern and don't spell them anymore. I wonder how this could be integrated in music notation.
Audio mastering for open minds: https://openmastering.studio
I master music using only free/libre softwares and DIY hardware.
Contact me for a free mastering sample.

bitsnpcs
Established Member
Posts: 90
Joined: Tue Jan 21, 2020 2:02 pm

Re: Let's talk notation, hacker style!

Post by bitsnpcs »

.
Last edited by bitsnpcs on Fri Mar 20, 2020 1:56 pm, edited 2 times in total.

S1gmoid
Established Member
Posts: 23
Joined: Thu Jan 23, 2020 8:25 am

Re: Let's talk notation, hacker style!

Post by S1gmoid »

... I don't see why drum notation would require automated translation. Honestly, I don't think it's a very good idea to overly rely on computers. I don't want a way so I wouldn't have to read classical notation - I'm hoping for a system that could be simpler to capture ideas in.

User avatar
CrocoDuck
Established Member
Posts: 1089
Joined: Sat May 05, 2012 6:12 pm
Contact:

Re: Let's talk notation, hacker style!

Post by CrocoDuck »

milo wrote:Does tablature count? I can learn a song from guitar tablature 100x faster than from standard notation, because tablature mimics the way I think about how the music is played.
I'd be very careful with tablature, as it has very severe limitations. The first one I can think of is that it is very hard to write compositions with tablatures that are intelligible to other instrument players. That is, tablature is not general enough (clearly) to allow other instrumentalist to easily read and understand your stuff.

I think this is a pretty big limitation as soon as you are composing for more than one instrument, as having musicians able to understand the whole piece actually helps them a great deal, especially when things turn complicated. Also, being able to read other instruments parts really helps you to gain a higher understanding of why a piece is arranged the way it is (if the person who wrote it notated in a reasonable way, especially when it comes to make clear how the rhythm should be interiorized). I think this really does help learning music beyond the confines of your instrument.

I am definitely not an expert about notation, and it has been years since I last read a score. I actually not sure whether I can do it anymore. But I remember that when I got able to read score in ordinary notation I never looked back at any tablature: it is like there is much more information in scores. I think it only took overall a full year of practice to get to the point I could read through scores and understand their contents quickly, like properly reading, not painfully parsing them bar by bar (or figure by figure). If you compare it with the 2-3 years you need to get basic proficiency with another language it doesn't really sound that bad.

Adam Neely mentions notation in few different videos. From the top of my mind: https://www.youtube.com/watch?v=ua-N1JuqO5M
Check my Linux audio experiments on my SoundCloud.
Browse my AUR packages.
Fancying a swim in the pond?

User avatar
thumbknuckle
Established Member
Posts: 35
Joined: Sun Oct 19, 2014 6:16 pm
Location: Western Massachusetts
Contact:

Re: Let's talk notation, hacker style!

Post by thumbknuckle »

S1gmoid wrote:I think it counts, if it's what you're most comfortable with. I prefer a representation that tries to capture the "soul" of the music for general use, but I do see the point of tabs. It's quick and dirty for both notating and playing by sight.
Is it though? With regular notation I can put a chart in front of my band and count to four and have music come out. With tab one needs to listen to the song to get the rhythm, which begs the question "if I need to listen to the song anyway, what good is the tab?"
Faster than a laser bullet.
Louder than an atom bomb.

User avatar
milo
Established Member
Posts: 421
Joined: Wed Sep 06, 2017 2:55 am
Location: Southern Utah, USA
Contact:

Re: Let's talk notation, hacker style!

Post by milo »

Tabs can include rhythm notation, but I think you are correct that most people who rely on tabs are the kind of people who play by ear.

One nice thing about tablature is that it is unambiguous. Depending on how many frets your guitar has, you may be able to find the exact same note in 5 or more places, and standard music notation doesn't tell you which one to choose. Which one do you play?

E---0---------------------------------------------
B--------5----------------------------------------
G-------------9-----------------------------------
D-------------------14----------------------------
A--------------------------19---------------------
E-----------------------------------24------------

The point of tablature is not so much to write down what the music is, but how it is played.

User avatar
CrocoDuck
Established Member
Posts: 1089
Joined: Sat May 05, 2012 6:12 pm
Contact:

Re: Let's talk notation, hacker style!

Post by CrocoDuck »

milo wrote:One nice thing about tablature is that it is unambiguous.
Yes and no.

Yes: in theory that really works well.

No: after I got a bit experienced I found that when I played things as in the tabs, especially chords, they would not sound quite the same as in the recording of the song the sheet was supposedly transcribed from. Also, very often tabs would be unreasonably hard to play. Turns out that In most of my tabs the person that transcribed them got most of the chord voincings wrong. I guess this could be for two reasons: either they were poorly transcribed by ear, or they were automatically generated from a score transcribed by traditional means. So, I guess that tabs are as unambiguous as the person writing them makes them to be, which unfortunately dilutes the pro they have about pinpointing exactly where the note has to be played. Clearly, not a con about the tabs themselves, tho. The risk here is to be forced by the tab to learn things as someone else understood them, rather than how they are. Which isn't necessarily wrong, but normally studying a piece yields much higher results when you figure it out on your own.

