social.noyu.me

fell in love with 🎸575585

oh shit 🎸020030->002200->000220 sounds pretty good

i love stumbling upon something first thing in the morning

this is one of those times where we realise that we will have to get relatively deep into the music theory weeds for “good-sounding guitar chords” to make sense

according to some chord identifier web app, this is Em7 -> Esus4 -> Dmaj7sus2\E ^^;;

i am told the app is sus and that this is actually Em7 -> two incomplete variations on A\E and that does make more sense

@hikari I came here to leave this comment

@hikari chord identifier app is sus: calling something with no F# any kind of “Dmaj7” SMDH.

@hikari this is Em7 - A/E, you just don’t get the A chord all at one time.

@steve oh yes that seems right, thank you

@hikari depending on your viewpoint the D in the second half of A/E is either an anticipation of what comes next or it’s just hanging out there being dirty because stacked fourths sound rad

@hikari (the alternative alternative viewpoint is that this is all “just” Em7 and the A major triad on the upper strings is one big suspension over the G major triad that “should” be there in Em7.)

@steve ah, thank you for giving me another reason to distrust it

@hikari the trouble with chord identification tools is they lead us to want to name every stack of notes at each moment in time as it’s own “chord” instead of stepping back and asking “where are we now, and when does that actually change?”

@hikari they’re not all bad, though. @liscio probably has some good thoughts about this.

@steve @hikari are we talking chord recognition (from audio) or “suggest some nice chords” tools? Both might share some components (e.g., transition probability matrices), but they are applied in very different ways.

I happen to have more experience on the recognition side, but am happy to try and answer questions about either.

(Edit: on second read, it might be neither! 😂)

@liscio @hikari I think this is a “here’s a pile of symbolic notes, give me a name for the chord” tool

@liscio @hikari (your “transition probability” comment is spot on, though. For any version of the task, the less musical context you have, the harder it is to give a reasonable answer.)

@steve @hikari Yeah—context is everything. A bag of notes, even when you arrange them by “pitch height”, cannot be reliably named. Rootless voicings, dropping the third, and other such changes to chords in practice make this a fool's errand.

FWIW, chord recognition infers symbols within a fixed vocabulary, and your performance/accuracy is helped if you allow it to say "not in the vocab.” A polyphonic note detector that names the output pitches would be painful to use in practice.

@liscio @hikari

Me: shows a jazz piano player (F# E C# B)
Jazz player: “of course, Dmaj”

@steve @liscio @hikari ooof. I see it, but as a non-Jazz, non-piano player, 😮

That being said, I love that someone looks at that and can say, Dmaj.

@woolie @liscio @hikari
- Piano pretty much never plays the root unless you don't have a bass player (because the bass will cover it)
- The fifth doesn't really add anything, so why bother
- F# C# would be the most minimal "shell voicing", but that's just an open fifth so it sounds boring and makes voice leading awkward
- So toss in the 6th (B) or 9th (E) or both, they don't really change anything for a maj7 but open up opportunities for voice leading to do something.

@steve @liscio @hikari guitar players have similar ways of making chords, although different choices. Fascinating!

@steve @liscio yep

@steve yes this makes sense. chord naming is also inherently a question of interpretation, because everything is an inversion of something else or something else with something added or removed