Re: [linux-audio-dev] 32/96 x 24 playback improvements, I think

New Message Reply About this list Date view Thread view Subject view Author view Other groups

Subject: Re: [linux-audio-dev] 32/96 x 24 playback improvements, I think
From: Benno Senoner (sbenno_AT_gardena.net)
Date: pe helmi  04 2000 - 11:10:05 EST


On Fri, 04 Feb 2000, Paul Barton-Davis wrote:

>
> Its true. But... the problem with EIDE controllers is that they
> require single-command interactions with the kernel. Its impossible to
> queue a bunch of commands, and then go and do and something else until
> the controller comes back and tells you that they are finished. As
> long as you only want to do one thing with the disk(s) at one time (as
> most standalone HDR systems do), this is not a problem. If you want to
> run Linux and do HDR and still be able to do other things that use the
> disk from time to time, this is a problem. Note: I am *not* talking
> about running the GIMP while an HDR session is going on. But I do find
> that I can run "ardour" in record mode, with the tape files on one of
> my regular system disks, then run fetchmail, do some web browsing, and
> program editing, without causing a single dropout. This has something
> to do with the fact that I have a dual CPU system, but it also has
> something to do with the fact that I am using a SCSI disk controller.
>
> --p

I think it's not the dual CPU but the 2nd disk which makes the difference,
I agree that on IDE if you run a master/slave configuration there could be
bandwidth problems if you load the channel heavily,
but assuming that you run with 1 IDE disk per channel,
then you have the fulll bandwidth available on each disk,
I am sure that using this configuration, you can browse the web and do HD
recording without problems.

But the funny thing would be to browse the web and do HD recording on the
SAME disk. :-)
but linux lacks the realtime disk scheduler, therefore we HAVE to use 2 disk,
wheter we use SCSI or not.
:-)
I remember the QLinux project has the goal of implementing a realtime disk
scheduler, but it will pass much time until this (if) becomes reality, and goes
into mainstream kernel.

Meanwhile we have to do our best to squeeze out as much as possible from
the disk using our sophisticated tricks.
:-)

Juhana: answering your question:
> But one thread per disk makes sense, or?

Yes, it makes perfect sense, that is what I had in mind,
because you have complete control over all audio disks.

Benno.


New Message Reply About this list Date view Thread view Subject view Author view Other groups

This archive was generated by hypermail 2b28 : pe maalis 10 2000 - 07:23:27 EST