LISP code for a specifica problem - Help

[fdsdb]

Hi All,

I am trying to resolve a specific poblem, but I can not find out the right Lisp code to do it, so any help will be appreciated, thanks in advance.

My problem is the following: given a whatever series of pitches, I want to randomly transpose some of them to superior octave, and then repeat the cycle until all pitches are in the superior octave.

i have set the code for random transposition:

(lambda (series)

(loop for i in (flat series) collect
(om::omif
(and
(< i (om::om+ 1200 (om::list-max series)))
(> i (om::list-min series)))
(om::om+ i (om:nth-random '(0 1200)))
i))
)

and tried to use it inside an iterative process:

(lambda (series)
(flat
(loop for i = series
while (equal (om::list-min i) (om::list-min series))
do (loop for k in (flat i) collect
(om::omif
(and
(< k (om::om+ 1200 (om::list-max series)))
(> k (om::list-min series)))
(om::om+ k (om:nth-random '(0 1200)))
k))
collect i))
)

but OpenMusic go into an infinite loop. Or, in another attempt woith different code, it transposes also more than one octave high the already transposed pitches.

I have looked for useful examples in some LISP book, but I can not arrive to a solution.

Many thanks in advance, ciao

Fabio

[haddad]

Dear Fabio,

your description of a superior octave is rather imprecise. But anyhow, i did an example here with a superior register based on an octave. I am sure you can adapt it. It is not in lisp code, but in VPL. In lisp code it is even easier.

Here you go:

The patch:
rand-reg-trans.omp (32.2 KB)

tell me if you need a lisp version or some modifications…

Best
K

[fdsdb]

Dear Karim,

Thank you very much for fast reply. Yes, sorry, superior octave indeed is higher octave. “Superior” is good in Italian…
The patch is good for my aims, now I am going to try to convert it into LISP, I hope I will need not further help… Otherwise we will meet again
Meanwhile thank you very much again, ciao

Fabio

[fdsdb]

Sorry, it’s always me.

I have tested more deeply the patch, there is only a little difference respect to what I am looking for: the already transposed pitches in successive loop cycles sometimes can return to original register, but I would not want that.

To try to more specific, I am searching something of similar: suppose we have the sequence c3, d3, e3, f3, g3; in first loop cycle I could have c3, d4, e3, f3, g4, for example, that is notes d and g have been randomly selected and transposed, and will remain always into the higher register. In second cycle, for example, I could have c4, d4, e3, f4, g4, that is the new transposed pitches now are c and f, so that in final loop cycle I will have only the last note to be transposed, that is e.

I hope I am clear, my English is not always very good…

Thanks in advance for help, ciao

Fabio

P.S.
Just for sake of exactness, that is a composition technique by Donatoni, he used it at least in “Lumen”, I discovered it during my analysis of the work (Guarda Figura, processo e articolazione parametrica in Lumen di Franco Donatoni - only in Italian, sorry)

[haddad]

Dear Fabio,

In the patch i sent you i doesn’t do what you are saying (pitches draws back to original register). I tried a lot of evaluations and here is one of them:

Maybe in your code you did something wrong???
Can you send please an example???

Best
K

[fdsdb]

HI,

My fault, you are right.

Only a last consideration, however, if I may: I have tested the patch starting from this simple sequence:

and in some evaluation the following occurs:

As you can see, the complete transposition is reached into five steps. But, because of random transposing mechanism, it happens that second and third step, if I read good, are the same, that is no pitches are transposed from one step to the other.
Is it possible to avoid it? I hope I am not inopportune in asking a supplement of help. Just to respect Donatoni’s process.

Besides your algorithm has given me another idea about to simulate another Donatoni’s composition process in “Lumen”, I must think a little about it… I am making a little update of my library, so I would be glad to enclose, with your help, also these composition techniques. This is the reason of my questions…

Really many thanks, ciao

Fabio

[haddad]

No problem Fabio. Glad to help.

Here is the way to remove duplicate series. Of course we remove these from the resulting list it is simpler:

And here is the patch:
rand-reg-trans2.omp (33.4 KB)
Best
K

[haddad]

Dear Fabio,

Here is a possible lisp code:
fabio-ran.lisp (985 Bytes)

Best
K

[fdsdb]

Dear Karim,

Thank you very much, it seems to me that everything runs perfectly.

Many Thanks, ciao

Fabio