As for ambiguity in ordinary notation, it can definitely happen but chords make it much harder: the notes are notated at the octave they are supposed to be, and this normally can happen only with selected fingering. When there are few chords the correct voicing bubbles up as the wrong ones are simply unplayable (although I reckon it is possible to construct examples in which there are multiple easy to play possibilities). At the end, this is how I pretty much dealt with my broken tabs too, but it took me a while to realize the tabs were broken as I was way too biased by where they told me to put my fingers.

Now, it might seem very hard to be able to parse the proper voicing out of different possibilities on the fly, but it isn't. It is kinda like when you read ordinary text and your brain already figures out the whole word from the first couple of characters and the context of the previously read text. For example, when I was studying ragtime music straight from the sheet the independent bass line was all I needed to figure out where the rest of the chord was, and there was only one option that would allow to play the body and the bass with the correct timing. Rather than thinking of all of the combinations, I remember my fingers just moving down to where it made sense as I read the cluster of notes.

That of ambiguity might be also a feature more than a bug at times. I mean, no, it is a bug, but it has useful ramifications when it gets to learning. For example, I think it unlocks the capability of mopping up mistakes by increasing awareness of how you can play the same thing in different ways. I found it easier to recover from a mistake when I was able to climb my way back by playing around where I got stuck to build a "mop up" phrase to get back to business rather than forcing me to go somewhere else on the neck. Also, it might force you to retro-engineer an arrangement: what kind of chord voicing is the most likely to be used here given the context?

So, back to the question:
milo wrote:Does tablature count?
Yes, I would say. It has its cons, but that is true of any notation, although I think the traditional one has the power-feature to apply to most instruments with little to no variation. That one on its own makes, at least for me, tablature not that much preferible. As for the other options discussed above, I am really not familiar with them, so I cannot comment.
Last edited by CrocoDuck on Thu Feb 13, 2020 8:20 pm, edited 1 time in total.
Check my Linux audio experiments on my SoundCloud.
Browse my AUR packages.
Fancying a swim in the pond?

S1gmoid
Established Member
Posts: 23
Joined: Thu Jan 23, 2020 8:25 am

Re: Let's talk notation, hacker style!

Post by S1gmoid »

Hm that's an interesting point you make about broken tabulatures...

I'm not a guitarist so I don't really know how prevalent the use of tabulature is to begin with. I'd think an intermediately experienced guitarist would probably be able to "instinctively" voicelead based on how their hand and fingers are comfortable moving on the fretboard, so I wonder if tabs are reserved to teach beginners (and as such, mostly left as an afterthought)?

BTW, strangely enough, we're back to an oral tradition in popular music. It makes some sort of sense that that written records of rock music may be less than accurate. Like how in early church music, the role of notation was not to capture the intricacies of the music, but rather just to remind the singers of something they already know. I believe nobody would endeavor to play a cover of Sweet Child of Mine without having heard the original. If any written record sounds strange compared to the original studio recording, the "oral" source would take precedence... :D

Basslint
Established Member
Posts: 246
Joined: Sun Jan 27, 2019 2:25 pm
Location: Italy

Re: Let's talk notation, hacker style!

Post by Basslint »

S1gmoid wrote:Hm that's an interesting point you make about broken tabulatures...

I'm not a guitarist so I don't really know how prevalent the use of tabulature is to begin with. I'd think an intermediately experienced guitarist would probably be able to "instinctively" voicelead based on how their hand and fingers are comfortable moving on the fretboard, so I wonder if tabs are reserved to teach beginners (and as such, mostly left as an afterthought)?
I am an amateur guitarist, guitar tabs are very common and definitely not reserved for beginners. Using traditional notation for a guitarist is not very useful, because they'd have to look ahead (or be well acquainted with the piece) to use efficient fingerings. Maybe using notation with chord diagrams for the fingering might be more useful, but IMHO nothing beats tablatures, also because on guitar the same note sounds different if played on different strings and things like legato can be played in different ways which sound different, so if a guitarist has to be accurate, they need tabs.
The community of believers was of one heart and mind, and no one claimed that any of his possessions was his own, but they had everything in common. [Acts 4:32]

merlyn
Established Member
Posts: 594
Joined: Thu Oct 11, 2018 4:13 pm

Re: Let's talk notation, hacker style!

Post by merlyn »

Basslint wrote:Using traditional notation for a guitarist is not very useful, because they'd have to look ahead (or be well acquainted with the piece) to use efficient fingerings.
Music specifically for the guitar includes positions, fingerings and strings. For example this is Romance a well known classical guitar piece :

Image

The roman numerals with dashes tell you to use a barre at the appropriate fret. The 1/2IX means a half barre at the ninth fret. The numbers tell you which finger to use and the numbers in circles tell you the string.

Post Reply