Subject: Re: [linux-audio-dev] protux, stereo and interleaving
From: David Olofson (david_AT_gardena.net)
Date: Sat Feb 10 2001 - 10:47:13 EET
On Friday 09 February 2001 15:56, Paul Davis wrote:
> >frames are interleaved by audio channel. Overall, interleaved data
> > seems much more generic, and real-time friendly. Is it really
> > such a performance-killer - the time to [de]interleave data is a
> > knats-p in a millpond compared to doing 600+ FFTs in real-time...
>
> true. but if you store a 40 minute 24 channel recording in
> interleaved format, editing it becomes extremely difficult, even
> with
> non-destructive techniques. destructive techniques would require
> rewriting all 24 channels at and beyond the location of any
> deletion. non-destructive techniques can avoid the rewrite, but you
> end up either doing massive amounts of duplicate data i/o and/or
> using a very complicated caching system between the file and the
> presentation system.
>
> i don't think its a coincidence that most high end audio software
> suites use non-interleaved storage formats.
It could *possibly* make sense to store stereo channel groups
interleaved to reduce seeking, but I'm not really thinking about
storage formats. I'm only interested in chains of plugins with lots
of channels in parallel.
> and as someone else pointed out, SIMD encourages this approach as
> well. i'm not sure that we have any good SIMD-capable compilers at
> this time, but thats just a detail :)
Well, there's always asm. ;-)
//David
.- M A I A -------------------------------------------------.
| Multimedia Application Integration Architecture |
| A Free/Open Source Plugin API for Professional Multimedia |
`----------------------> http://www.linuxaudiodev.com/maia -'
.- David Olofson -------------------------------------------.
| Audio Hacker - Open Source Advocate - Singer - Songwriter |
`--------------------------------------> david_AT_linuxdj.com -'
This archive was generated by hypermail 2b28 : Sat Feb 10 2001 - 13:28:08 EET