Hey everybody,

The 4th Horseman requested that I do this a long time ago, and it (due to my life getting in the way and the fact that real book charts take about 5 minutes) has been swept under the rug.

Consider it unswept.

I've attached a guitar pro file, because that's all we have to go off of.

Let's begin. I'm taking this bar-by-bar, so follow along.

Bars 3-6

Our guitars enter here and outline this progression:

F#dim/A - Gmaj7 - Em9 - F#dim7 (see #2)

VIIdim (V7*) - Imaj7 -VIm9 - VIIdim7

This is a tonal progression in G major. Two things of note here:

1. VIIdim is basically also expressed as a rootless V7, so this all but confirms G major for now.

2. I'm calling this F#dim7 for now, but it could also be VIIm7. The elusive 5th has not shown up yet, and we can interpret this multiple ways. We can look at this as being borrowed from G Lydian, thus "blurring" the status of G as I.

However, Occam's razor says its probably the same chord as before, and this is confirmed later.

Bars 6-10

This is actually the same chord progression, but the guitars have switched to a more dense texture with passing tones. The bass is moving now.

Note the 6/4 time sig, and the harmonic rhythm.

Our eighth notes are beamed in the tab as 6+6, but the chords are changing 7+5, perhaps it'd be better beamed as a non-shuffled 12/8? Who knows? I don't!

This keeps the tonic (Gmaj7) and dominant chord (F#dim7) on weak locations in the bar. This helps create the hazed, swirling effect of the melodies, and the lack of strong resolutions.

Bars 11-16

The same stuff as before, however one guitar switches to chords and confirms the F#dim7. The guitars then break back into the more complex riffery until...

Bars 16-24

One of the guitars moves to an upper harmony and extends the chords to:

F#dim7/A - Gmaj9 - Em11 - F#dim7 (with upper tones)*

* I don't know if I'd call the notes over this chord chord tones (they're mostly in passing), but dim7 with upper neighbor tones is a common sonority in a more modern setting, so more power to you.

Bars 24-28

We change into 4/4, and outline the following:

Cm/Eb - Bb6/9* - Cm9 (Bar 28) Gm/C - Cm/Eb - Bb/F

*This chord is expressed quartally (4ths) in one guitar. Quartal chords are open to many different interpretations(any note can be the root, even ones not voiced ), but here its use with the bass implies this chord.

Note the hybrid voicing of Gm/C, giving us Cm9, without the third. Here's an analysis.

All the Cm chords are II chords and the Bb chords are I.

Not only have we modulated to C minor immediately, but something cool has happened. Note the lack of any cadential chords, and the alteration between Cm and Bb.


This section is a true modal framework in C Aeolian. We alternate between a tonic chord and the cadence chord, a Bb, containing (one of) the character pitches of C Aeolian, Bb. There is no functional harmony in this passage.

Bars 29-32

Now things get fun.

We have a unison(almost) line in the guitars that seem to imply a quick loop of:

Cm7- F7 - Cm7 - Gm*

Im7 - IV7 - Im7 - Vm

The unison riff in 32 implies a Gm chord.

Now, this is another modal vamp, the IV7 is a dead giveaway for C dorian.

Also of note is the Gm chord. Although the lack of (A) in this chord limits its use as a cadence chord, allow me to remind you at no point is Gm actually played, rather, we have an 8 note riff that revolves around scale degree V as our center, function has still been avoided because there isn't really a chord there that makes a ton of sense to name.



Ebmaj7#11. Don't believe me? Look at the riff. It's all there. Watch the accents, it helps determine chord tones.

This is bIII in Cm, but we are modulating to...

Bars 35-40

Immediate key change to what appears to be C# major.

C# - B - F#

I - bVII - IV.

Note the plagal cadence; no mixolydian here.

The second time around, we get a sweet trick instead of the F#. We replace it with:

C# - B - Eb - F

I - bVII - V/V (OR II we can't confirm or deny either) - (III???) constant structure

Note the third relations here. Eb (D#) is a major third above B. In addition, the F (NOT a III in C# major, but part of our constant structure chain here...), works perfectly as our pivot chord to get back to...

Bars 41-46

C Dorian. This is the same music as bars 29-34 with a counter melody.

Bars 47-End

Here you have almost the same music as 35-40 with two differences.

1. One guitar is playing some extended chords. They appear dissonant, but that's because they are bringing out both 5 AND #11, extending each major chord in our collection to a quasi Lydian sound. Some Steve Vai crap right there.

