Rule-based Improvisation Instruction systems?

[docbop]

Not the same thing.

Point is, someone with perfect pitch has advantages in some areas. Transcription is always going to be easier, and it is not necessary for that person to practice it, or do ear training exercises.

(Which is not to say those with pitch can 'hear everything' right away either.)

Obvious extreme example, I could do all the training in the world, and I would still lose against Michael Phelps in the 200m freestyle.

There are different levels of perfect pitch so can deal with variations in tuning and some can't. I heard about a classical violinist in Europe who's sense of pitch was so sensitive he had to quit playing in the orchestra he was in because pitch variances between players. He makes he living now doing orchestra transcriptions from recordings at home. He can't deal with live music the pitch variances get to him so much.

Some say Perfect Pitch you have to be born with people who practice and develop their pitch are said to have Absolute Pitch. An another thing I heard recently is that ability to train the ear is tied to how good a memory a person has. The better a person's memory the more likely they can develop Absolute Pitch.

Kenny Werner the pianist was born with Perfect Pitch and said it was years later after he was already starting to gig that he discovered it was a unique ability. Kenny said he just thought everyone could hear like that. For some reason I find that really interesting.

[pingu]

IT IS an existential threat if you are, for instance, living the year 2017.

You know, the Luddites were RIGHT about their future. They weren't stupid. That's what doesn't get remembered:

Luddite - Wikipedia

The problem is not machines, but a world that since the industrial revolution has viewed people as machines and unsurprisingly, finds machines to be better machines.

OK I'll take that ...
I wasn't saying the Luddites were wrong ...

Only that I'm not a Luddite , I do think progress
has overall been a good thing for mankind

But , we do need to grow up and start taking
responsibility for our creations , robots/AI
being an important one right now ...

Up till now , we've just created stuff because
we can , we need to discuss the morality of
what we create ...

A certain future may be possible
(eg an automated world) but is it desirable ?
It may be or it may not be ....

We're not really discussing it in society
were just going right ahead and creating it anyway ... !

eg. If we automate everything , we'll have to pay
everyone to do nothing ... Is that gonna work ?

I saw a great Tedtalk interview with Elon Musk
The interviewer asked 'why are you doing all this
future stuff' and Musk answered words to the effect of
"I just want to be able to think about the future and not get depressed"

Me too

[plasticpigeon]

Progress is good for mankind by definition. Not all change is progress though.

Maybe Christian is actually a very advanced AI bot. How else could anyone (or thing) become so accomplished at jazz guitar and yet spend most of their life writing messages on online forums? When do you get time to practice, or have you assimilated 100 years of music content a picosecond after you were turned on?

[christianm77]

Some say Perfect Pitch you have to be born with people who practice and develop their pitch are said to have Absolute Pitch. An another thing I heard recently is that ability to train the ear is tied to how good a memory a person has. The better a person's memory the more likely they can develop Absolute Pitch.

Do you have any evidence of this?

I have heard someone researching the area to say that in her opinion is impossible to develop perfect pitch in adulthood. But - I'd be interested to hear another perspective.

[christianm77]

OK I'll take that ...
I wasn't saying the Luddites were wrong ...

Only that I'm not a Luddite , I do think progress
has overall been a good thing for mankind

But , we do need to grow up and start taking
responsibility for our creations , robots/AI
being an important one right now ...

Up till now , we've just created stuff because
we can , we need to discuss the morality of
what we create ...

A certain future may be possible
(eg an automated world) but is it desirable ?
It may be or it may not be ....

We're not really discussing it in society
were just going right ahead and creating it anyway ... !

eg. If we automate everything , we'll have to pay
everyone to do nothing ... Is that gonna work ?

I saw a great Tedtalk interview with Elon Musk
The interviewer asked 'why are you doing all this
future stuff' and Musk answered words to the effect of
"I just want to be able to think about the future and not get depressed"

Me too

I do like this about Elon Musk. I wish more people were like him.

No. I'm not anti-progress, but neither were the luddites.

We are actually on the edge of a Golden Age, this is the frustrating thing. But its not so simple in the specifics.

[destinytot]

Not the same thing.

Point is, someone with perfect pitch has advantages in some areas. Transcription is always going to be easier, and it is not necessary for that person to practice it, or do ear training exercises.

