David Medina

I'm trying to translate an old Pure Data patch that I built to create a chorus of Text-To-Speech voices (each voice just saying syllables that were prerendered using a TTS system that could produce decent results in spanish at the time). The system used another script (made in Processing) that calculated the sun position (altitude and azimuth every few milliseconds). These numbers were then used to trigger each "singer". The singers (around 8 different voices) were built with an Euclidian sequencer and some effects, each one only playing files from a assigned folder. Then each voice audio would be sent to a circular panning subpatch where the audio signal would be assigned in a 360 degree space between 8 speakers.  A partial example bellow (excuse the horror):

I'm trying to work the very first section with max and put together this simple sample player for some random syllables in a folder:

Excuse all the newbie questions and thanks for any help.

Hi!

I'm trying to translate an old Pure Data patch that I built to create a chorus of Text-To-Speech voices (each voice just saying...

octophonic-chorus-euclidian-sequencers-sample-buffers-fx-chain-sends

not sure if that helps, but polybuffer~ roughly corresponds to buffer~  inside a poly~, which is a bit more work but also offers more flexibility. 

poly~ has a certain learning curve and is quite different from pd´s clone. 

preloading media or not? should imho be a matter of a conscious decision what you want, not of programming style. audio is today mostly small enough to fit in RAM, on the other hand there are also contraindications about having poly~s or polybuffer~s with 1000 voices.

you´ll find a post somewhere here where someone suggests to use an algorithm lended from a (graphics) line antialiasing method, which outperforms any possible björklund algo implementation by some aspects.

but for up to 16 beats, precalculating and writing into a [coll] should be "best".

Thanks for the suggestions, Roman, I'll try this out.

Octophonic Chorus - Euclidian sequencers, sample buffers, FX chain sends