Developing a heuristic for evolving guitar blues licks

Towards the development of a pseudo simulated annealing genetic heuristic for evolving blues ideas

Architecture, graphic design, your projects.

Moderators: Calilasseia, Blip, ADParker

Re: Developing a heuristic for evolving guitar blues licks

#61  Postby John Platko » Dec 17, 2015 4:11 pm

And, as we did for pitch transitions, we can create network maps showing how licks transition from one count pattern to the next. For example, Lick 14 - which is:

Image

can be mapped like:

Image

And a map showing the transition probabilities for sets of licks can also be created. Here's such a map for the founder set of licks. Only basic count patterns were involved.

Image

And here's the map for the mutated licks that the count pattern was presented for a couple of comments back. Now 7 common count patterns have developed.

Image

And here's the result of a several generation run that ended with a set of about 700 licks.

Image

Now there are licks that have triplets that cross beat boundaries. Which may or may not be a good thing. Any opinions on that? In any case, it's easy to see how this kind of representation lets you see the forest from the trees and find licks that stand out as being "creative". For example, we can look at one with a triplet that crosses a beat boundary that the network map helped find.

Image

And we can look at its network map.

Image
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Ads by Google


Re: Developing a heuristic for evolving guitar blues licks

#62  Postby John Platko » Dec 17, 2015 6:53 pm

I stumbled upon this video lecture which is somewhat related to what I'm doing and I thought it might be of interest. It does present some of the concepts that I'm using and also uses Python and LilyPond. It's more 'proper" music focused, not devil at the crossroads music like I'm shooting for, but it's still kind of interesting.

I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#63  Postby John Platko » Dec 21, 2015 5:01 pm

And following along on the theme of counts and rhythm. Licks can be mutated by using the rhythm of a different lick. This can be done to the original lick, leaving no trace of it as it was or perhaps more usefully as another form of spawning two licks. i.E. using the rhythm of one lick and the pitch family of another lick. This is a bit trickier to do then it might at first seem because not all mutations are possible if adding pitches is not desired. For example. If a 1/2 step bend is called for but those notes are not available in the lick pitches then perhaps a whole step bend or a curl (1/4 bend) should be used. But if the note is on an open string then not even a curl is possible. All in all this is a bit tricky to account for.

Here's an example of the founder set mutated in place. I.E. every lick is mutated to conform to the rhythm of lick 0.

Image

And here's what that sounds like.
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#64  Postby THWOTH » Dec 21, 2015 5:35 pm

This is all good for generating practice material, but that only offers a hint to the true heuristic for working in an improvised medium: practice, with the aim of developing a personal musical language that can be drawn on in context. A improvisatory tradition like the blues cannot be naturalised into equations and quantities because it's not about the notes but about the in-the-moment choices of the musician, and how that makes them, and you, feel about the music.
"No-one is exempt from speaking nonsense – the only misfortune is to do it solemnly."
Michel de Montaigne, Essais, 1580
User avatar
THWOTH
RS Donator
 
Name: Penrose
Posts: 37081
Age: 54

Country: Untied Kingdom
United Kingdom (uk)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#65  Postby John Platko » Dec 29, 2015 3:56 pm

THWOTH wrote:This is all good for generating practice material, but that only offers a hint to the true heuristic for working in an improvised medium: practice, with the aim of developing a personal musical language that can be drawn on in context.


Very true. At this point I find what I've generated to be most useful for practice material. It helps me see variations, both in note order and places to put rests in ways that help me out of the ruts that I somehow find myself getting stuck in playing over and over again. It also helps me understand and get a better feel for subtle timing differences.

But beyond that, I find the process also helpful in bringing consciousness to what I'm going about doing when I make my attempts at playing the blues. It makes me think about what is a lick, and what is not, and if such a counterfactual actually is real. For example, I've come to the conclusion that a lick must be resolved enough, it must have a "period' on it of sorts to be a lick. And when I did some checking, all the licks I looked at did just that. There's a long list of things that an exercise like this forces me to think about in quantitative ways which give added insight to the metaphors that are more commonly used to help communicate about what is going on. For example, people often say that the blues is a mixture of sweet and sour, i.e. major and minor and that mixture in various combinations helps communicate feelings. But, what does that really mean? Is that even true? What is the difference between major and minor? Can we model that and quantify that? Which brings up other questions like: is that model just another mythology added on top of the sweet-sour metaphor or is it solidly based in physical reality and human physiology - not that I hope to resolve that question in this thread but diving deep into nuts and bolts of licks makes me more aware of questions like that. So beyond practice material- I'm developing a more detailed understanding of the nuts and bolts of licks in a way that is effective for me.

And as you say, I've only hinted at a heuristic so far - very true. This is a bottom up development approach and so far I've just been laying the ground work which I need as building blocks for a heuristic to develop a dictionary of blues licks. And I envision that dictionary itself, as just a set of building blocks, a vocabulary, for actual blues music. Now when I started the thread, creating dictionary of licks and showing the relationships of various licks to each other was the extent of my project but now that I'm getting close to being able to start working on heuristic for the dictionary I'm thinking the next step after that would be to pick licks from the dictionary to create blues solos. Which brings us to:


A improvisatory tradition like the blues cannot be naturalised into equations and quantities because it's not about the notes but about the in-the-moment choices of the musician, and how that makes them, and you, feel about the music.


I agree. I think that's what it is ultimately all about. At some point in my musical struggle I got that. But I'm hoping to shed some light on what's involved in those choices, or at least present another way of looking at those choices that may be helpful to someone like me who is a bit musically challenged and needs to use other skills to help overcome those deficiencies.

This seems like a good time to summarize what has been covered so far and to lay out a bit of road map of where I'm planning to go next.

Up to this point I've presented a very brief overview of blues music, with emphasis on blues licks. After that I've been building a software foundation that allows me to mutate or transform one set of licks into other sets of licks in subtle and less subtle ways. I've also presented some measuring techniques that can be used to help describe the qualities of different licks. i.e. how much information they contain - or their entropy. Are they more major or minor. How close are they to the note structure of a particular chord. Ways to think about their musical timing. And ways to create network graphs that allow us to look at characteristics of entire sets of licks.

At this point these low level functions have be demonstrated in isolation as they are being developed because I need them (and a few more) in place before I can use them in a meaningful way all together in some grand lick dictionary heuristic.

So before I start on the heuristic I must still cover:

1) different fret positions. So far I've been using the key of E and the part of the fret board near the nut so I need to talk a bit about what's involved in moving licks around the neck and what effect that may have.

2) I've covered cool and warm mutations but I need some hot or really out of the box mutations.

3) At this point I calculate pitch and rhythm entropy with rest notes part of the rhythm calculation but, for various reasons, I want a three dimensional entropy vector, pitch, rhythm, and harmony, where harmony describes the vertical complexity of the music. So I need to fix my entropy calculator.

4) I've mentioned I, IV, and V chord licks a bit but I need to introduce turn-around licks and deal with them as a separate class of licks.

5) I have some internal housekeeping work to do, in particular, I need an improved hashing mechanism so that some lick operations are more efficient.

6) I need to nail down the metric I want to use for major- minor and perhaps major - minor - dom classification.