(Which is not to say those with pitch can 'hear everything' right away either.)

Obvious extreme example, I could do all the training in the world, and I would still lose against Michael Phelps in the 200m freestyle.

Neither would you need to compete against Michael Phelps in order to validate your training.

[docbop]

Do you have any evidence of this?

I have heard someone researching the area to say that in her opinion is impossible to develop perfect pitch in adulthood. But - I'd be interested to hear another perspective.

Rick Beato says that all the time and uses his son Dylan as proof. Now others I've been in discussions with say Perfect pitch you're are born with, but a person can develop Absolute pitch which in practice is same thing.

Now my question is people born with Perfect pitch don't they need to be taught the pitch names so they know what to call the pitches they hear. So really they don't just know the notes, they just can accurately identify frequency something the good recording engineer can do, hear a pitch and tell you the frequency. I worked with recording engineers you could tell them I want a sound like <fill in the blank> and if they've heard it they know how to EQ things to simulate the sound. Guess you could say it's all about ability to remember and memorize pitch/frequency.

We were talking to a woman at the music store she was buying reeds for her son in school. She said she played in school band and marching band for fun. Somehow topic of reading music came up she siad I can't read a note, never wanted to. She said I can play anything I hear I learn where the sounds are on th sax. She said it took awhile for the school band director to realize she can't read music that she was just doubling what she heard the other saxes play. She never used the term Perfect pitch, it was just if I hear it I can play it. She implied it only seemed to work on saxophone not other instruments.

Nature has a sense of humor I guess.

[destinytot]

Do you have any evidence of this?

I have heard someone researching the area to say that in her opinion is impossible to develop perfect pitch in adulthood. But - I'd be interested to hear another perspective.

I bet lifestyle and diet come into play.

[goldenwave77]

.

Kenny Werner the pianist was born with Perfect Pitch and said it was years later after he was already starting to gig that he discovered it was a unique ability. Kenny said he just thought everyone could hear like that. For some reason I find that really interesting.

Actually I think many psychologists believe most infants have perfect pitch capability, but lose it pretty quickly. Has to do with language-acquisitions skills being most impressionable at an early age, but falling off quickly if not being used. (Some believe that with languages where inflection of identical words matters a lot....such as (I'm told) Chinese....this bestows an advantage on these native speakers in learning music.)

My gf's college son, a psych. major, said they were taught this pretty recently. He graduated a couple of yrs. ago.

Rick Beato has a lot of discussion on this, as he says one of his sons has p.p. and he thinks it was developed by listening to a lot of complex, unpredictable music.

Beato also believes it is impossible to develop p.p fully as an adult, whereas really good rel. pitch can be developed, and as far as music is concerned, plenty of great musicians and composers did not have p.p.

[docbop]

Actually I think many psychologists believe most infants have perfect pitch capability, but lose it pretty quickly. Has to do with language-acquisitions skills being most impressionable at an early age, but falling off quickly if not being used. (Some believe that with languages where inflection of identical words matters a lot....such as (I'm told) Chinese....this bestows an advantage on these native speakers in learning music.)

My gf's college son, a psych. major, said they were taught this pretty recently. He graduated a couple of yrs. ago.

Rick Beato has a lot of discussion on this, as he says one of his sons has p.p. and he thinks it was developed by listening to a lot of complex, unpredictable music.

Beato also believes it is impossible to develop p.p fully as an adult, whereas really good rel. pitch can be developed, and as far as music is concerned, plenty of great musicians and composers did not have p.p.

I've heard the same thing in that most people are born with perfect pitch, but since nothing is done with it, it goes away. Beato has said all the playing of complex music before and after birth make his son open to complex music that others would not like. Interesting all that listening and training his son's interest in music is pretty typical of a kid his age. He does the little bit of practice Beato tells him to but after that he's into video games and rock and roll. So all this listening and music training has done little to create interest in music for his son.

I would say a lot of the greats had perfect pitch or developed absolute pitch especially composers and the early Jazz greats. I hear these guys talk and absolute pitch and relative pitch are requirement to them it's all about the ears to them. They say there were musicians that didn't have great ears, but they weren't the composers or great improvisors. That's how I see it.

[zirenius]

I think if you'd mentioned possible educational applications, you might have gotten better responses. I for one, probably implied something that wasn't there, so I'll apologize here.

