So I have been playing around with the good'ol ii V I standard and I have found, just by playing around, that instead of going to the V, it sounds odd, but you can substitute it for a bV.
Key: D
Normal:
iim - Vmaj - Imaj | Em - Amaj - Dmaj
Substituted:
iim9 - bV7 - Imaj7 | Em9 - Ab7 - Dmaj7
Can we even use this substitution or am I just hearing things? :^)
If this does work,
Why can we substitute the V for bV?
Does this substitution work for every perfect cadence?
This is actually something that I stumbled upon a few years back and explored quite a bit, even extending it to a VI-II-V-I with bVI(minor or dominant)-II-bV-I. I think the sound is very interesting and clearly not common, so it has some unique characteristics compared to most of what we hear. There are, however, a few issues with treating this as a true substitution in the same sense as a tritone substitution. I would define a true substitution as an alternate chord to be used in place of another that would serve the same function; this substitution would also be able to be used in most songs/pieces without disrupting the function or melody.
The reason a tritone substitution works so well is that it shares the same guide tones as the chord it is substituting (G7 guide tones: B, F; Db guide tones F, Cb (enharmonically equivalent to B). Because of this, the voice leading works the same for the ever important tritone of 7 and 3. This allows the substitution to have the same function without taking away from the important resolution.
When using bV, the guide tones are also a half-step lower, which changes the notes that cause the tension that will resolve to the following chord. This resolution is less strong, largely because the leading tone is no longer a part of the chord but also because of the voice leading. With traditional Jazz voicings in a II-V-I, each guide tone's line will follow a pattern of moving down by step, then held as a common chord tone, or vice versa (slightly different with the resolution to minor where both guide tones move down by step when resolving to I). For example, a II-V-I in C would have guide tones: C-B-B; F-F-E. When we substitute bV, the guide tone line starting on C would be: C-Bb-B (the other guide tone line still holds the pattern with F-E-E). This disrupts the linear motion that the guide tones usually create in a II-V-I, taking away from its smoothness. So the combination of no leading tone and the lack of linear motion in the guiding voices ends up disrupting the smooth motion that a typical II-V-I has, as well as having a less strong resolution to tonic.
Speaking to the melodic aspect, the bV brings a lot of issues in terms of a true substitution. The tritone sub doesn't have this issue because it ends up just creating an Alt7 chord, with all of the diatonic notes from the key still being chord tones/alterations; the only diatonic note that is not accounted for in the tritone sub is the tonic, which is usually not a melodic note over a V anyway. This means that any diatonic melody can use a tritone sub in place of the V. With bV, you end up having some important melodic notes not being able to be properly harmonized within a diatonic melody, specifically ^7 and ^4, the two strongest tendency tones.
That's essentially my case against bV being a true substitution but I wouldn't say that I'm against its use, in fact I have written many songs including this chord and love how it sounds. The main issue that I have faced is trying to write a nice melody that goes along with the changes. This stems from the fact that the bV is derived from the major scale a half-step below the tonic. This sort of tonal relationship only has two notes in common, ^7 and ^3 of the tonic key, and ^7 is not properly harmonized by the bV chord. You can alter the bV and get ^2, ^5 and ^6 as common tones, which helps, but I still find that the melodies I was able to come up with did not sound very nice to me, presumably because of the dissonant nature of the harmony combined with the weakened resolution.
I also found that I liked the resolution to minor better than major, which I attribute to the voice leading being more smooth. This also made the bVI-II relationship that I referenced at the beginning of this answer seem more at home, as the standard in minor is to use a bVI, though in my case I used a dominant or minor chord instead of the expected maj7.
I've also seen bV in some Jazz songs as part of a sequence (though I can't remember which): II-V-bII-bV-I. It's arguable as to the functionality of the bV being a substitution for V here and instead thinking of it just as a sequence. I also believe that Debussy's Prelude to Afternoon of a Faun uses a bV as a substitution but I would argue this is not a true substitution, as it is specific to that piece, not a concept that can be applied in all settings.
I always like to remind people that theory is not a set of rules; it is an explanation. As such, you can do whatever you want, wherever you want, as long as it sounds good. If it sounds good to others and you gain some sort of audience, your approach here could end up creating new theoretical concepts. I clearly referred to things using words like 'properly', which may sounds like it subverts what I'm just now saying, but ultimately theory becomes rules when applied to specific genres and idioms. So my use of 'properly' and such is basically just referring to how things are handled in a traditional approach to Jazz. So be creative and write these songs but you just need to be aware of how using any sort of substitution in a more traditional setting can cause some issues with functionality and harmonic consistency that may subvert the idiom you are performing/writing within.
Em9 A7 AbV7 makes some sense: A7 in this context is a tritone substitution of the Eb7 going to AbM7, which is swapped with Ab7 upon landing.
Then Ab7 is the tritone substitution of D7. Swapping D7 with DM7 is going to sound weird (definitely weirder that the other way round) and not a very satisfactory resolution but it is conceivable.
In any case this is not in my list of "usual substitutions" but I reckon it can work.
maybe if you try to see chords notes you can find some analogy. Usually V substitution is to its tritone, so in your case is EB: Eb-G-Bb-Db(-F-Ab). Now Ab: Ab-C-Eb-G-(Bb-Db). Then it can be Eb7-11 > Ab7-9-11, in this case Ab7-9-11 shares 5 notes with Eb7-11, but in different position. Conceptually, I think that any chord could be superimposed to another when adding 9-11-13, it depends on how it sounds and... mainly on who is playing it! Music is full of transgress rules, which really build its theory and history. Hope this helps.