And perhaps other things will come up along the way. But after that, I'll start getting into the actual heuristic for the evolution of licks. I have decided on the general direction I plan to go with this and I'm pretty excited about it because I'll be building off a concept which is new to me and I'm hoping that trying to use it will help me understand it. And perhaps others will find it interesting too. So for those who are bored with the minutia that I still have to deal with before getting into the more interesting stuff, here's a preview of where I'm going and an invitation to investigate this topic on your own.

I'll be using Constructor theory (as best I understand it) to provide the overall structure for my heuristic with simulated annealing sitting under that framework. I like the idea of using constructor theory because the general concept of using transformations along with describing what is impossible fits well with what I've been doing. The kind of mutations I've presented so far can be seen as transformations and the various metrics, entropy, major/minor, rhythm pattern, can be combined to define various universes where certain types of licks are possible or impossible. And simulated annealing will be used to distribute licks in those universes.

But perhaps it's best and far more pleasant to learn a bit about Constructor theory from Dr. Marletto.



Great comment THWOTH :cheers:
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#66  Postby John Platko » Jan 02, 2016 8:13 pm

Working down the list of basic functions I need to complete before starting the actual lick heuristic ...

1) FIngerboard positions.

Up to this point licks have been using notes at the lower part of the fingerboard near the nut- and they've been in the key of E. But there are other keys and other fingerboard positions. I'm not planning to do much with different keys because, for the most part, it's a fairly simple matter of transposing licks in a given key to another key by changing where on the fingerboard you play the lick. (There are some issues like, use of open frets, that get in the way sometimes but that's a nit that's not too important to what I'm doing.) I'll just give an example of how a lick in the key of E can be transposed to the key of A by moving all notes in the lick up 5 positions on the fretboard. LickMaker now has a function that can do this. Here's lick 0 transposed to the key of A.

Image

Similar to the previous example. Moving all notes in the lick 12 frets up the fingerboard transposes the lick an octave up. So we get the same key but now the pitches in the lick are an octave higher and, because of the nature of guitars, each note will have a bit different harmonic content too.

Here's lick 0 an octave up:

Image

Often guitar players can play the same lick in different positions on the fretboard. That is, they can as long as all the notes in the lick are available at or near that position. For notes that are not available they need to either play them in some other position or play those notes in an octave available at the new position. For example: here's lick 0 moved up to the third fret (which many systems consider the next position up for this key).

Image

And here's the same lick moved up to the 5th fret.

Image

Moving the lick to different locations on the fingerboard as in the above examples will be eventually useful to help LickMaker populate the fingerboard, or physical dimensions, of the "lick universe" that it will try to fill. But I think I'll simplify things a bit at first and save that for last.

However, transposing lick position is also useful in changing the chord/scale of the lick. And by making a couple of position transformations (of different types) , a lick that has more of a minor chord and minor pentatonic feel can be made to have more of a major chord and major pentatonic feel - as I've done in the following example: (I think I'll just leave the whammy bar bend)

Image

Doing a similar major minor position transformation is common for major minor scale and chord transposing but the double position transformation shown in the last example is new to me and I need to think about it more and explore what I can do with it.
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#67  Postby John Platko » Jan 04, 2016 4:43 pm

Next on my "to do" list is hot mutations. This is a bit difficult to define, that is how hot should hot be? How out of the box should I be willing to try. I can't answer that at this point. The plan is to set up a mechanism where a lick universe is defined and then let LickMaker try to fill that universe by transformations of the members of that universe. If it fails to adequately fill the universe then I'll have to modify the heuristic and add transformations until it can.

Also, at this point I imagine mating selection will have a big effect on how hot the mutations get. Mating chords within the same family vs with other chord families should have large effects on pitch variety.

All that said, I've add a hot mutation which mutates a note in a lick which is pretty much in one scale to a note from other scales. For example, a minor or major pentatonic lick will get a Dorian or Mixolydian pitch either added to the lick or replacing another note in the lick. For Lick 0, here are some examples of what this type of mutation creates.

Image

And then there's the questions, what about just random pitch mutations? And how random should random be? Again, I don't know - we'll see. But I added another hot mutation which either adds a random pitch (which doesn't already appear in the lick) to the lick or replaces another note in the lick. The caveat being, this "random" note has to be somewhat near and on the same string as another note already in the lick. Over time (generations of mutations) I imagine this would produce pretty random results but I haven't verified that yet. I think I'll just wait and see how my lick uninverse(s) get populated to judge how effective this hot random mutation is. But here are some examples of what such a random mutation does to lick 0:

Image

I didn't bother to post audio of this but I did listen to it and for one generation all licks produced seemed perfectly usable to me- I don't expect that would continue over future generations though.
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Ads by Google


Re: Developing a heuristic for evolving guitar blues licks

#68  Postby THWOTH » Jan 06, 2016 1:16 am

John Platko wrote:
THWOTH wrote:This is all good for generating practice material, but that only offers a hint to the true heuristic for working in an improvised medium: practice, with the aim of developing a personal musical language that can be drawn on in context.


Very true. At this point I find what I've generated to be most useful for practice material. It helps me see variations, both in note order and places to put rests in ways that help me out of the ruts that I somehow find myself getting stuck in playing over and over again. It also helps me understand and get a better feel for subtle timing differences.

But beyond that, I find the process also helpful in bringing consciousness to what I'm going about doing when I make my attempts at playing the blues. It makes me think about what is a lick, and what is not, and if such a counterfactual actually is real. For example, I've come to the conclusion that a lick must be resolved enough, it must have a "period' on it of sorts to be a lick. And when I did some checking, all the licks I looked at did just that. There's a long list of things that an exercise like this forces me to think about in quantitative ways which give added insight to the metaphors that are more commonly used to help communicate about what is going on. For example, people often say that the blues is a mixture of sweet and sour, i.e. major and minor and that mixture in various combinations helps communicate feelings. But, what does that really mean? Is that even true? What is the difference between major and minor? Can we model that and quantify that? Which brings up other questions like: is that model just another mythology added on top of the sweet-sour metaphor or is it solidly based in physical reality and human physiology - not that I hope to resolve that question in this thread but diving deep into nuts and bolts of licks makes me more aware of questions like that. So beyond practice material- I'm developing a more detailed understanding of the nuts and bolts of licks in a way that is effective for me.

And as you say, I've only hinted at a heuristic so far - very true. This is a bottom up development approach and so far I've just been laying the ground work which I need as building blocks for a heuristic to develop a dictionary of blues licks. And I envision that dictionary itself, as just a set of building blocks, a vocabulary, for actual blues music. Now when I started the thread, creating dictionary of licks and showing the relationships of various licks to each other was the extent of my project but now that I'm getting close to being able to start working on heuristic for the dictionary I'm thinking the next step after that would be to pick licks from the dictionary to create blues solos. Which brings us to:


A improvisatory tradition like the blues cannot be naturalised into equations and quantities because it's not about the notes but about the in-the-moment choices of the musician, and how that makes them, and you, feel about the music.


I agree. I think that's what it is ultimately all about. At some point in my musical struggle I got that. But I'm hoping to shed some light on what's involved in those choices, or at least present another way of looking at those choices that may be helpful to someone like me who is a bit musically challenged and needs to use other skills to help overcome those deficiencies.

I find it helpful to think of a lick as a fragment: a brick in a wall, a sentence fragment in broader narrative, a colour on a palette ready to be mixed with others and applied to the musical canvas - on it's own it means nothing, but it can form part of a larger structure and gain meaning from that context. Of course, in an improvised tradition that larger structure is built on-the-fly, in-the-moment - the lick is not like a motif in a more formal structure that has been deliberated over and tested outside of the bounds and immediate pressures of musical time.

