Aight so listdev shoes the one I want as iten number three on both input and output.
Has anyone found Pd to be an improvement over max/msp in terms of CPU? specifically when it comes to using lots of audio cables, cause I like to do that but my computer doesn’t quite so much.
Any opinions about Purr-Data? It seems like a more modern choice to me (especially if embedded systems aren’t a focus), and I like that it was named because of cats.
Thinking about making the switch from max
I don’t know about the comparison between PD and Max, but Purr Data is definitely slower than PD Vanilla, due to its modern GUI. This probably doesn’t matter on a modern machine though.
Not done any comparison either, but I cannot think why pure data would perform better ( and a few why it might not). If performance is critical
gen~ on Max could become a key factor.
( if your using loads of externals then their performance might become important)
Has anyone here worked with neural networks in Pd?
I’d love to hear about your approach and/or externals used. Has anyone used ml-lib or ann library, or have you had Pd to talk to python or something else that runs a NN?
Personally I’d like to gets something working within Pd although I’m looking at a couple of options. Python seems like an excellent solution but I don’t have the time to get my head around that within the confines of my current project. The documentation for ml-lib is almost non-existant so I’ve been battling to get the ‘ann’ library up and running but I can’t seem to properly compile the ‘Fann’ library that it requires. I’m working on OSX with this at the moment but will eventually be on a raspberry pi. Has anybody out there encountered this? Is anybody even trying to do this kind of thing?
So this isn’t Pd necessarily, but looking at this, I was curious to know of the possibilities of taking just the engine of the nsynth and making some kind of program for Rpi or other setups.
NYC-area Pure Data (learning) event coming up:
i made a 4 voice neural patch for organelle last yer
–it would ahve been 6 but organelle could not handle the overhead
let me know if you want the patch.
ann~ the neural external i am pretty sure is not maintained but if you want the patch i’ll share it
but to be clear [after re-reading my post – the patch does NOT use ann~
I’m trying to make a patch in the latest version on Mac and cannot get [netreceive] to actually receive anything. I’ve used it many times in the past. Has anyone else run into this? I’m trying to figure out if it’s an issue with PD or something in my config.
I just went through setting up my latest acquisition (a beautiful 2011 grid) with vanilla puredata and thought this might be useful reference for anyone else wanting to try it out:
Now that Pd extended is really dead, I’m planning on updating the pd grid study to work with vanilla so that people don’t have to worry about downloading externals. It’s slightly more awkward than the routeOsc and unpackOsc objects, but simplicity and portability is worth it IMO. @tehn, before I start modifying the docs and preparing a pull request: is a pd-vanilla update of the grid study something you want and are likely to accept a pull request for? If not, I’ll write it on my own site.
Count me interested !
Is there an application like mlr designed in pure data, available to use ?
Axiome 0.4, someone took over the reigns!
Oh wonderful, I’d love to have a look at the patch! I’m not familiar with VaMilla~ ? Do you mean vanilla?
I’ve wound up using ml.star in Max for my current project, but I’m still keen to look at other options for the future!
This is my latest patch (eventually to be attached to the monome):
It needs any wav file beginning with a number such as “1.wav”.
The problem I’m having is that it doesnt run efficiently and Im not sure what to do.
It plays back a sound file but when I need to load another file the whole patch freezes so much so that it stops the audio play for a moment. Its very annoying becuase I’ve spent a long time figuring out the patch and now It works it at the same time doesnt. Please may I have some help improving this to make it usable?
i would load all the audio files you want to use at the beginning then use [set( to switch between arrays.
in the Pd Help Browser:
you may also want to look at the [pd define] subpatch in the [array] help file.
[array define -k] optional -k flag to keep contents
would this work for 30 odd samples that last for 10 seconds or so for an hour set?
I just discovered:
and this loads audio in the same why as my patch so far ? although there is also and gap of silence but alot shorter than mine somehow?..What does mlrv do, surely that loads files into tables as well? But that seems to work successfully?
i don’t see why not. but i guess it depends on what machine you are running the patch on.
here is how i go about it sometimes (an incomplete in process abstraction) there are many other ways of doing this
p.s. if you prefer using one table or array only you may be able to reduce audio dropout time slightly by using the [table] or [array] objects or putting the table in a subpatch. could be because the array itself is visible in your patch as you are loading the file.