Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[IIO]

my current "max farm" is an XServe 8x2,25 Xeon and i can tell you it can do a LOT (compared to a G4!)

with its dual gigabit network you can control it with OSC or IPMidi easily when used a music instrument - but of course i mostly use it to render stuff offline.

first thing i will do with a working-audio OS9-mini is releated to tokio rather than max. or a combination of both?

[dr bu]

hjälp, hjälp!

Right, thats the hotline IIO-button/menu. Not fully implemented yet.

[IIO]

tomorrow a bunch of small audio IOs arrive here, and i will start setting up an OS9 mini with max & co.

[IIO]

i deleted the above paf patches.

1. the naming scheme was weird; the phase is not the carrier, it is the phase.
"carrier" and "modulator" are now properly used as specified by the ucsd.edu tutorial.

2. phase of master phasor can be reset at audio rate now, so that you can synchronize many oscillators of the same (or different) frequencies.

3.  included alternative layout using single connections between the modules in the "-x" versions of the patches. see example file.

use the "-x" ones as default; only go back to the multiple-inlet versions when you have trouble with the "-x"  stuff.

(the zip is made in OS9 using stuff it - for OS9 pls remove the .mxb)

[dr bu]

Ok.. interesting 

[IIO]

eventually it is a bit confusing to use note numbers (or actually: numbers of notes) for the bandwidth. you might want to use octaves instead.

[dr bu]

puckette made [formant~] back in `95 (found in IRCAM PMA-lib) presenting "live formant tracking"!
i want to feed its output (3 formants) into pafs and fofs but helpfile is way over my head.   

your nomenclature is impressive. it does not always suit my needs, but i steal what i can.   

[IIO]

i dont think i ever used that, if it is for the FOF object then it is very oldschool.

however, i have tried to make the two available vowel libraries work with the PAF long ago - and it is not so easy.

the IRCAM one only works with partials and has no bandwidth parameter, and the CSound database has bandwidth - but less vowels.

you can try to use my new helpfile and apply 3-formant vowel data to it, you will see that it doesnt work until you start to freely scale the data.

and of course, if you apply soprano vowels to you filters/partials/fof/paf, make sure you only play base frequencies in the correct range.

[IIO]

oh, i misread, that was for recording/analysing? i´ve used praat once but it feels like trial and error.

[dr bu]

i´ve used praat once but it feels like trial and error.

sure does. and its not real time.

[formant~] comes in PD too. if i were not so shy and lazy i would ask those guys to explain it..

[IIO]

i´ve made up this data by averaging the values from dozens of different sources.

russian 28 48
black 36 51
basso 38 63
baritone 43 68
tenor 48 70
alto 52 75
mezzo 55 78
soprano 60 81
falsetto 52 89

only when you stay in this frequency range, applying vowel formats will "work" as supposed to.

[IIO]

installed formant~

probably not difficutl to tell whart the output is - but i dont get the output data to change at all (when testing with the FM) wtf?

however, i think it´s LPC. i.e. not so much "fiddle" at all here, rather its own class.

on OSX you have plenty of alternatives... FTM, rtcmix, the zsa stuff...

p.s.
there is also an object called lpc from ircam, which is OS9 only. but it only does the lpc itself and leaves you with a bunch of IIR coefficients.

p.p.s
while you probably understand what my "justine" does , since the helpfile is missing i should add that you have to add the key as argument in order to get the right ratios (0-11 for C-Bb)

[dr bu]

So yhis guy: https://www.cycfi.com/2018/03/fast-and-efficient-pitch-detection-bitstream-autocorrelation/ appearently found an interesting way to grab f0. I dont know phyton so pondering if it possible to realize in old squarebox Max?

https://github.com/cycfi/bitstream_autocorrelation/blob/master/bcf.py
zerox/weave, + bit operators + ??

[IIO]

when i remember correctly, correlation analysis patch is difficult in OS9 because it requires an object we dont have.