2. Our constant structure pattern continues for a two bar coda outlining a series of major triads.

Eb - F(7) - Ab - Db - D

These chords may be hard to find (the accents flipped, and the last three are for two notes each) but they are there.

After you hit the Ab, each of those chords is best though of as I in its own mode.

Anyway, that's the tune.

I apologize to 4th for making this take forever, and I hope I didn't miss anything. Seeing as there's no real strumming to be done here, I had to derive everything from the melodies. Feel free to ask tons of questions, this is a confusing one compared to a Real Book chart.
work in progress.gp5
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp
Wow, thanks a lot man. I figured this would probably be difficult to parse through. I just did a quick skim through and it looks great, but now I'm going to work my way through every chord and then get back to you with questions.
Yeah. No prob.

I also want to note that I'm inclined to say more things are modal than they may actually be, due to the fact that you have a large amount of fourths and the piece is more "riffage" than "chord progression"
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp

F#dim/A - Gmaj7 - Em9 - F#dim7
VIIdim (V7*) - Imaj7 -VIm9 - VIIdim7

This works for me, I've been very unsure about where the key center was on this one and what to even call the first and last chords because, as you said, the accents keep everything pretty hazy. Seeing that the first chord is F#dim instead of F#m or some kinda Adim chord makes it pretty obvious G is the tonic. I honestly had no idea how the rhythm should be notated so I just left it at guitar pro's default for 6/4. I think it should be beams of 7+5, and maybe in 12/8 as you said.

For the higher harmony section I simply thought, "I wonder what harmonizing a third up would sound like?" then tried it and liked it. I would consider them more extensions of the original chords than passing tones.

Bars 24-28:
Cm/Eb - Bb6/9* - Cm9 (Bar 28) Gm/C - Cm/Eb - Bb/F

Wow, I didn't realize there was a Gm in there, and I definitely didn't know I'd written something that was actually aeolian. I was just trying to create a little tension builder to transition to the next section. Cool stuff.

Bars 29-32:
Cm7- F7 - Cm7 - Gm*
Why Cm7 and not like Cmadd9? I don't see a Bb in there until the next bar, which you're calling F7.
I didn't understand how bars 30 and 32 could be different and was about to be very mean to you , until I remembered the bass is playing a descending F triad in bar 30 and realized that's what you meant when you said there's a unison in bar 32. Why not call the Gm a Gm7? That riff contains and even ends on the F.

Bar 34:
wow, I never would have thought there was a chord there, but there it is.

Bars 35-40:
It was really hard for me to find the F# chord in there. The F# section contains every note in the F# major scale except B. The Eb and F chords were hard to see as well, but I get it now. What do you mean by constant structure, and how does bar 40 fit in?

Yea, I wrote the second guitar part about a week after I had the main riff and had no idea how to build chords off of it and just used my ear and went for a lydian kinda sound.

Eb - F(7) - Ab - Db - D
I honestly really like this part and cannot remember at all how I came up with it, except that I was trying to sorta match what was going on with the other guitar, but not. I'm surprised you were able to find chords in this section, but of course, after looking at it for a few minutes I see the little bits and pieces of chords in there. Again, what's constant structure?

Thanks lot for doing this. It's really helping me see what's happening a lot better. I knew there was a lot going on, but I wasn't sure exactly what in a lot of places. Now that I know where the piece has been I just need to figure out where it's going next. I've been thinking I should add something that grooves more (something that's not straight 8th notes) right before the modal riff.
You don't need to have chords to have chords. It's more or less implied by the melody anyway.
Quote by Elintasokas
You don't need to have chords to have chords. It's more or less implied by the melody anyway.

yea, I knew this, but there are a couple of spots in this where it's kinda surprising to me.
1. It should be a Gm7, I must have forgotten to type it. The Cm7 could also be Cmadd9, the jazz guy in me took over and slapped 7 on everything.

We just want to make sure the 9 is actually a chord tone, and not a passing note before we label it add9.

2. Constant structure is when you have a chain of chords of all the same type (maj7, min, etc.). In your example we have a large chain of major triads through various root movements. The progression is given unity by the chord qualities and structures, not the key or function. So every chord is usually I once the structure gets going and leaves the key.

In bar 40, you use V/V to start the structure and plane it up a step to F, before modulating back to Cm.
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp
Oh, so basically even though the chords are leaving the key pretty far behind keeping the chord quality the same still gives it structure?

