Piano World Home Page Forums Our Most Popular Forums Adult Beginners Forum Jazz Study Group 2: Advanced Players

Excellent, there is nothing like playing with a really good drummer

Very cool knotty!

Beeboss,

Check out this tune on Youtube. I think it's pretty good, and an example of what I'm talking about re algorithms.

http://www.youtube.com/watch?v=QEjdiE0AoCU&feature=related

It is very interesting, but in a sense all the computer is doing is mashing up lots of pieces that it has already been told are good and then producing a composite of those according to the instructions of its human programmer. We have no way of knowing how many 'bad' compositions it came up with first before the human programmer chose that piece and even that piece wouldn't sound half as good if it was played on a bad synth.
But I do take your point. Maybe at some level there is very little difference between what our brains do when we create and what the computer is doing. It is an interesting philosophical problem.

But even if the computer can 'choose' combinations of notes that, when processed in the right way, we perceive as pleasant that doesn't mean that other combination cannot be made to sound good in the right context. I wonder if a string of notes can be found that are impossible to turn into good music?

Band in a box can generate decent solos on the fly.

Heres a blues a played today
It won't stay there for long as I am not very happy with it
I was trying out some different ideas

http://soundcloud.com/david-beebee/blues-beebee

The programmer was/is a composer that became obsessed with writing a memorable composition--something that would be timeless and unquestionably beautiful. This lead him to think of what constitutes good melodies, harmonies and rhythms, and since he was a computer programmer thought to write out the algorithm that may help him better understand what it takes to compose. So, apparently Emily writes only 'beautiful' tunes. I believe the algorithm is limited to the range of classical music (baroque to romantic, I think), but evidently it has produced works that have fooled many music professors when they have been subjected to blindfold tests.

I'm not sure I see it as a problem or a breakthrough. I can see how many composers and musicians and music lovers would become upset to know that the secrets of music are slowly becoming revealed. Somewhat like discovering that the earth is not flat, or that the Sun is the centre of our little universe, or that the process of putting caramel into the middle of the Caramilk bars really isn't too mysterious.
I think it gives me hope that if I continue down the path of learning how to play, that I'll come closer and closer to being able to express myself with consistent musicality.

I'm not sure what you mean. I think the point is that there really is NOT a limitless amount of possibilities for any given musical vertical and horizontal slice, but that the number is exponentially larger than what we may be able to imagine, and also larger than what most people would be able to think of, especially in an improvising situation. We are all conditioned by our environment, past and present, whereas the computer IS the environment. And if the programmer changes the environment, the resulting output is altered as well.

Define good. Maybe someone out there loves the sound of a succession of b9 intervals, but I think most people from most cultures might find it unpleasant. So, good to some is always possible. Paul Bley and Cecil Taylor seem to find music in the cracks, so to speak.

Ya, I agree you should take this down right away. Horrible stuff, David. You really ought to be ashamed.

I say leave it up unless you have something equally cool that you're equally unhappy with.

How do you program the drum endings? Are there loops for last bars? I think I need to check out that drummers website...

+1

You've said this before and I didn't quite believe you then. But Pilc's comment matches what several of my teachers say.

At the beginning, I was given strict rules on outlining harmony. Then they showed me how they threw those same rules out the window.

Beeboss -- Blues -- first of all that was a nice groove! I liked the surprise at 1:44.

Leave it on.

I wonder if musicians were forced to follow just the 'expected' rules, would McCoy Tyner's music exist?

I suppose I just disagree that the secrets are becoming revealed. I don't have any objection to computers improvising or composing and it may be possible that at some time their skill at doing it may surpass human skill, BUT (and it is a big one) they are only doing what they are being told, the programmer is really doing the composition but in a different kind of way than is usual. Also the programmer is the one that is eventually evaluating the quality of the composition, deciding whether it is good or not.

In a sense all the secrets are already revealed and open to us - if we agree on which works are 'good' then the notes in those pieces are already completely accessible to us.
However the mental processes that went on to put those notes in the exact places will always remain hidden. We can only have a guess at what they were.





Whether the choice is infinite or just exceedingly vast is not really that relevant I think. It is surprising how fast the possibilities add up, how the number of possible choices of a few random notes quickly exceed the number of atoms in the universe. I calculated that once and was surprised at the low number of random notes it would take. That is not even taking into account rhythm, dynamics, timbre.






That is exactly my point, that we rely upon human subjectivity to decide what is good and it is not conceivable that there could be an objective way to do it. That is why I don't think there are likely to be any objective rules to create such goodness.

The drum track just ends and I played along with it. It is not exactly in the ideal place but it is just for practice. I could have edited it up I guess if I was bothered.

There are no loops, just audio tracks at different tempos.

There are no rules, there are only choices that musicians and composers make.
imo

Daughter nr 3 is taking her mid-day nap, daughter nr 2 is out on her sledge - dad fools around with BiaB.
Giant Steps: ten sax solo + bass line, drums all generated within BiaB. No editing (just a little balancing and reverb), I use the RealTracks + RealDrums packages.
http://www.box.com/s/9u8uoj9zxjkrl45ofd99

Drums, bass and pno comp:
http://www.box.com/s/f82b3c0x8ti5yft6b921

Drums, bass and pno solo:
http://www.box.com/s/rpcsuai2h7n8um8j16nc

Well played on Giant steps Chris. Your solo illustrates how far the computer model has to go ;-)

>>http://www.box.com/s/9u8uoj9zxjkrl45ofd99
It's really amazing what this thing can do.
He even almost quotes the melody sometimes (1:50).

Do they have a McCoy plugging where they'd weave in and out ?

It gets repetitive but at least he never loses the form

>>http://www.box.com/s/rpcsuai2h7n8um8j16nc
Apart for the lame sound, the solo's not bad. Nice chops (duh) and a few nice piano cliches. Some odd intervals, not very typical of piano playing, but this is really crazy what the computer can do.

Yeah, it's crazy - and fun in a perverse sort of way.
A couple of clicks, change the Style, change the sounds - and hey presto there's a Scofield wannabe inside all of us pianists.
http://www.box.com/s/jbj5fhd537ysie4bukof

wow, amazing. The pub would be on fire!

Sounds awesome to me! Might be a little slippery in a couple spots compared to your usual stuff... but it works that way.