I know this question has been asked a couple of times before but i see that the posts are now quite old and, to be honest, i couldn't make sense of the replies as my understanding of puredata is somewhat limited. Basically i've been tasked with creating a patch that will accept full quality audio files, playing sections of the file back in a random order, to do this i understand that i will have to load the file into an array but the number of samples seems to be limited. Is there a way to lift this limit? I'm running windows 7 if that is relevant in any way. Thanks, Dan.
hi daniel
may this patch may help you
der.brandt
Zitat von "Daniel Embleton" crumbly.bwfc@hotmail.co.uk:
I know this question has been asked a couple of times before but i
see that the posts are now quite old and, to be honest, i couldn't
make sense of the replies as my understanding of puredata is
somewhat limited. Basically i've been tasked with creating a patch
that will accept full quality audio files, playing sections of the
file back in a random order, to do this i understand that i will
have to load the file into an array but the number of samples seems
to be limited. Is there a way to lift this limit? I'm running
windows 7 if that is relevant in any way. Thanks, Dan.
If you're just playing random sections of the soundfile, you could use [readsf~] and give an onset
-Jonathan
From: Daniel Embleton crumbly.bwfc@hotmail.co.uk To: pd-list@iem.at Sent: Friday, October 21, 2011 9:17 AM Subject: [PD] Array Size limits
I know this question has been asked a couple of times before but i see that the posts are now quite old and, to be honest, i couldn't make sense of the replies as my understanding of puredata is somewhat limited. Basically i've been tasked with creating a patch that will accept full quality audio files, playing sections of the file back in a random order, to do this i understand that i will have to load the file into an array but the number of samples seems to be limited. Is there a way to lift this limit? I'm running windows 7 if that is relevant in any way. Thanks, Dan.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
I think the OP wants to know what's the maximum size for an array in PD, so as to know how much audio he can fit in it unless I misunderstud his question.
I made a little non-scientific test here growing a table on a linux machine with 2 GiB RAM (and quad processor @ 2.4 GHz but I think it's not so relevant for array size). The machine started to get a little less responsive at array size 4.2984e+08 that is about 2.5 (mono) hours at 48Khz sampling rate and about 2.7 hours at 44.1Khz (still mono). That would give you roughly 1 and a quarter worth of stereo audio to put in arrays. I think RAM quantity is the key here.
Of course you should keep in mind extra processing etc. not just the array size.
Do you know in advance how long the audio will be?
If you have multiple files (instead of a single big one) you might consider using 2 or more arrays and switching between each other while loading them. Another approach might be to do the random cutting offline by splitting the big file(s) into small random chunks through a programme/script and then loading and playing those (could be another random layer there too)
Also I'm not aware if there are externals to stream audio from disk, and furthermore being able to start at a discrete point in the audio
Lorenzo.
On 21/10/2011 17:08, Jonathan Wilkes wrote:
If you're just playing random sections of the soundfile, you could use [readsf~] and give an onset
-Jonathan
------------------------------------------------------------------------ *From:* Daniel Embleton <crumbly.bwfc@hotmail.co.uk> *To:* pd-list@iem.at *Sent:* Friday, October 21, 2011 9:17 AM *Subject:* [PD] Array Size limits I know this question has been asked a couple of times before but i see that the posts are now quite old and, to be honest, i couldn't make sense of the replies as my understanding of puredata is somewhat limited. Basically i've been tasked with creating a patch that will accept full quality audio files, playing sections of the file back in a random order, to do this i understand that i will have to load the file into an array but the number of samples seems to be limited. Is there a way to lift this limit? I'm running windows 7 if that is relevant in any way. Thanks, Dan. _______________________________________________ Pd-list@iem.at <mailto:Pd-list@iem.at> mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Le 2011-10-21 à 13:17:00, Daniel Embleton a écrit :
I know this question has been asked a couple of times before but i see that the posts are now quite old and, to be honest, i couldn't make sense of the replies as my understanding of puredata is somewhat limited. Basically i've been tasked with creating a patch that will accept full quality audio files, playing sections of the file back in a random order, to do this i understand that i will have to load the file into an array but the number of samples seems to be limited. Is there a way to lift this limit? I'm running windows 7 if that is relevant in any way.
You can set your own -maxsize while using the [soundfiler].
If you are in 32-bit mode, you'll have a max size limit of something between 2GB and 4GB, before pd claims out-of-memory (nowadays this can happen due to 32-bit mode, instead of because of RAM size).
Be careful about the precision of numbers in [tabread~] and [tabread4~]. Use the right-inlet (available since pd 42, I think).
[readsf~] is good if your files are too big, or if [soundfiler]'s startdup time is too long, or if you don't need the kind of instantaneousness that [tabread~] has.
But with [readsf~] the messages take funny arguments when you want to skip any part of the sound file, because there is no easy «seek» option in milliseconds... you need to close and reopen it with six arguments, one of which is a value that you have to compute using information that you don't really have.
| Mathieu BOUCHARD ----- téléphone : +1.514.383.3801 ----- Montréal, QC
-
brandt＠subnet.at -
Daniel Embleton -
Jonathan Wilkes -
Lorenzo Sutton -
Mathieu Bouchard