Modulation Methods and Analysis by Mark Thornbury
(Condensation from his postings in the Ted Greene.com Forums)

I'd like to jump-start an idea here...I'm going to take a tiny kernel and throw out an analysis, ala Ted, and let's see where it goes...

Ted's intro to Old Man River uses some cool techniques, starting with a Max Steiner (among many other film composers) concept..."Dixie" as a relevant thematic concept to place you South of the Mason-Dixon line...and through a beautiful deceptive cadence, using a pivot chord for modulation, takes the key center up a major third, resulting in a rather bright, uplifting sound.

Ted showed me how to do this, as he loved to modulate up major 3rd key cycles, and I'll share it here for those of us who might not have had the chance to check this out...

If you start out in a certain key, at some point just go to the vi chord, and make it a vi6 (meaning a vi minor 6th, not a vi chord with the 3rd in the bass, like Baroque style figured bass), then consider the vi6 to simultaneously be a iv6 of the key a major third higher.

This “up a major third” pivot chord concept looks like this (example in the key of C):

         C  –  Am6  –  E.
C:     I  –   vi6
E:              iv6    –   I      

This works because the Am6 introduces an F# note to the C scale, and while this is not diatonic to the key, Ted pointed out that it is actually more natural to the key center in that the overtone series naturally has a raised 4th degree, so it doesn't sound all that weird.  As a sidelight to all of this (as Ted would say), the F# note is only a whole step away from G#, which is the third of E major, and hence sounds rather natural moving up to it.

Analysis:
B:    I  vii  vi7  I/5  IV  vi7/5  ii7  vi6/6
Eb:                                                iv6/6  I/3  vi7  iii7  IV  #ivm7(b5)  V  vi6  ivm6  I  vi  I  IV  I
Does this make sense?

[Comment from Steve Brodie:  Another approach to modulations, other than the anchor chord method, excellently explained by Mark, is what I will call the “cycle method.”  Follow the diatonic sequence of 4th progression (I – IV – vii – iii – vi – ii – V – I), but instead of going to the IV, go to #iv, and treat as ii of the new key, followed by the V7, and then to the I.  Then start the pattern again in the new key.  Again, the key changes are upward in major 3rd cycles.]

[Mark:]
That sounds right to me.  In fact, I remember this is a 'cousin' variation on the same idea, in that Ted liked to go to the #ivm7(b5), then consider it to be a iim7(b5) to a V of the key of III, which is almost exactly what you have described!
I remember Ted showing me something like going from I to IV to #iv7 (with a natural 5) and proceeding as you described.  This is great stuff!  

[Comment from Steve Brodie:  Ted also got me into cycle of diatonic 4ths, but when you get to the V chord, you change it to minor, and treat it as ii – then you’re in a new key a 4th higher.  You can go on forever that way.]

[Mark:]
I think the sounds are quite similar to "Bird" changes, although the chord "qualities" are diatonic while moving through the cycle, whereas "Bird" changes are when you take a normal tune and add a lot of ii-V's in them through back-cycling.
A good example of this is the tune "Blues for Alice" which takes a normal 12-bar jazz blues and inserts a long series of ii-V's into the progression.

Here are a couple examples of taking a normal progression and adding some cycles:

Example #1:
| Dm7          | G7           | Cmaj7 |  
becomes:
| Ebm7 Ab7 | Dm7 G7 | Cmaj7 |

The above example works whenever you have a ii-V over two bars.  

Example #2: (first 5 bars of a blues in F)
| F7 | Bb7       | F7            | F7         | Bb7 |  
becomes:
| F7 | Em7 A7 | Dm7 G7 | Cm7 F7 | Bb7 |

The Bb7 in the above example is the target chord, and a cycle of fifths is simply added before it, taking up however many bars desired. This is what is called "back-cycling" because you're cycling back from a target chord through the cycle of fifths. 

[Comment from Hal:  One of things that interested me was the method of using a common chord tone on a chord that is implied, but not played.  On a ii-V-I, you can dump the I and go to any chord that has a common chord tone with that same I that you did not play.]

[Mark:]
I missed that lesson, but I was just experimenting with the concept, and it seems to work, especially if you anticipate the target note and then fill in the rest.  For example, I played a Dm11 – G13b9 to a solitary C note, then play an Ab chord (which contains the C note as its third), and it sounds nice. Voila – instant modulation down a major 3rd!
Likewise, play Dm11 – G7/6(with G on top) – Ebmaj add9...very nice.

Here is another modulation trick that Ted really liked, that uses a similar idea to move down a 1/2 step:

I - I/5 - vi6 - #IV7(b5b9)/3
                   = V7 of VII
                              = I of new key. 

I realize that this looks ghastly, so in real terms it looks like:

C - C/G - Am6 – C/A#    -   B (new key)
                           F#7(b5b9)

For the curious, the reason that we have a C triad over an A# instead of a Bb is that the chord is actually an F#7(b5b9), which sounds and looks identical to a C/Bb.  I remember Ted showing me how he got there by playing the progression, and from the Am6 he first went to an F#7/3, voiced A#, F#, C#, E (bottom to top), then he moved the F# up to a G to create an F#7b9/3.  He then dropped the C# down to C in order to arrive at the target in question.  It really does sound neat.

Here’s a lesson that Ted wrote out for me on a blank sheet of paper:

Modulation Studies:
Practice voice-leading on vi6 to I, iv6 to I

       vi6                           iv6
C   Cm6     Eb            C  Cm6     G
C   C#m6   E              C  C#m6   Ab
C   Em6     G              C  Abm6   Eb
C   Fm6     Ab            C  Am6     E
C   G#m6   B              C  Bbm6   F

– Mark Thornbury