I actually believe, pretty wholeheartedly, in computer-aided drilling and reinforcement.

(I've always avoided the idea of a band camp: Listening to a concept, is easy, but putting it into practice just takes a lot of time, and I don't care who is teaching you....to train your brain, to get it under your fingers, and to execute at speed just takes time.)

In the end though, we all choose our own path. Didn't Blake say something like "I must invent my own system, or be a slave to someone else's" ?

Thank you for that. I am happy that this thread has sparked a vivid discussion and have some reading to do in the morning. I really appreciate all the comments and will try to catch up with the points being discussed first thing tomorrow.

[zirenius]

After a zillion posts and jibes, and ripostes, I'm not sure anyone has gotten to what I see as the essence of what's wrong with the OP's proposal, so I'll try to state it clearly.

- Yes, it is possible to come up with rules that can generate sequences of notes based on some sort of input or seed (e.g., a chord progression and melody); those rules can be found in many method books.
- It is possible to include some sort of rule-based phrasing logic that will make a sequence seem like a line rather than just an even sequence.
- It's possible to build rules for different song genres, rhythm, and feels (e.g., degrees of swing). Machines have been doing this for many years.

All of that adds up to simulating composition (again, something that has been around for quite a while), not improvisation. Improvisation isn't just spontaneous composition. It's also performance and (typically) group interaction, which strike me as encompassing huge amounts of activity that are not readily conceptualized via rules. So, I think the premise of the OP is misstated. He's trying to model spontaneous composition of lines via rules. I think that's eminently do-able (I don't have the coding or the rules chops for the task, but know enough about both to sort of see how to do it), and potentially interesting, especially if someone can figure out a novel approach or implementation. But that's a much more limited thing than improvisation.

John

You are correct, that is a disctinction that needs to be made and discussed. I am very (!) sceptical about whether the result of my work could by definition be considered to be improvising and I am happy to live with that limitation and still learn some things along the way.

The current state of the code is so very much dependent on my harmonic analysis and input that I can safely assure you that nobody here needs to be afraid of what I am doing.

Regarding the many method books you mentioned: do any come to mind that might lend themselves to the task at hand?

[christianm77]

Actually I think many psychologists believe most infants have perfect pitch capability, but lose it pretty quickly. Has to do with language-acquisitions skills being most impressionable at an early age, but falling off quickly if not being used. (Some believe that with languages where inflection of identical words matters a lot....such as (I'm told) Chinese....this bestows an advantage on these native speakers in learning music.)

My gf's college son, a psych. major, said they were taught this pretty recently. He graduated a couple of yrs. ago.

Rick Beato has a lot of discussion on this, as he says one of his sons has p.p. and he thinks it was developed by listening to a lot of complex, unpredictable music.

Beato also believes it is impossible to develop p.p fully as an adult, whereas really good rel. pitch can be developed, and as far as music is concerned, plenty of great musicians and composers did not have p.p.

I'm not sure how much of an authority on this stuff Rick is, but what he says broadly agrees with what I have heard generally.

That said - Bruce Arnold sells these perfect pitch courses (for instance) and I was wondering if there's anyone who's given them a go?

[John A.]

You are correct, that is a disctinction that needs to be made and discussed. I am very (!) sceptical about whether the result of my work could by definition be considered to be improvising and I am happy to live with that limitation and still learn some things along the way.

The current state of the code is so very much dependent on my harmonic analysis and input that I can safely assure you that nobody here needs to be afraid of what I am doing.

Regarding the many method books you mentioned: do any come to mind that might lend themselves to the task at hand?

Read literally any jazz improv method book and see what rules you can distill from it. Others in this thread have cited the Bergonzi and Coker book. There's tons of free material on the Jamey Aebersold site or the Berklee MOOC site, including a number of free items. You have to figure out which (if any) of these lend themselves to your systemization. Don't ask others to figure that out for you (unless you actually are an AI, in which case, congrats to your programmer for getting me to cough up even this much).

John

[Jabbarn]

The OP’s concept sounds to me like a fun and fascinating project. Nothing more.

[cosmic gumbo]

Thelonious Monk said all musicians were subconsciously mathematicians. It has to do with pitch and rhythm, which have applied math values, and we learn how to put it into practice.

