Re: [linux-audio-dev] hdr disk throughput

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

Subject: Re: [linux-audio-dev] hdr disk throughput
From: Andrew Clausen (clausen_AT_gnu.org)
Date: Mon Mar 13 2000 - 02:03:32 EST


Paul Barton-Davis wrote:
>
> i am a little worried.
>
> it seems to me that in order to offer "true professional level" HDR,
> one has to sustain twice the apparent rate of disk i/o.
>
> the feature that causes this is the ability to move instantly in and
> out of record mode. when transitioning from record to play, people
> expect no break in the audio stream. that means that you must have the
> playback data (from the disk) already read into a memory buffer ready
> for use. but since you have no idea when this transition may occur,
> you actually have to do this *all the time* you are recording - punch
> out could happen anywhere.
>
> thus, in a record-all-tracks situation, to offer this feature, you
> have to be able to stream the given data rate both to *and* from the
> disk.
>
> comments ?

I don't think this is a problem, as long as you record to disk in
the same place (or near to) the place you read from. The major
pressure on disk IO is the seek time.

I wrote a little test program, that
(1) Writes a buffer
(2) Reads in the buffer for playback (a few buffers ahead of the
write buffer)
(3) Syncs the buffer written in (1). This shouldn't happen
before (2).

If I comment out the write, there is no drop in performance.
I get about 4.5Mb of reads per second on my disk, either way.

I attached the program. Are there any issues I missed?

BTW, you might want to run them on a raw partition, and a
filesystem, to compare performance. If you don't have
a spare partition, get GNU Parted ftp.gnu.org/gnu/parted
(plug, plug...)

Andrew Clausen


overwrite-1.tar.gz


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

This archive was generated by hypermail 2b28 : Mon Mar 13 2000 - 23:40:51 EST