Subject: Re: [linux-audio-dev] Software filter engines for high end audio and now DSP's
From: David Olofson (david_AT_gardena.net)
Date: ke helmi 16 2000 - 23:25:47 EST
On Wed, 16 Feb 2000, Anders Torger wrote:
> On Wed, 16 Feb 2000, you wrote:
> > That is why I am still dreaming about an "audio operating system",
> > ( MuCoS) which would provide an abstraction layer to run your
> > native audio processing components on top on a high performance OS.
> > (in this case Linux).
>
> How about RT Linux? It exists today.
RTLinux is a real time kernel without full support for hardware, like
soundcards. (You can't use standard Linux drivers w/o losing the RT
performance.) Further, it requires root access rights (to load
modules), and it lacks protected memory, partly for performance
reasons. It's as far from "an abstraction layer to run your native
audio processing components on top on a high performance OS" you will
ever get...
And, MuCoS is a multimedia plugin/integration infrastructure - not
an OS kernel. Although there will be a low level transport layer
that reminds of/could be used like IPC + shared memory of a normal
RTOS, it's still a rather specialized multimedia API.
> By the way, has anyone tried clustering machines (preferrably Linux boxes) to
> do real time audio processing?
Not tried yet, but I'm very interested in it. :-)
> I'm especially interested in FIR filters with
> more than 15000 taps. It must be a problem to keep latency low in these
> systems.
Yep, you can forget about any normal SAN solutions, with the
possible exception of hardware memory mirroring interfaces. The
problem here is most likely the protocols and not the hardware. From
the figures I've seen so far, dedicated ethernet should be just fine,
at least with RTLinux drivers.
The important part is to trade the ultra low network latency for
bounded maximum latency and guaranteed bandwidth.
> I have thought of the idea to use digital soundcards for data transfer
> between the machines in the clusters, must be perfect for low latency hard real
> time data transfers [at limited rates]. However multichannel digital sound cards
> are a lot more expensive than network adaptors, so I guess it is just a wierd
> idea :-)
Not really; professional audio cards can provide sample accurate sync
between the systems, which is very important in some situations.
However, for the raw data transfer between the cluster nodes, only
$$$$-solutions can beat ordinary Fast or GBit ethernet PCI cards with
point-to-point connections. You probably have to avoid switches and
the like and use more cards instead, but as they're almost approaching
the price class of the cables that you use to connect them, that's no
big deal... ;-)
//David
.- M u C o S --------------------------------. .- David Olofson ------.
| A Free/Open Multimedia | | Audio Hacker |
| Plugin and Integration Standard | | Linux Advocate |
`------------> http://www.linuxdj.com/mucos -' | Open Source Advocate |
.- A u d i a l i t y ------------------------. | Singer |
| Rock Solid Low Latency Signal Processing | | Songwriter |
`---> http://www.angelfire.com/or/audiality -' `-> david_AT_linuxdj.com -'
This archive was generated by hypermail 2b28 : pe maalis 10 2000 - 07:23:27 EST