Hi all, I'm working on a patch that pulls a random wav file from a directory, plays it and then when it's getting close to the end, opens another random file and fades from one to the other. It's a pretty basic idea, but I'm pretty new to pd, so almost every step of that process has been challenging.

I've managed to get most of it put together by watching different tutorials and using example patches people have posted, but there seems to be a big limitation that I haven't found a good solution for yet. I am hoping to use wav files ranging from about 5~10 minutes long.

I'm currently reading them into arrays and playing them with tabread4~, which basically works except that it freezes up when loading either file, meaning it will be seemingly impossible to use this method to get an uninterrupted stream of sound. It also doesn't seem to handle 10 minute long files very well - they seem to be cut off from what I can tell in the array.

Alternatively, it seems like I could use readsf~, but then I'm not sure how I could know when to load/start the coming file, since it doesn't seem to offer the control that tabread4~ and arrays have.

Oh, I'm using Linux by the way. Thanks for any help you're able to offer!

Dear Psiforce

check out this topic:

http://puredata.hurleur.com/sujet-3507-long-soundfiles

there also is a problem with adressing long files by the limitation of the 32bit floats pd normaly uses, you can use the fantastic pd-double:

explenation:

http://www.katjaas.nl/doubleprecision/doubleprecision.html

download:

http://puredata.info/downloads/pd-double/releases/0.43.1

cheers,

-j-

a (not very elegant) solution in pd vanilla could be the
use of 2 readsf~ objects, the 2nd with an offset which has the length of
the transition. when the 2nd readsf reaches the end, it is time to crossfade.

and there might be an object in extended to get the filesize without loading
the entire audio into in array. (system? shell?)

Hi Johnny, I did see that topic when I was looking for a solution. I also watched katjav's youtube video (and her other videos, which are all awesome!). Thanks for linking her explanation, which I didn't see before. While I was impressed with her efforts, I didn't really understand how it related to soundfiler~ until reading that page.

It actually brings up an interesting point further down though. Maybe my problem isn't necessarily the use of floats, but the inability to load sound files in the background without the sound dropping out. I thought the dropout was related to the long files, but maybe that isn't the case. Will pd-double even solve that issue?

mnb, I think that second suggestion might allow me to simply use readsf~ if I can find out how to do that. I will look into it.

you dont have to use shell or something like that in extended, there is filesize, or fsize, which outputs the size of a file in bytes. if you know the samplerate, bit-depth
and header-size, it is easy to calculate the length of the audiofile.

if you want audio in tables (without dropouts while loading big files),
you can use readsf~ in a reblocked subpatch, http://puredata.hurleur.com/viewtopic.php?pid=5387#p5387
and you could split huge audiofiles into multiple tables, to avoid the problems with float precision. but i dont think its worth the effort.

Yeah, actually I have already redone it with readsf~ and overall it's way simpler and more elegant so far than it was with soundfiler~, arrays and tabread4~. As of now, I am fine loading new files and playing them without dropouts. I have also figured out how to get the length using [soundfile_info] (which also give me the samplerate as well.

I still haven't quite gotten it working yet though. I am getting weird issues not being able to use [timer] or [realtime] to compare with the time I want. I wanted to port part of functionality into a separate patch, but I think I may be creating unnecessary problems for myself that way and may have to just do the whole thing in one patch after all. Anyways, the original issue seems to work well combining [readsf~] and [soundfile_info], so thanks for the suggestions!