As a lick is a base unit of musical information it can be manipulated in two significant ways: by time and by pitch. The possibilities here are virtually limitless, but as I said, the goal of working on licks in an improvised tradition is to build and integrate musical components into a language schema that can be deployed on the hoof, and so the routines that we develop to manipulate musical information also have to be integrated into our musical lexicons - not just the base units to which those manipulations have been applied beforehand.

What I'm saying is that I think this formal cataloguing project of yours, though interesting in itself, is basically a distraction which displaces you from what you really need to focus on: generating a lexicon of coherent musical information that can be drawn on in performance. That can only be achieved through practising musical performance, which is to say, by 'practising' - because if you're not practising the performance of music you're not really practising at all, you're just don't the musical equivalent of housework or weightlifting.

I used to flatshare with an alto sax player who was obsessed with Cannonball Adderly - I mean like totally obsessed. Not only did he have pretty much every recording Adderly appeared on, but he even had a copy of one of his suits made from an album cover, so that he looked like Adderly (in his own mind) on gigs. He held the instrument like Adderly, used the same mouthpiece as Adderly, the same make of reeds in the same strength, and he obsessively transcribed as many Adderly solos and lick as he could muster - which was a doomed project imo because it would take a lifetime to write all that shit down.

He felt that if he could collate all of Adderly's solos he could get at the very essence of what made Adderly such an incredible improviser, but in this he spent far more time listening to Adderly and and transcribing his solos, and writing them out neatly, and cataloguing them by recording session, by tune title, by band members, by tempo, and by key, that he didn't have much time left for practising - and I mean really practising. Not that he didn't play the alto a lot, he did, every day, and for quite a few hours - going over Adderly solos and playing along with the records and get that material 'just right'. He certainly put in the hourse with his hands on the horn, and In fact as a result he was an excellent and accomplished alto player. He could read anything you put in front of him, he knew all his chords and all the modes, and he could play really fast and really loud - which all sax players secretly admire and aspire to.

I asked him once what it was about Adderly's playing that got him so excited, and his reply went something like this: "He's just so good. I mean, in this solo I'm working on at the moment he does this thing where he pops out a low B from nowhere, and then does this quick arpeggio in fourths up to G and then, get this, he does this five-note repeated figure all the way down the wholetone scale until he's back on that B again. It's amazing!" Now this was obviously a keen observation and a pretty good musical description of a particular phrase, but it said nothing to me about how he responded to Adderly's playing beyond his capacity to combine notes in a musical phrase in a way my friend found musically satisfying. Of course, what he really responded to was the way Adderly's solos started, and developed, where they came from and where they went to and how they integrated with what the rest of the ensemble was doing; the musical journey Adderly took him on and the way that made him feel about the music, about Adderly the man, Adderly the musician, Adderly the amazing sax player, and about himself in relation to all that. Without incite into this aspect of the ultimate spontaneous musical form my friend was content (and as I put it to a mutual friend 'condemned') to reproduce material he'd collected from Cannonball Adderly, and so his own improvising, although far more technically accomplished than many others, was essentially lifeless. It lacked spontaneity. It lacked certain and necessary spirit and dynamic energy. It lacked personality, and without that it said next to nothing. It sounded like jazz but it wasn't Jazz, it was a musical jigsaw where the pieces had been put together from different boxes without reference to the lids. In effect, without the aforethought to develop his own personal musical lexicon my intense and sincerely obsessed friend mostly sounded like a mechanical pastiche of someone else.

"No-one is exempt from speaking nonsense – the only misfortune is to do it solemnly."
Michel de Montaigne, Essais, 1580
User avatar
THWOTH
RS Donator
 
Name: Penrose
Posts: 37081
Age: 54

Country: Untied Kingdom
United Kingdom (uk)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#69  Postby John Platko » Jan 06, 2016 3:27 pm

THWOTH wrote:
John Platko wrote:
THWOTH wrote:This is all good for generating practice material, but that only offers a hint to the true heuristic for working in an improvised medium: practice, with the aim of developing a personal musical language that can be drawn on in context.


Very true. At this point I find what I've generated to be most useful for practice material. It helps me see variations, both in note order and places to put rests in ways that help me out of the ruts that I somehow find myself getting stuck in playing over and over again. It also helps me understand and get a better feel for subtle timing differences.

But beyond that, I find the process also helpful in bringing consciousness to what I'm going about doing when I make my attempts at playing the blues. It makes me think about what is a lick, and what is not, and if such a counterfactual actually is real. For example, I've come to the conclusion that a lick must be resolved enough, it must have a "period' on it of sorts to be a lick. And when I did some checking, all the licks I looked at did just that. There's a long list of things that an exercise like this forces me to think about in quantitative ways which give added insight to the metaphors that are more commonly used to help communicate about what is going on. For example, people often say that the blues is a mixture of sweet and sour, i.e. major and minor and that mixture in various combinations helps communicate feelings. But, what does that really mean? Is that even true? What is the difference between major and minor? Can we model that and quantify that? Which brings up other questions like: is that model just another mythology added on top of the sweet-sour metaphor or is it solidly based in physical reality and human physiology - not that I hope to resolve that question in this thread but diving deep into nuts and bolts of licks makes me more aware of questions like that. So beyond practice material- I'm developing a more detailed understanding of the nuts and bolts of licks in a way that is effective for me.

And as you say, I've only hinted at a heuristic so far - very true. This is a bottom up development approach and so far I've just been laying the ground work which I need as building blocks for a heuristic to develop a dictionary of blues licks. And I envision that dictionary itself, as just a set of building blocks, a vocabulary, for actual blues music. Now when I started the thread, creating dictionary of licks and showing the relationships of various licks to each other was the extent of my project but now that I'm getting close to being able to start working on heuristic for the dictionary I'm thinking the next step after that would be to pick licks from the dictionary to create blues solos. Which brings us to:


A improvisatory tradition like the blues cannot be naturalised into equations and quantities because it's not about the notes but about the in-the-moment choices of the musician, and how that makes them, and you, feel about the music.


I agree. I think that's what it is ultimately all about. At some point in my musical struggle I got that. But I'm hoping to shed some light on what's involved in those choices, or at least present another way of looking at those choices that may be helpful to someone like me who is a bit musically challenged and needs to use other skills to help overcome those deficiencies.

I find it helpful to think of a lick as a fragment: a brick in a wall, a sentence fragment in broader narrative, a colour on a palette ready to be mixed with others and applied to the musical canvas - on it's own it means nothing, but it can form part of a larger structure and gain meaning from that context. Of course, in an improvised tradition that larger structure is built on-the-fly, in-the-moment - the lick is not like a motif in a more formal structure that has been deliberated over and tested outside of the bounds and immediate pressures of musical time.


My wife is taking a water color course and she recently showed me an exercise that she was doing for class creating a palette of many varieties of red. She explained how she was thinking about them, trying to organize them, ... - and I explained to her that is just the sort of thing I'm trying to do with my "lick dictionary" project and I think, for the first time, she sort of got it. Maybe I should call it a lick palette instead of a lick dictionary ...



