thread progress report
|
View:
New views
6 Messages
—
Rating Filter:
Alert me
|
 |
thread progress report
by Bill Schottstaedt
::
Rate this Message:
Some first impressions of Snd/CLM and threads: on a machine with
2 quad-core 3.0 Ghz 86-64 processors (8 "cores"), I can currently get with-threaded-sound to run about 4 to 6 times faster than with-sound, and multichannel edits (filter-channel) run 6 to 8 times faster. I'm still chasing race conditions and so on, but this looks promising. Also, if you have 8 GBytes or more RAM, you can now use that 2^32 element table-lookup you've always wanted -- I think all vct, sound-data, and clm table sizes can be off_t's. Use mus-fft for an fft in that size range (and take a vacation while it computes). _______________________________________________ Cmdist mailing list Cmdist@... http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist |
|
|
Re: thread progress report
by Fernando Lopez-Lezcano
::
Rate this Message:
On Fri, 2008-07-04 at 07:20 -0700, Bill Schottstaedt wrote:
> Some first impressions of Snd/CLM and threads: on a machine with > 2 quad-core 3.0 Ghz 86-64 processors (8 "cores"), I can currently > get with-threaded-sound to run about 4 to 6 times faster than > with-sound, and multichannel edits (filter-channel) run 6 to 8 > times faster. I'm still chasing race conditions and so on, but this > looks promising. Wow, _very_ nice! I presume this applies to scheme only right? (I forget which lisp you were testing this with). -- Fernando > Also, if you have 8 GBytes or more RAM, you > can now use that 2^32 element table-lookup you've always > wanted -- I think all vct, sound-data, and clm table sizes can > be off_t's. Use mus-fft for an fft in that size range (and take > a vacation while it computes). _______________________________________________ Cmdist mailing list Cmdist@... http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist |
|
|
Re: thread progress report
by Bill Schottstaedt
::
Rate this Message:
> Wow, _very_ nice! I presume this applies to scheme only right? (I forget
> which lisp you were testing this with). Thanks! I'm kinda pleased too. I'm using guile right now -- I can't figure out sbcl's threads, can't remember whether ruby has threads, and I don't think fth has them. _______________________________________________ Cmdist mailing list Cmdist@... http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist |
|
|
multichannel files with up to 36 channels interleaved
by Ludger Bruemmer
::
Rate this Message:
Hi Bill,
I tried to generate interleaved multichannel files, especially for our 47 channel sound dome and ran into problems. I think interleaved files would be better to read for the discs because the head doesnt have to jump as much. It looks that the size of multichannel files is limited in absolute size. Computing those files resulted in shortened sounds-durations. The more channel there are the shorter was the duration. Is there any way I could create longer soundfiles, something like 36 channels, 48 (or even 96) kHz, 24bit and 25 minute of duration? would the counter of the runloop need a to be special format? The duration limit of the mentioned format is currently limited to a duration of 8 min53 seconds. thanks Ludger Prof. Ludger Brümmer Leitung Institut für Musik und Akustik Head of Institute for Music and Acoustics /////// |< |||| ZKM|Zentrum für Kunst und Medientechnologie /////// |< |||| Center for Art and Media /////// |< |||| Centre d'Art et de Technologie des Medias Lorenzstrasse 19 . D - 76135 Karlsruhe P.O. Box 6909 . D - 76049 Karlsruhe Phone +49 721 8100-1600 Fax +49 721 8100-1699 email: lb@... http://www.zkm.de/musik http://www.sumtone.com/lb Am 04.07.2008 um 20:24 schrieb Bill Schottstaedt:
_______________________________________________ Cmdist mailing list Cmdist@... http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist |
|
|
Re: multichannel files with up to 36 channels interleaved
by Bill Schottstaedt
::
Rate this Message:
Are you writing aifc files? If I remember correctly, they have a 32 bit limit
on the data size, which sounds like 8 minutes at 44100 36 chans -- (/ (expt 2 32) (* 44100 60 36 4)) is about 10? In "snd" (i.e. NeXT) files, the file size field should be ignored if it is wrong, and in "riff" (wav) files, Snd should automatically move to using the new extended format -- I forget the acronym. Apple introduced the "caf" format partly to get around the aifc limitation. In any case, 36 chans at 44Khz for say 36 minutes looks like 13 GBytes -- I just wrote such a file, but now I'm muttering about Snd which is taking forever to display it. _______________________________________________ Cmdist mailing list Cmdist@... http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist |
|
|
Re: multichannel files with up to 36 channels interleaved
by Ludger Bruemmer
::
Rate this Message:
Thanks Bill - that was the solution. It worked quite well with the snd format.
Just in case you would like to poke around with our easy to use N-Channel N-Speaker Vektor-based-panning Sound-Dome software, running on OSX, its public domain and open source soon: Ludger Prof. Ludger Brümmer Leitung Institut für Musik und Akustik Head of Institute for Music and Acoustics /////// |< |||| ZKM|Zentrum für Kunst und Medientechnologie /////// |< |||| Center for Art and Media /////// |< |||| Centre d'Art et de Technologie des Medias Lorenzstrasse 19 . D - 76135 Karlsruhe P.O. Box 6909 . D - 76049 Karlsruhe Phone +49 721 8100-1600 Fax +49 721 8100-1699 email: lb@... http://www.zkm.de/musik http://www.sumtone.com/lb Am 06.07.2008 um 20:53 schrieb Bill Schottstaedt:
_______________________________________________ Cmdist mailing list Cmdist@... http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist |
| Free Forum Powered by Nabble | Forum Help |