Does it make sense to call that Eb a V/V without actually having the V? Why call it that instead of the II? Just because it doesn't go to the actual V after that? I never did get very familiar with secondary dominants and how they work, I'm just kinda aware that such a thing exists. And now I just noticed a little weirdness with the note names in the
C# B F#
C# B Eb F
Should the Eb technically be D# and the F be E#? Or am I missing something?
1. Yeah basically. You still have logic and unity. It's the harmonic equivalent of planing triads through key centers.

2. You can call it either. Here's why:

The difference between II7 and V7/V is that the latter has dominant function, and the first does not.

Dominant chords tend to resolve down by P5 (V/x) or half step (SubV/x).

If a 7th chord with a diatonic root moves to another diatonic root in a way atypical of dominant chords, it is likely a (x)7 and not a V7/(x).

We thus label it with its actual roman numeral and not V/something.

The exception to this is a backdoor dominant, which is a deceptively resolving dominant chord with a non diatonic root, that often resolves up by whole step (Bb7 in the key of C). We usually label that bVII7, because the other way would be confusing as hell.

This chord moves from a diatonic root to a non-diatonic root, which is typical of secondary and substitute dominants. But it also heads out of the key, so you could make a case for either...

3. It should be D# and E#, I used used the names of the nots in the GP5.

Let me know if that makes sense, I can talk dominants all day if necessary. Just remember the categories of dominants other than V7 and VIIdim7:

1. Secondary = V/x
2. Substitute = SubV/X
3. Backdoor = X7, usually bVII
4. Non-dominant dominants: usually I7, II7, III7, IV7, VI7
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp
ok, I kinda get it. I'll have to do some reading on the different kinds of dominants later, and yea I forget that GP5 doesn't automatically choose the right note names (and how could it anyway?) since I tend to just use the tab part and ignore the notes other than the rhythm. Reminds me that I really need to get better at reading sheet music, but whatever.

Again, thanks a lot man. I found this really helpful and I learned a few things.
Hey any time, it's no big deal.

Using the same numbers as my last post. Here's how a G7 would resolve in each category:

0. Regular V: G7 -> C

1. Secondary: G7 -> is V of something that isn't I.

2. Substitute: G7 -> Gbmaj7

3. Backdoor: G7 -> Amaj7

4. Non dominant: G7 is diatonic and moves to a diatonic root with motion that is NOT #s 0-3.
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp
Does a secondary dominant have to go to something that's not diatonic? Otherwise there's some overlap between non dominant and secondary dominant.
No it doesn't. In the progression C (I) - C7 (V/IV) - F(IV) - G(V) - C (I) the target chord (F, IV) is diatonic and C7 functions as a secondary dominant.

Another example: C(I) - E7(V/vi) - Am (vi) - A7 (V/ii) - Dm7 (iim7) - G7 (V7) - C (I). Two secondary dominants in this. Both move to diatonic chords.

And I think it's better to call "non-dominants" non-functioning dominant.

I guess an example of that would be something like: Cm - D7 - Ab - G7 - Cm. Sounds better as Cm - D7 - Db7 - C, though, but that would be #2.

Most of the secondary dominants you're gonna see DO move to diatonic chords, but secondary dominants are also a nice device to do modulations that are less abrupt than just switching key without warning. They have that strong V - I pull.
Last edited by Elintasokas at Mar 25, 2015,
^This. Jani's got it down.

Here's the most blatant example of a non functioning dominant. So many songs (Eight Days a Week FTW) go:

I - II7 - IV - I

That second chord in this case is II7, not V/V.
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp
Oh I get it! For some reason it just didn't click when I read that. Basically a non dominant is a dominant 7th chord that doesn't actually function as a dominant. Calling it a "non-functioning" dominant got it through. For some reason the name non dominant didn't fully register in my head so I was still expecting it to function as a dominant in some way.

Why does it have to be diatonic? What about a non-diatonic non-dominant like a I bII7 II V I, or his example with the Db7? (do you mean b2?)
They're usually diatonic, because they need to have some other function in the key in order to be...well...functional.