As a lick is a base unit of musical information it can be manipulated in two significant ways: by time and by pitch. The possibilities here are virtually limitless, but as I said, the goal of working on licks in an improvised tradition is to build and integrate musical components into a language schema that can be deployed on the hoof, and so the routines that we develop to manipulate musical information also have to be integrated into our musical lexicons - not just the base units to which those manipulations have been applied beforehand.


That makes sense to me. And I agree that in its "purist" form, blues is improvised music. But sometimes, often actually, like a prepared speech, i.e. a roadmap of where a given performance is likely to go, the spots it's likely to visit have been laid down and thought about ahead of time- I guess I'm saying, there's a bit of flexibility in where one draws the line with what is live and what is memorex.

I like your phrase: "a lick is a base unit of musical information it can be manipulated in two significant ways: by time and by pitch."

I'm going to be building on something very close to that. Something like: a lick is a base unit of musical information that can be characterized and manipulated in three significant ways, its pitch or harmony information, its rhythm information, and its vertical pitch or simultaneous harmony information. And I want to drill down a bit deeper and specify exactly what I mean by that, take it out of the realm of metaphor and quantify what I'm describing, hoping that endeavor will add consciousness and deeper understanding of what that musical information, and what music itself is all about. And I also have the hope that deeper understanding of the base units will aid in how those licks are selected and/or manipulated on the fly- not in the sense of my playing along and thinking - "a change in pitch entropy to 1.84 would be good about now" but rather having internalized what such an entropy feels like would help me come up with a lick like that in real time because that kind of understanding is helpful to me and my internal processes- even though to many it may not be at all helpful. - Anyway ... that's the idea I'm going for.

And right on cue, my next comment is going to drill deep on musical information content, i.e. musical entropy.





What I'm saying is that I think this formal cataloguing project of yours, though interesting in itself, is basically a distraction which displaces you from what you really need to focus on: generating a lexicon of coherent musical information that can be drawn on in performance. That can only be achieved through practising musical performance, which is to say, by 'practising' - because if you're not practising the performance of music you're not really practising at all, you're just don't the musical equivalent of housework or weightlifting.


Ummm. indeed, this effort of mine is no substitute for practice of an actual musical instrument. In fact when I started it was very inspirational for one aspect of practice because I had to play the scores I produced to get a sense of them. I remember doing this once and my wife stepping in and saying, "that sounds nice, honey" - which was nice since my practice is sometimes met with something more like: "you're driving me crazy" because my continued repetition of some small detail can have that effect ... the point being, the minor changes, the added variety, added something to one aspect of my practice. - Still need repetition though.

I only added the capability to use midi to play the licks because I thought it might make some of the ideas I'm presenting more accessible to others who don't read music or play an instrument. But now that's it's there I'm less likely to grab a guitar and play the output and more likely to just pop in into Ableton and play it. But I think that's a temporary condition. And creating the midi emphasized another aspect, perhaps the reality of what music is, a series of on and off notes with or without bends.



I used to flatshare with an alto sax player who was obsessed with Cannonball Adderly - I mean like totally obsessed. Not only did he have pretty much every recording Adderly appeared on, but he even had a copy of one of his suits made from an album cover, so that he looked like Adderly (in his own mind) on gigs. He held the instrument like Adderly, used the same mouthpiece as Adderly, the same make of reeds in the same strength, and he obsessively transcribed as many Adderly solos and lick as he could muster - which was a doomed project imo because it would take a lifetime to write all that shit down.

He felt that if he could collate all of Adderly's solos he could get at the very essence of what made Adderly such an incredible improviser, but in this he spent far more time listening to Adderly and and transcribing his solos, and writing them out neatly, and cataloguing them by recording session, by tune title, by band members, by tempo, and by key, that he didn't have much time left for practising - and I mean really practising. Not that he didn't play the alto a lot, he did, every day, and for quite a few hours - going over Adderly solos and playing along with the records and get that material 'just right'. He certainly put in the hourse with his hands on the horn, and In fact as a result he was an excellent and accomplished alto player. He could read anything you put in front of him, he knew all his chords and all the modes, and he could play really fast and really loud - which all sax players secretly admire and aspire to.

I asked him once what it was about Adderly's playing that got him so excited, and his reply went something like this: "He's just so good. I mean, in this solo I'm working on at the moment he does this thing where he pops out a low B from nowhere, and then does this quick arpeggio in fourths up to G and then, get this, he does this five-note repeated figure all the way down the wholetone scale until he's back on that B again. It's amazing!" Now this was obviously a keen observation and a pretty good musical description of a particular phrase, but it said nothing to me about how he responded to Adderly's playing beyond his capacity to combine notes in a musical phrase in a way my friend found musically satisfying. Of course, what he really responded to was the way Adderly's solos started, and developed, where they came from and where they went to and how they integrated with what the rest of the ensemble was doing; the musical journey Adderly took him on and the way that made him feel about the music, about Adderly the man, Adderly the musician, Adderly the amazing sax player, and about himself in relation to all that. Without incite into this aspect of the ultimate spontaneous musical form my friend was content (and as I put it to a mutual friend 'condemned') to reproduce material he'd collected from Cannonball Adderly, and so his own improvising, although far more technically accomplished than many others, was essentially lifeless. It lacked spontaneity. It lacked certain and necessary spirit and dynamic energy. It lacked personality, and without that it said next to nothing. It sounded like jazz but it wasn't Jazz, it was a musical jigsaw where the pieces had been put together from different boxes without reference to the lids. In effect, without the aforethought to develop his own personal musical lexicon my intense and sincerely obsessed friend mostly sounded like a mechanical pastiche of someone else.



Thanks for the good advice, gentle warnings of potential pitfalls, and the great musical selection - I'll try to keep them all in mind! :cheers:

Ohhhhh and I should add. The time I'm spending on this effort instead of practicing is small potatoes compared to the time I spent on the subject in the neighboring thread, building a steel string guitar- now actually building guitars is a time suck.
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#70  Postby John Platko » Jan 06, 2016 10:44 pm

As it's next on my to do list, and ties in nicely with THWOTH's statement "As a lick is a base unit of musical information it can be manipulated in two significant ways: by time and by pitch. " it is time to drill into what exactly is musical information.

I've mentioned the concept of entropy as a measure of the amount of musical information in a lick from time to time in the thread and have given brief explanations of what I mean, and some examples, but now it's time to get more specific.

Some of this is review and perhaps very familiar material to some but it may be helpful to start at the beginning.

Claude Shannon introduced the concept that the amount of information being communicated has something to do with the probability of the underlying symbols being sent. (There are some issues with the language, the symbols used to communicate, i.e. "Shaka, When the Walls Fell" has a certain measure of information content when English letters are used to communicate it, but that doesn't measure the actual amount of information that " "Shaka, When the Walls Fell" can actually communicate. Soooo. we're dealing with the symbols being sent not the underlying information content of those symbols.

An overview of information theory can be found here. but this comment should explain all you need to know for this thread.

Shannon, building on work previously done in statistical mechanics, realized that the amount of information being communicated can be calculated to be:

Image

So I don't mess up the meaning of this I'll just quote from the Wiki article:

The entropy, H, of a discrete random variable X intuitively is a measure of the amount of uncertainty associated with the value of X when only its distribution is known. So, for example, if the distribution associated with a random variable was a constant distribution, (i.e. equal to some known value with probability 1), then entropy is minimal, and equal to 0. Furthermore, in the case of a distribution restricted to take on a finite number of values, entropy is maximized with a uniform distribution over the values that the distribution takes on.