Software developers, always on the lookout for talent that can code/program have established that musicians are superior candidates for having the abstract skills needed to learn and excel at this type of work. They've gone as far as try to recruit college music students into changing their major to computer science, offering them lucrative career options, that music has a hard time competing against in lifetime income. Plus, they tell them, you don't have to give up music, just learn to write code, be a musician too....I play with a bone player that is a corporate programmer, and he's worked from home for the past 7 years.

We think jazz is pretty abstract in practice, but it involves certain deeply imbedded musical logic in many of the choices made in the moment, ideally without thinking about it.

So...for the OP, Coltrane liked math and rules to help create. John Coltrane's Music & Geometry

Everyone's favorite jazz internet indexer Bob Keller ( Bob Keller's Jazz Page) is involved with his school's computer science department and by applying "improvisation rules", has had this software available for years, newly updated again this year. Maybe it will help your research. Impro-Visor. Welcome to Impro-Visor

[Jazzstdnt]

The OP’s concept sounds to me like a fun and fascinating project. Nothing more.

In the end it's grad school. We must do what we must do, to earn our 4.00 GPA. Been there, done that.

[pauln]

I would have thought that this myth would be seen as such by musicians. The concept of perfect pitch is a fundamentally flawed concept because of various reasons, some being:

1] It assumes a concert pitch reference like A=440Hz, but actual concert pitch has historically varied over more than half an octave in the last few hundred years. Modern concert pitch of A=440Hz was standardized only about 80 years ago.

2] It assumes a temperament, but there have been over two dozen temperaments, a half dozen enjoying historical popular use, the convergence to the modern standard being relatively recent.

Perfect pitch would have to be in regard to a specific concert pitch and in a specific temperament, yet I have never read or heard of anyone confronted with a perfect pitcher asking the most obvious of all possible questions, "What is the concert pitch and temperament of your perfect pitch?"

Are we to believe that perfect pitch just happens to comprise both the modern standards of concert pitch and temperament?

3] Chromatic notes are 100 cents apart... how much error in naming a pitch is allowed to count as correct for a perfect pitch test? The maximum error is +/- 50 cents before coming closer to a contiguous note.

4] Louder is perceived as sharper, softer as flatter (the reason why headphones cause problems in the studio monitoring pitch when playing very loud).

5] Lower in frequency is perceived as sharper, higher as flatter (the reason for "German Tuning" on pianos and "Sweetened Tuning" on guitars, to compensate).

The variations in perceived pitch with level and frequency range are corrected by ordinary musicians in order to sound in tune. Does the perfect pitcher hear perfectly based on frequency (and hear the corrected music as out of tune) or does he hear the same variations and correct for them?

Until the very concept and definition of "perfect pitch" is rigorously developed and made clear, it just serves as an uninformed label for someone with a "good ear", altogether confounded with relative pitch.

[destinytot]

Never fall in love with an operating system:

[grahambop]

(unless you actually are an AI, in which case, congrats to your programmer for getting me to cough up even this much).

Maybe we are all AI, with the ultimate programmer. At least that’s what you’re liable to read in the bible...

[grahambop]

Some Wayne Krantz ideas:

Get In The Zone: An Intriguing Improvisation Strategy from Wayne Krantz - GuitarPlayer.com

[destinytot]

[Jabberwocky]

Sebastian, there is a computer science professor at Harvey Mudd College who created a programme which did something like that. It is on the web if you do a search for it.

Fun stuff. Ah, Welcome to Impro-Visor .

[Jabberwocky]

REU: Music Software

[diminix]

Bob Keller published something about the design abstracts of his learning tool:

Blues for Gary: Design Abstractions for a Jazz Improvisation Assistant - ScienceDirect

PDF can be downloaded there, might be interesting for the OP.

A lot of discussion in this thread is rotating around the prospect of a machine improvising jazz. Though that might be a horrible vision for some, isn't almost all beginners searching in this and other forums exactly the same? How can I improvise jazz? Is there any system, any rules helping me to learn that? And after reading stuff here for about five years, I am still waiting for the emergence of an reliable system to learn jazz improv. I haven't seen too many JGBE members that where brought onto a learning system here and came out as good improvisers... But I'm still not giving up!

Being a software developer myself with a mindset working in a logical and systematic way, the OP is working in an interesting field.


Robert