If they aren't going to have dominant function, what's left?

bII7 doesn't tell us much about what's happening. It'd be more enlightening to say that chord is SubV and is resolving deceptively up a half step.
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp
Yea, I guess non-diatonic non-dominant would be getting pretty non-functional. I was just trying to think of a non-diatonic non-dominant and the first thing I thought of was bII, but it looks like I missed my mark lol.
There might not be any. Any non diatonic dominant is either going to be V/X or SubV/X.
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp
*smacks forehead* You can have substitute secondary dominants! Well, I think you mentioned them a few times in this thread, but I just now got it.

I think the last thing I'm unsure on is the backdoor dominant. I just did a google search and think I get it, but want to make sure. Basically it's when you replace the V7 of a progression with the bVII7, like II bVII7 I or IV bVII7 I, right?

Now, the question of learning to accurately identify these dominants and other chords with roman numerals. I can easily identify a basic progression, but I get lost in jazz songs. I've also noticed that some people use lower case to indicate minor and others (including you) just capitalize them all and add m to indicate minor. Which is the more commonly used practice? Also, I know I'm heading pretty off topic at this point, do you think I should just start a new thread? I'm starting to veer more towards how to analyze in general, like how to handle modulation and such.
Lol nah it's fine. It's all relevant, anyway.

Yeah that's basically a backdoor dominant. Here's a little more info from another thread.

3. What the heck is a backdoor II-V?
-Check it out. Picture a classic IV-I plagal cadence:

F - C.

Imagine we want that plagal cadence, but we want to preserve the strong root motion found in a V-I "authentic" cadence. Can we do that? Is it legal? Yeah. We're gonna stick that root motion between the IV and I chords:

F - Bb - C.

Cool, but now we have this weird succession of major chords. Let's fix that, smoothing the voice leading by creating a II-V:


Awesome, but that dominant 7 makes the motion to C a little heavy, maybe. Can't we do something about that whole step resolution to make it smoother? Yep. We can anticipate the 3rd of the I chord and put it into the bVII chord:

Fm7 - Bb7(#11) - C. We now need a Bb Lydian Dominant chord scale. That scale is usually played over both chords, you don't need F Dorian.

That's where a backdoor dominant comes from. It isn't unique to jazz at all.

The "traditional" practice has always been to use lowercase, but there's been a recent shift in the academic circles (probably due to the inclusion of jazz and other styles into the academic world) to use all capitals, and follow it up with a chord symbol.

That way, you don't have to indicate the thirds twice. IV7 and IVm7 can't be confused easily, but
IV7 and iv7? Slippery Slope.......
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp
I just played that Fm7 Bb7(#11) C and totally recognized it. I don't know where from, but apparently I already knew what a backdoor dominant sounds like and didn't know it.

I see, yea it does make more sense to just capitalize everything. Saying iv7 would basically be saying X minor minor 7.

As for figuring out the correct Roman numerals, how do you handle modulation and such? I'll use the pretty simple example of Autumn Leaves.

Am7 D7 Gmaj7 Cmaj7 F#ø B7 Em Em x2
F#ø B7b9 Em Em
Am7 D7 Gmaj7 Gmaj7
F#ø B7b9 Em7 Eb7 Dm7 Db7 Cmaj7 B7b9 Em Em

My first instinct is to label them:
IVm7 bVII7 bIIImaj7 bVImaj7 IIø V7 Im
IIø V7b9 Im
IVm7 bVII7 bIIImaj7
IIø V7b9 Im7 VII7 bVIIm7 VI7 bVImaj7 V7b9 Im

But I feel like the Am7 D7 Gmaj7 is actually a IIm7 V7 Imaj7 in G major, but if I call it that it's obviously going to look like F# B E instead of A D G, so I guess it makes more sense to call it IVm7 bVII7 bIIImaj7. Or would you even do something like call the "I" on the previous section V/IV and so then you can call the A D G section II V I, but then if you did it that way I don't see how to get back to E=I again since right after that is a II V I in Em. And what about songs where there's an actual key change, not just a short little modulation to the relative key?? How do you know both as the person writing the roman numerals and the person reading them at what point your numerals are in relation to a different tonic?

I also see that the Em7 Eb7 Dm7 Db7 Cmaj7 section is basically a cycle of 5th progression where the Eb7 is a sub for A7 and Db7 is a sub for G7, so it could basically be seen as Em7 A7 Dm7 G7 Cmaj7, but I have no idea how to correctly label it in roman numerals.

edit: also do you know of an easier way to to talk about the half-diminished chord using a keyboard? m7b5 just feels clunky to me, but I don't like having to find the ø symbol and copying and pasting it every time. Maybe I should just get over it and use m7b5.
Last edited by The4thHorsemen at Mar 26, 2015,
Your'e overcomplicating Autumn Leaves.