A simple example should clear up any confusion. We can use Shannon's equation to calculate how many bist we need to code n distinct symbols. This is most familiar with there is an equal probability of each symbol occurring. So, if we want to represents four symbols or states, each with equal probability.

The Entropy (H or E which every you like) = - the sum of (the probability of each symbol) * Log(probability of each symbol)

in our example the probability of each symbol is .25
we want our answer in bits so we use log base 2
log2 .25 = -2
so the sum value for symbol 1) = probability of symbol 1 * Log(probability of symbol1) = .25 * -2 = -.5

and since each of the symbols are the same, the sum is just 4 times that of the first symbol. -.5 *4 = -2
and taking the - in front of the sum into account we get 2 - that's 2 bits are required to communicate four symbols with equal probability.

Of course, things get a bit more tricky to wrap your head around when the probability for the symbols occurring differ - but that's what computers are for.

I'm applying this concept to measure the symbol content of a lick. There are multiple ways this could be done- it depends on how you define the symbols used to communicate music. One could boil down all the information content of a lick to one number, it's information entropy (more or less), but music has several fairly obvious different information properties and it's useful to measure the information content of each of those properties separately. Doing this, I end up with an entropy vector for each lick. The Entropy vector has a pitch component, a rhythm component, and a vertical (simultaneous harmony- multiple pitches played at the same time) component. If you want one number to represent the entire vector you can use the magnitude of the vector.

In short, the more symbol variety of a lick in any of those components the greater that component's entropy. Some examples without the math should clarify this nicely. I'll use my entropy test licks because they are easy to understand.

Image

The figure above shows four licks, 0-3. An Entropy vector with 3 components is printed above each lick. The components, in order are, pitch, rhythm, vertical harmony.

Lick 0 is just a bar of rest. One symbol, with a probability of 1 is all that is needed to describe it. The log of 1 =0 so it's entropy is 0. In this case there are no pitch and harmony symbols - again 0 entropy.

Lick 1 is musically the same as Lick 0 so we expect and get the same answer.

Lick 2 is effectively composed of two symbols (the three rests could have been written as a dotted half note rest- which is one symbol) and the quarter note. So, we have two symbols with equal probability that's 1 bit of information.

Lick 3 is once again 2 symbols with equal probability.

There is more than one way to factor rests into the entropy calculation. I'm ignoring rests in pitch and vertical entropy and only considering it in rhythm entropy. But I'm open to other ideas on how this should be done.

Some more licks:

Image

Lick 4 is the same note with the same duration always being played. The probability of what symbol is next is 1 so the licks entropy is 0.

Lick 5 has a lot of variety in note duration, this gives it high rhythm entropy but nothing much is happening pitch wise or harmonically.

Lick 6 shows some pitch variety and it's pitch entropy show that, Lick 7 has even more pitch variety and a greater pitch entropy.

Next we get to some examples that show vertical or what I sometimes call simultaneous harmony entropy.

Image

To measure vertical entropy I consider different simultaneous note intervals as different symbols. That is, if the chord has a root, Maj3rd, P5 interval content then it gets the same symbol no matter the root. If different chords are played that shows up in the pitch entropy.

Lick 8 has the highest pitch entropy because every chord has different intervals. Like 9 has a perfectly predictable vertical pitch interval so its vertical entropy is 0. Lick 10 has effectively 1 bit of vertical entropy. Lick 11 has lower vertical entropy.

Now all of this is way more tedious to explain than understand and relate to. It can all be said simply as: the greater the pitch entropy the greater the pitch varies, the greater the rhythm entropy the greater the rhythm varies, the greater the vertical entropy the greater the vertical intervals vary.

There are some approximations made here and there in my calculations but they're mostly nits and I won't go into that. And I'm still checking this so there might be a bug here or there- but I think you get the idea.

Now that we have a measure of various information components of licks we can plot a set of licks and see how their information content compares. Below is a plot of the test licks I described above. (The vertical axis is vertical (harmony) entropy)

Image

And we can plot the founder set of licks that I've been using throughout the thread. I won't post the scores for those licks as I suspect this is already way to tedious but if someone wants it - just ask.

There wasn't a lot happening vertically in those licks.

Image

And after a couple of generation of mutations of all sorts except no spawning . (resulting in about 700 licks)

Image

After a couple generations with spawning - still about 700 licks.

Image

And that's my multidimensional entropy calculation. The plan is for it to play a center role as I develop my simulated annealing heuristic based on constructor theory. Briefly, the idea is that one way to guide, that is, one knob of the heuristic, is the definition of a volume in entropy space and a density function. The heuristic will attempt to fill the space with the specified density function, licks with entropies that fall outside the defined volume will not be able to exist.

What else ... There is other information in music that this doesn't cover. ie. dynamics, i.e volume, ways to hit a note, picked, vs hammer ons, but most of what would appear on the score is being taken into consideration. By looking at where a lick falls in my entropy space you should be able to have some idea of what the lick might feel like. Of course knowing the amount of information says nothing about the actual information- so more metrics are needed.

Any questions.
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#71  Postby THWOTH » Jan 08, 2016 11:59 pm

I've read your last post but I just want to address this point from the one before.

John Platko wrote:I like your phrase: "a lick is a base unit of musical information it can be manipulated in two significant ways: by time and by pitch."

I'm going to be building on something very close to that. Something like: a lick is a base unit of musical information that can be characterized and manipulated in three significant ways, its pitch or harmony information, its rhythm information, and its vertical pitch or simultaneous harmony information.

Pitch and harmony are two different constructs, pitch being a description of a an singular or individual tone and harmony being a description of combined pitches. The notion of an additional characteristic that you call 'vertical pitch or simultaneous harmony' is redundant here because although pitch and harmony are two different constructs and each lends musical context to the other harmony is pitch-dependent.

'Licks' for musicians of single tone instruments (brass, woodwind etc) contain nothing but pitch and rhythm information. Licks for players of instruments that can play multiple notes (guitar, piano, violin, etc) may also contian 'harmony' in the sense of employing simultaneous note combinations, but any apparent harmonic variation/manipulation occurs through pitch manipulation.

Look at the first phrase in your post here. On paper it appears to have a particular harmonic context, lent to it by the key signature and the fact that it appears to resolve to the tonic of that key, but really it's just seven pitches of a (relatively) defined duration. Play that over a Cmajor7 chord and it could work just fine (as it could over an Am7, C#m7b5, G6/9, B7b9, F+#4, G#m-major7b5 or any number of other chords), but in the end the phrase can only be manipulated by pitch (displacement or transposition) and/or by time (displacement or translocation).

Pitch and time are the base components of music, and the context they lend to each other is generally what distinguishes 'music' from 'sound' or 'noise'. All other musical constructs can be fully described in terms of pitch and time.

Of course, whenever I say 'pitch' here what I'm really talking about is timbre, but that's a whole different thing and probably best left for another day.

Please don't think I'm trying to discourage you in any way, in fact quite the opposite. I'm just wanting to emphasise that you don't need to bring in or generate extraneous information to develop a skill into an art - all you need to do is play. The more hours you spend playing the better you'll get at expressing yourself, because playing is ultimately what develops, expands, deepens, populates (or whatever) your musical capacities and capabilities - what Jazz musicians call 'chops'.


