How would you notate the chord G# F B E as a substitute for a G7 chord leading to C?

Question

An example of this sort of sound seems to happen at the very beginning of Joao Gilberto's Aguarela do Brasil. And I have been using it a lot in my guitar playing recently. But I don't know what to call this chord.

I don't have much experience actually writing down chords for other people so looking for a name for the chord that might indicate to another jazz guitarist "hey play " (meaning something like 4x345x in tab notation) if they see it in context (say as a substitution in a 2 5 1 progression).

Edit seems there is some controversy about my spelling of the chord so wanted to provide some context as to why I spelled it this way. Basically the way I discovered this voicing is as a substitution for the dominant chord leading to tonic. So with this specific voicing in this specific context, really the way I am thinking about it is as a G13 with a sharp root. An alteration of the G13 chord.

Answer 1

It's E7b9.

Probably the most known substitute of a dominant chord is a tritone substitution, where e.g. Db7 (Db F Ab Cb), which shares the third (Cb = B enharmonically), and the seventh (F) of the original chord G7 (G B D F).

But if you add b9 to G7, making it G7b9 (G B D F Ab), two more potential substitutes appear: E7b9 (E G#=Ab B D F) and Bb7b9 (Bb D F Ab Cb=B).

Edit: Following the discussion in the comments, as piiperi points, enharmonic substitution of Ab with G# is against the traditional harmony rules. In G chord, the note Ab is b9, and it resolves downwards, to the note G (fifth of the chord C). So one could try to call the chord "G7b9(13) no root". Now, G7b9(13) chord symbol isn't too exotic, but "no root" is. E7b9 is a way more straight forward symbol to describe these notes.

The controversy is that in E, the note G# resolves upwards, to A, not downwards. But in jazz, the classical voice leading rules are not always strictly followed... Though actually the consecutive chord is C6, thus including the note A! So maybe writing G# isn't that stupid?

Edit2: note that in the recording linked by the OP, all notes are a whole tone up w.r.t. this discussion.

Answer 2

(I chose to answer despite your already having accepted an answer because I felt I had some information to offer that is relevant to your question)

A specific chord voicing cannot always be accurately captured in a chord symbol. This is one of these instances. What you basically have here is a shell rootless G13 chord (FBE) with an Ab in the bass. You CAN call it a G13/Ab(no root) but don’t expect it will always sound the way you want if someone else reads and interprets that chord. For example, some players might choose to add a 5th or a 9th to it.

This chord has a diminished quality to it. If you replace the E with a D it is simply a G#o7: 4x343x. There are times when a diminished chord has non-diminished chord tones added to it. If you spell it from the F it is a Fo(maj7)/Ab (1st inversion). A M7 note is sometimes used as either a chord tone or a melody note in diminished chords (“Spring is Here” and “Stella By Starlight*, both in bar 1). You could also spell it from the root: G#o7(b6,no 5). Both of these are accurate but awkward chord symbols so pick your poison. Personally I prefer the G13/Ab(no root). I think this would best be understood by another jazz or Brazilian guitarist and in most instances I think the “no root” designation would be a given.

This chord does resolve to a C in a slightly unusual way because of the G# root but in making my case for the diminished quality you will see it also resolves nicely to Amaj7 or Am7. Keep in mind dominant 7th and diminished chords are closely related.

*In the case of “Stella”, the original first chord is a Idim, not the 2-5 commonly used in fake books

Answer 3

This is a common voicing in bossa nova.

You can see that in different ways depending on the context. For instance in Insensatez:

Am7     (5x555x)
E7b9/G# (4x345x)
Gm6     (3x233x)
D/F#.   (2x023x)
etc...  (How smooth!)

Picture the chord with an E instead of F: you get an E in first inversion (4x245x) to make it more evident.

You can also see it as Db7#9/Ab7 (tritone sub of G7) in a 2-5-1 in C like:

Dm9      (x5355x)
Db7#9/Ab (4x345x)
C6/9     (x3223x)

(see O Pato)

This time imagine it with Db in the bass (x4345x) instead, you'll get the Hendrix chord.

Since you bounce between the root an the 5th on the ii (D/A), the Ab gives more chromatic movement than a plain G13.