It's literally a II-V-I in G, followed by a II-V-I in Em.

Remember G major and E minor are not the same key. Ever.

That "chain of chords" is just an extended progression tonicizing C. Behold:

Em7 - Eb7 - Dm7 - Db7 - Cmaj7

III- SubV/II -II - SubV - I

Now that's just if we don't relate it back to Em. Related to Em, it's a bit different, because Em is clearly I.

What's actually happening is we're taking the chord we want to move to (Cmaj7, aka VI in Em) and sticking it's II-V in front. Then we take another II-V and put it in front of that.

When we add II-Vs, the II usually isn't analyzed as anything other than part of the II-V. By that same token, analyzing it with respect to Em yields:

Em7 - Eb7 - Dm7 - Db7 - Cmaj7

I - SubV/(Dm) - I (Dm) - SubV/VI - VI

Note the First II-V and how it isn't functional long term, so isn't analyzed in the key. You could do it all in the key like this:

I - SubV/bVII - bVIIm7 - SubV/VI - VI

Which, IMO, is less accurate because bVIIm7 is not a thing .

My point being that II-V's in front of other II-V's are usually just tonciziations or part of larger chains, and shouldn't be analyzed with respect to the long-term key.

And on your final point:

In pop/jazz/vertical music, it's m7b5

In classical/horizontal music, its half-diminished with the little symbol.

They don't intersect because in classical land m7b5 implies that the fifth is chromatically altered. It isn't. Not in a minor key anyways, where the chord type is derived/interchanged from.
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp
Yea, I know Autumn leaves is pretty simple in comparison to a lot of other jazz standards, which is why I picked it as an example. I think the problem may be that I'm trying to over complicate roman numeral analysis in general.

I get that G major and E minor are two different things, what I'm not getting would even apply if the key's weren't related, basically how do you notate the change from one to the other with Roman numeral analysis without specifying that the first II V I is in G and the second is in E? Like it wouldn't make sense to say it's II V I VI II V I because that would be F# B E C F# B E and it's supposed to be A D G C F# B E.

I guess what I'm asking is, "Is there a way you could make a chord chart of a song with tonicizations and/or actual actual key changes without using any letter names so that you could just read the numerals and play the chords all in relation to any overall key that you want, while showing the chord functions?" Like basically a way to show somewhere in a chord progression that from that point on bVIm=I or whatever. Or do you just have to specify using actual chord names to show that? I'm starting to think maybe that's what my hangup is, numerals are just a tool for analysis and I'm thinking of it as like a combination of notation and analysis, which could be the reason I've hard so much trouble thinking about roman numerals beyond really simple progressions.

I - SubV/bVII - bVIIm7 - SubV/VI - VI seems to make the most sense to me the way I was thinking about it, but I get that the Dm isn't functioning as a bVII, so that's not really accurate. This is kinda tied in with what I was talking about in the previous paragraph.

Ok, I guess I was intuitively thinking of m7b5 as being altered, which of course it isn't if it's diatonic, so it felt wrong to call it that even though I knew it's a common thing to do.
Last edited by The4thHorsemen at Mar 26, 2015,
I just use a third line that shows the key centers and when they change.

Roman Numerals
Modulations when applicable
"There are two styles of music. Good music and bad music." -Duke Ellington

"If you really think about it, the guitar is a pointless instrument." - Robert Fripp
Such a simple thing that's bugged me for so long. I thought there was something more to it that I was missing, but it was just me trying to use it for something it's not meant for. I did already tend to think of things in separate chunks, but I thought I was missing something.

I've learned a lot more about music that will be really beneficial to me in this thread than I have in a long time. I already knew the very basics of harmonic function and kinda sorta knew about secondary and substitute dominants, but the things that got cleared up for me in this thread (particularly that last thing) were holding me back from thinking about functions in a precise manner. Of course there's more for me to learn and actually internalize in this area, but I feel like I'm in a better place to do that now.

I'm thinking maybe I'll try to do an analysis of a song in a few days and make a thread so you or whoever can double check me. Do you have any recommendations for something that isn't painfully obvious but not super hard? I'd prefer if it's something with an easy to find chord chart since my ear for chords is still kinda shaky, especially if they're not particularly obvious or across multiple instruments. If not I'll just look through some jazz standards and pick one.