:smoke:
"No-one is exempt from speaking nonsense – the only misfortune is to do it solemnly."
Michel de Montaigne, Essais, 1580
User avatar
THWOTH
RS Donator
 
Name: Penrose
Posts: 37081
Age: 54

Country: Untied Kingdom
United Kingdom (uk)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#72  Postby John Platko » Jan 09, 2016 3:06 pm

THWOTH wrote:I've read your last post but I just want to address this point from the one before.

John Platko wrote:I like your phrase: "a lick is a base unit of musical information it can be manipulated in two significant ways: by time and by pitch."

I'm going to be building on something very close to that. Something like: a lick is a base unit of musical information that can be characterized and manipulated in three significant ways, its pitch or harmony information, its rhythm information, and its vertical pitch or simultaneous harmony information.

Pitch and harmony are two different constructs, pitch being a description of a an singular or individual tone and harmony being a description of combined pitches. The notion of an additional characteristic that you call 'vertical pitch or simultaneous harmony' is redundant here because although pitch and harmony are two different constructs and each lends musical context to the other harmony is pitch-dependent.


The components of my three dimensional entropy vector, pitch, rhythm, and vertical harmony are not redundant. They are representing how pitch differs from harmony as you described it. Perhaps I over complicated the language. I sometimes hesitate to simply use the words pitch and harmony because harmony invokes the idea of chords and that can invoke the idea of simultaneous pitches, but chords can be arpeggiated or "spread out" too. In the context of a blues lick, it is sometimes useful to think of the chord or blend of chords, or even approximate chord that the pitches in the lick contain.

So my use of "vertical harmony" is just to make clear that I mean simultaneous pitches (as opposed to arpeggiated pitches). As long as that's understood, then I'm fine with naming the components of my entropy vector: pitch, rhythm, and harmony. And harmony is, as you suggest it should be, measured in a pitch independent way, that is, the intervals between the notes are what is determining their symbol value, not the exact pitches of the notes.



'Licks' for musicians of single tone instruments (brass, woodwind etc) contain nothing but pitch and rhythm information. Licks for players of instruments that can play multiple notes (guitar, piano, violin, etc) may also contian 'harmony' in the sense of employing simultaneous note combinations,


Yes, that's what my entropy measurement should show.

but any apparent harmonic variation/manipulation occurs through pitch manipulation.


:scratch: I'm thinking you mean stings bends and the like. I've classified such manipulations as pitch changes and put them in the pitch entropy calculation. :scratch: I'll chew on that some more to see if I can understand why they should be in the harmony calculation but I'm not seeing that at the moment. Is there a reason why they have more to do with vertical harmony?



Look at the first phrase in your post here. On paper it appears to have a particular harmonic context, lent to it by the key signature and the fact that it appears to resolve to the tonic of that key, but really it's just seven pitches of a (relatively) defined duration. Play that over a Cmajor7 chord and it could work just fine (as it could over an Am7, C#m7b5, G6/9, B7b9, F+#4, G#m-major7b5 or any number of other chords), but in the end the phrase can only be manipulated by pitch (displacement or transposition) and/or by time (displacement or translocation).


I could also add one or more double stops to it and that would manipulate what I've been calling the vertical harmony. For example, after the bend, I could play the open B with the open E. That would give a bump to my harmony entropy. The vertical entropy dictionary would show two symbol types: single pitch and a P4 interval from the vertical lowest note (the exact pitches would not be registered.)


Pitch and time are the base components of music, and the context they lend to each other is generally what distinguishes 'music' from 'sound' or 'noise'. All other musical constructs can be fully described in terms of pitch and time.

Of course, whenever I say 'pitch' here what I'm really talking about is timbre, but that's a whole different thing and probably best left for another day.


I'm been talking a fair amount about harmonics in my thread on building a guitar, I'm thinking about adding a demonstration on how harmonics add to create the sound we associate with a steel string guitar.



Please don't think I'm trying to discourage you in any way, in fact quite the opposite.


I don't think that at all. I appreciate your comments and welcome sanity checks on my work. This is especially true since, for various reasons, I often prefer to come up with my own ways to think about music and that approach can be error prone and problematic. Know your comments are well welcomed and appreciated.



I'm just wanting to emphasise that you don't need to bring in or generate extraneous information to develop a skill into an art - all you need to do is play. The more hours you spend playing the better you'll get at expressing yourself, because playing is ultimately what develops, expands, deepens, populates (or whatever) your musical capacities and capabilities - what Jazz musicians call 'chops'.


I see how that makes sense for most people if the goal is to be a better musician. However, I have found that for me, improvement in playing skills ofetn comes faster with added consciousness of what I'm doing. And thinking about this, it's not just the direct application of knowledge that is helpful but the added confidence that I somehow get by that added knowledge - what I'm describing isn't completely rational but it is what it is.

But also, I'm interested in the concepts being explored by this thread. It's one thing to talk about entropy or read it in a text book or using it in a simple binary calculation and a whole different experience of the concept applying it to something like this. Likewise, there are many other technical concepts that I'm exploring in this thread, the blues just happens to be an enjoyable way for me to explore those concepts. And soon, I'll be starting on the actual heuristic and toying with constructor theory, in my own way, which will help me gain insight into what it's about. And if nothing else, it feels like a constructive way to contribute to the forum. :cheers:
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#73  Postby John Platko » Jan 09, 2016 11:12 pm

Fourth on my to do list that I previously posted was to mention turnarounds. A turnaround is used at the end of a 12 or x bar progression to get you back to the beginning of the loop. Like most things in the blues, there's more than one way to do this but a common method, perhaps the most common, is to end the twelve bars on a note that begs for resolution to the root of the key. And fifths do this nicely. For the key of E that would be a B chord or B note. Often turnarounds have familiar structure and sound - everybody knows where the music is at when they hear the turnaround. Because of this, I plan to treat turnarounds like they are a different species from other licks. Turnarounds won't spawn with non turnaround licks, They'll mutate but they should keep enough of their founder characteristic that they won't loose their feel of being a turnaround.

I could say a lot more about turnarounds but I think that can wait for later, I just wanted to make sure I mentioned them and had basic mechanisms in place to handle them separately from other licks. And those mechanisms can be used to segregate other licks too, as desired.

Here is a small founder set of turnaround licks:

Image

And you can hear them here.
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#74  Postby THWOTH » Jan 10, 2016 3:19 am

OK, let's try it this way.


  1. What is musical authenticity?
  2. What constitutes an Authentic Blues?
"No-one is exempt from speaking nonsense – the only misfortune is to do it solemnly."
Michel de Montaigne, Essais, 1580
User avatar
THWOTH
RS Donator
 
Name: Penrose
Posts: 37081
Age: 54

Country: Untied Kingdom
United Kingdom (uk)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#75  Postby John Platko » Jan 10, 2016 4:09 pm

THWOTH wrote:OK, let's try it this way.


  1. What is musical authenticity?
  2. What constitutes an Authentic Blues?


Those questions somehow remind me of a video I posted earlier in the thread which I find a good starting point for answering such questions. chruck Berry schooling Keith Richards

I would say in that case, what Chuck is looking for is "Authenticity of performance" defined here to be:

With performance arts such as music and theater, both the composer or playwright and the performers are involved in creating an instance of the work. There are some who consider that a performance is only truly authentic if it approximates as closely as possible what the original author would have expected to see and hear. In a historically informed performance, the actors or musicians will make every effort to achieve this effect by using replicas of historical instruments, studying historical guides to acting and so on.


