Converting words to specific pitch

michael watson's icon

Hi!

I occasionally use a composition technique called a musical cryptogram. I would love to try build a version in max. Simply, it extends the musical alphabet across all of the alphabet rather than the musical alphabet. There are a few versions but this is the simplest version.

Would love to build something that could generate music from typing a word in.
Do you think this is possible?
Thanks!

Roman Thilenius's icon

mine is called "guido generator" and transforms the vocals from texts into musical keys.

it distinguishes between a e i o u, ah eh ie au ei, and ümlauts (german) or likewise y ee ou and so on (for english) and voila, you have note numbers 0-11.

the assignments are done by the user, the frequency statistics are precalculated and shown to the user.

textedit, atoi, if 34 then 11.

Source Audio's icon

typing a word would mean a chord ?
and single letter one note ?

All you need is to assign whatever note number you want to a letter.
triggered word would simply recall notes as group, single letter 1 note.
you can try coll with indexes A-Z, and notes of your choice.

your letter lines, are they octaves ?
you can split alphabet into different then natural minor diatonic as on that
picture, like for example pentatonic, you can shift tonalities ...

I woud use textedit to capture key input and output typed word,
words or letters.
all it needs is to set some rules, for example if more than 1 letter or word is allowed per output, if yes, how much time should pass between separated words or letters, etc.

Are you using only capital letters, and if not should small caps trigger different octaves ?

Peter Ostry's icon

I work coll based. There are note numbers for letters according to frequency of occurence in my language and pauses for paragraphs, sentences, words, syllables and punctuation. For the text I use my own syntax with brackets and special characters.

Of course, this can be extended endlessly with emphasis, pitch algorithms, treatment of vowels and consonants. I did it in a relatively simple version just out of interest.

I wanted to explore what differences the musical interpretation of words and verse measures yields in different poems. I didn't get very far. It's not easy to control the boundaries between cacophony, comic music, randomness and washed-out drones so that interesting music comes out. It is a very large field of experimentation and Max offers many possibilities for this.

michael watson's icon

Amazing stuff! Would you be open to sharing any of these?

@Source Audio, I have been taking a word and creating a melody of single pitches. Once I've got the melody I work out what chords fit. I haven't been using different octaves for letters past g.
Cool, sounds like Coll is the go. I've attempted to learn max over the years but I think it's time to take the plunge and learn properly. Any other tips aside from the in app tutorials, YouTube vids and good old fun experimenting? :)

Source Audio's icon

with only 7 notes assigned,
you can easilly map all letters to your scale.

here is example that let's you select scale and fill coll.
typed letters and words will be ouput in a note list.
space is ignored

Max Patch
Copy patch and select New From Clipboard in Max.


michael watson's icon

Great thank you! I'll see how I go with all those.
Wondering if I end up struggling too much still, is there a place I can pitch the full project to pay someone to build it?
Thanks again :)

Roman Thilenius's icon

no [coll] required for 2D translation of smaller sets, [zl.nth] does it, too.

Source Audio's icon

using only zl objecs instead makes storing and recalling of scales impossible

michael watson's icon

Sorry I think I made things confusing saying pitch there.
Do any of you max geniuses hire out your services to build full patches?

Source Audio's icon

if you mean pitches are midi note numbers,
than there is no confusion at all.
What else could output be ?

I am not taking any jobs, but am willing to help to create such simple patches, but only if one says in plain text what one wants.
from the beginning to the end.