:scratch: although Berry seems cool with Richards using a Strat. :scratch:

And as that article points out, Berry's quest for authenticity of performance interferes with Richards' "authenticity of expression".

Dutton's concept of expressive authenticity is based on the Oxford English Dictionary alternative definition of "possessing original or inherent authority". In this sense, authenticity is a measure of the degree to which the artist's work is a committed, personal expression rather than derived from other work. It includes concepts of originality, honesty and integrity. In the case of a musical performance, authenticity of expression may conflict with authenticity of performance. The player is true to their personal musical sense and does not imitate someone else's method of playing. Their performance may thus differ significantly from that of a player attempting to follow the style common at the time the musical work was composed.


:scratch: So which is the authentic authentic?

And reading further :coffee:

Expressive authenticity is related to the technical term authenticity as used in existential philosophy. It has always been thought right to know oneself and to act accordingly, and in existential psychology this form of authenticity is seen as central to mental health.[33] Prominent artists such as the Abstract Expressionists Jackson Pollock, Arshile Gorky, and Willem de Kooning have been understood in existentialist terms, as have filmmakers such as Jean-Luc Godard and Ingmar Bergman.[34] The greater popularity of performer-based music as opposed to composition-based music is relatively recent. It seems to reflect a growing interest in expressive authenticity, and thus in musicians who have a unique and charismatic style.[35]

The question of whether an artistic work is an authentic expression depends on the artist's background, beliefs and ideals. Andrew Potter cites the example of Avril Lavigne, a teenage singer from Napanee, Ontario who burst onto the pop music scene in 2002. She claimed to be a small-town skateboarder, with her background providing the subjects of her songs, and said these songs were her own compositions. These claims of authenticity of expression and of provenance were both challenged. However, her work could have been authentic in expression even if Lavigne had not written it, or authentic in provenance if she had written it but not authentic in expression if the carefully cultivated skater-girl image were false.[36]

Authenticity of expression may thus be linked with authenticity of style or tradition. Many feel it is not permissible for someone to speak in the voice of another culture or racial background, and that such an expression cannot be authentic.[36] For example, hip hop was originally an art form through which underprivileged minorities in the United States protested against their condition. As it has become less of an underground culture, there is debate over whether the spirit of hip hop can survive in a marketable integrated version.[37] In "Authenticity Within Hip Hop and Other Cultures Threatened with Assimilation," Kembrew McLeod argues that hip hop culture is threatened with assimilation by a larger, mainstream culture, and that authenticity of expression in this genre is being lost.



Which sounds to me like the usual nonsense that is dished out when the gate keepers arrive on scene. I assume you mean something very different than all that.

But getting back to Chuck and Keith, while it's understandable why Chuck would want his music to be authentic to his vision of it, would Keith's version be any less authentic to most people? If it's authentic enough for Keith, wouldn't that be authentic enough for most? And what about this conflict of authenticity, the sacrifice of Keith's personal artistic expression for Chuck's vision of what should be? Maybe the whole thing would have really been more authentic if Chuck just let Keith play it his way, after all, it's never going to be like the original version if Keith is in the picture.

Relating these ideas of authenticity to what I'm doing in this thread, I consciously worked out a paradigm that embodied how blues ideas evolved over time. The notion of starting with a "founder" set of licks and then mutating and adding to them, creating something new after learning a bit about what's be done before - that's authentic blues tradition and culture. And coming up with ways to describe those licks and how they play off each other, using metaphors like sweet vs sour, etc. that's authentic blues tradition and culture. I'm just changing the old metaphors to new metaphors that use numbers. And while it may be more authentic blues culture to not get too technical about the music ( although there were/are some very technical blues players), it's standard operating procedure for Jazz musicians to be very technical about the music. So, I would say that so far, I'm finding LickMaker to be authentically bluesy. And by and large, the licks that it has produced so far are viable variations of traditional blues licks and sometimes something that seems a bit new (at least to me) and still bluesy.

I have no truck with:

White men Computers can't play the blues.

What do you think is required for authentic music and authentic blues?
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Ads by Google


Re: Developing a heuristic for evolving guitar blues licks

#76  Postby John Platko » Jan 11, 2016 7:21 pm

Next on my "to do" list was some internal hashing that I needed to add to make checking for duplicate licks more efficient. Not much to directly show about that but now that it's in place it's easy to generate large sets of mutated licks. Here's a 3d entropy plot of about 10,000 licks.

Image

It can be hard to figure out what's going on in a list like that. If I printed the score for that with about 30 licks per page it would result in over 300 pages of licks to sort through. Here's were network graphs can help a bit. This one shows the probability of note transitions and beginning and ending notes of licks. It's interesting to see how the Root, P5 and P4 show up front and center in the graph.

Image

And you can see what count patterns appear in all the bars and the probability of transitioning from one count pattern to another.

Image
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#77  Postby John Platko » Jan 12, 2016 10:36 pm

Before tackling the last item on my "to do" list for this phase of the project, I decided to add pre-bends, which were missing up to now. This turned out to be harder than I imagined because LillyPond doesn't have a built in mechanism specifically for pre-bends. I did find an add on package from someone that encountered this problem before, and he can create all kinds of nice bends, but I didn't have much luck with his package - but for now, I found a hack way of creating pre-bends that are fine for my purpose. Here's some examples that were mutated into existence:

Image

And you can hear them here. I didn't put a lot of work into making them sound great but you'll get the idea.
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#78  Postby John Platko » Jan 13, 2016 6:32 pm

And now I arrive at the last issue I need to resolve before moving on to the next phase of my project. I'm rather happy with how the entropy vector worked out in measuring the amount of information in a lick but now I need metrics that say something about what that information is, that is, not just how big the pile is, and not the details about what exactly is in the pile- I know that, but more like, what does the pile taste like. Dealing with the rhythm taste doesn't seem too hard to me so I'll deal with that in the next phase. But I want to tackle the harmony - both vertical and horizontal, taste now.

Blues is often said to be a mixture of sweet and sour- but what does that mean? I think it means that the blues mixes major and minor tonality, or perhaps more accurately," harmony" and "dissonance". But what exactly do these terms mean? As far as I can tell, there is no general agreement as to what they mean. There is no general agreement as to why major chords sound happy and minor chords sound sad, or even if such characterizations are true. However, people have demonstrated that brains "light up" differently to major and minor chords- with minor chords "lighting up" more of the brain. (e.g.) The exact physiological reasons for this are unknown. I'm going to proceed with my- mmmmm, well let's just call it, my fantasy, that:

1) Minor chords are more dissonant than major chords because the components that make up minor chords don't fit together as well as the components that make up major chords.

2) When the components of a chord fit together well enough, there is not much puzzle to be solved, we can sit back and enjoy the sound and not fret over "what is wrong with this picture."

3) When the components of a chord don't fit together well enough, there is a mystery afoot, a puzzle to be solved, and our limbic system arouses us to experience, in varying degrees, sour instead of sweet.

4) All of this is fundamentally due to the actual components present in the air, and as such, can be measured and quantified.

5) However, one persons sour milk is another persons gravy for mashed potatoes - so individual judgments on these matters vary, i.e. we all don't have exactly the same tastes.

That said, I'm hoping to come up with a "sweet - sour" metric based on how compatible the harmonic content of each note is with the other notes in the chord. Perhaps someone(s) has already done this but I haven't found anything to my liking. (note: this is different than the chord pattern matching that I presented earlier in the thread where a lick could be compared to known chords to see which one(s) it was most like and a weighting given to how much like that chord it is. This is attempting to come up with an absolute sweet-sour scale for groups of notes played together.

I'm going to start with work I did in a thread in the music section of the forum where someone asked a question about why minor chords sound sad and major chords sound 'bright'. That question made me curious.

Here I posted a chart I created that tries to organize all the relevant data for this in one place. The chart shows the harmonic series associated with each note being played (a given note is composed of many frequencies, each an integer multiple of the fundamental frequency. I plan to show exactly how these harmonics shape the sound of an acoustic guitar shortly in my guitar build thread.) So, if you look at the A note in any of the three chords shown, you see, from left to right a line with marks where a harmonic may be for that note. For example, after the fundamental for the A note, at 2 * the frequency (an octave up) the harmonic is at also at the pitch of a musical A. The next harmonic up is 3 * the fundamental - which happens to be at the frequency of an E note - a P5 up from A. 4 * A is the next octave. 5 * A gives a frequency that happens to be at the frequency of C# - a M3 up from A. 6 * A is another P5, 7 * A is about a m7, 8 * A is another octave up.

And as you keep adding harmonics more of the entire musical scale (more or less) is added to the note. And all this would probably sound terrible if it wasn't for the fact that the higher harmonics have a much lower presence than the lower ones in real musical instruments. But they are there and important to the sound - which I'll demonstrate in the other thread.

Ok, so 1) the chart shows how the harmonics of a note line up with other musical notes.

The chart also vertically shows the notes of three chords grouped together. From bottom to top, A min7, A Maj7, and A 7. And the harmonics for all of the notes in every chord are also marked. ( I use some special symbols so you can easily see if any harmonic falls on a frequency that is root, P5, or M3 of the root note of the chord.

Finally, the chart draws lines between harmonics that are close in frequency to each other. Different colors are used for m2, M2, or m3 distances. My hypothesis is that the more of these close frequency components, especially in lower harmonics, the more dissonance, or the more sour the chord, and the more P5s and M3s or the root in the lower harmonics the more sweet the chord. There are other ways to say this.

This is the path I'm heading down for now to develop my sweet-sour music scale. There is another approach just using the notes that make up the chord but I like this method that uses the underlying harmonic structure in a direct way.

Any questions or comments?
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#79  Postby John Platko » Jan 18, 2016 4:23 pm

The last major item on my to do list for the first phase of my project was to come up with a sweet vs sour measurement for the set of pitches in a lick. It seems that people have been trying to come up with such a measurement and understanding of chords for a very long time, and while there are many putative theories on how this all works there doesn't seem to be one that puts the issue to bed. In short, the theories vary from: we respond to chords the way our cultural conditions us to respond to, we respond to chords because of how basic audio neural processing responds to the superposition of the harmonics in the notes. I, like many, imagine it's some combination of both; and that is what makes it difficult to come up with a definite theory. I'm still wading through some of the research on this but for now I have come up with my own measurement technique that attempts to take both cultural norms and the harmonic relationships in pitch sets into consideration. I could describe this in more detail but I'll just present the results and add detail if need be.

Lickmaker creates a Sweet measure and a Sour measure for each pitch set that it evaluates and then ranks each set on a scale from 1 to 10. A 10 on the sweet scale is the most sweet, likewise, a 10 on the sour scale is the most sour. The plot below shows where standard chords fall on the scale.

Image

The sweet and sour components form a vector, I take the magnitude of the vector to come up with the sweetness magnitude. Here's a list of the sweetness magnitudes for standard chords.

Pow 10.0
Maj_i2 9.3
Maj 9.3
Maj_i1 9.2
Sus4 8.82
Sus4_i1 8.81
Maj6 8.29
Maj7 8.08
Add4 7.65
Maj13 7.14
7 7.07
Majb5 6.93
Maj11 6.62
7_#5 6.57
Maj6_Add9 6.53
11n9 6.49
min_i2 6.46
min_i1 6.43
Maj9 6.36
min 6.23
Sus2 6.23
13 6.17
7_b5 5.69
9_i1 5.55
7_#9 5.43
9 5.43
dim 5.01
7_b9 4.99
min6 4.77
dim7 4.62
11 4.43
min7 4.33
min7_b5 3.95
min11 3.91
min13 3.86
min9 3.46


Power chords - root and perfect fifth rank a 10. (the scale can go higher, e.g. a root and an octave would be an 11- which is important because ...)

Looking at the chart we can see that the chords are generally divided along the lines that people generally think about chords, Major, Minor and dominant chords. Within these groups one my quibble over specific values and orders - but then again tastes vary.

I can now apply this metric to the founder set of licks I've been working with. The chart below shows where they fall.

Image

This chart can't be directly compared to the one above for all chords because the Platko sweet/scour scale is adaptive. That is, if it "tastes" something that seems more sour it recalibrates what it "thinks" a sourness of 10 is. Sooooo. after tasting the blues licks in the founder set, it ranks the standard chords as being less our. as the figure below shows.

Image

Now these last two figures can be compared to each other. And we can learn some things. Lick 3 seems to be about as sweet as a Power chord- and that's correct. Lick 10,11 which have minor third to Major 3rd hammer ons fall between the new major and minor chord groups. Lick 0 a standard blues run is measured as being more sour than standard chords. I imagine I'll tweak this over time but, all in all, I think it's not bad for now.

I should just note. This is looking at the notes as if they are being arpeggiated. If you strummed a 7 b9 chord it would sound more sour than Lick 0. The same sort of sweet sour measurement could be used to measure the sweet/sourness in a given beat or set unit of time but I'm not sure how useful that is - perhaps it would be a good way to weed out doublets created by adding a random note to check if the result is just too sour.

Any questions?
I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

Re: Developing a heuristic for evolving guitar blues licks

#80  Postby John Platko » Jan 19, 2016 6:03 pm

Ok, I think I laid enough of a foundation to make it worthwhile to move on to actually starting to create the heuristic. To be sure, there are some areas that will need to be backfilled. I still need a simple metric for measuring rhythm and I think I also need a metric for measuring how the tension of the lick rises and falls with time. I'll develop those as they are needed. (There's also a few details missing from the lick language, specifically, I'm not happy with how various bends of notes that are parts of triplets are handled, I'll fix that at some point too.)

Moving on ...

I'm going to use constructor theory as the foundation of the simulated annealing heuristic. So an introduction to constructor theory in some detail seems appropriate now. I would think the following videos would be of interest to anyone interested in physics even if they have no interest in blues, or my heuristic.

This video is a good overall introduction to constructor theory:



And in this video David Deutsch drills down into what he's going on about. David starts talking about 11:00 minutes in. It's well worth watching.

Here's a paper where David explains his theory.

And if you want a bit more:



and

I like to imagine ...
User avatar
John Platko
THREAD STARTER
 
Name: John Platko
Posts: 9411
Male

Country: US
United States (us)
Print view this post

PreviousNext

Return to Technical Design and Engineering.

Who is online

Users viewing this topic: No registered users and 2 guests