The Logic Off Topic list

Hi,

The last part of my tests: using different audio files. Well, as I
thought: this shows the bottleneck of IDE vs. SCSI: not using any DSP
plugins already causes lots of CPU load when using the ATA66 drive.

I got 24 MONO tracks, all using different 16bit 44.1 kHz files. The
entire system (OS, apps and audio files) are on the same ATA66 harddisk.
When adding a 25th audio file, I got an overload message. Very strange:
the metering is not correct: the audio load showed 40% and the disk I/O
showed 25% and still an overload when adding just one more file. I guess
that's a bug in the disk I/O metering. So, 24 tracks is the maximum
(adding a Fat EQ to every track got the CPU load to 50%, so there's
still room for some DSP stuff) using only this one harddisk.

Another thing I've noticed is that the disk I/O metering sometimes
doesn't work after an overload.

Ciao,
Joeri
--
Joeri Vankeirsbilck
joeri@...

List-admin   Logic-users/SoundD*ver-users/Logic-TDM
FAQ & Info:  http://www.nbdj.com/Logic/mailinglists.htm

http://www.nbdj.com  - Natural Born Deejays
http://www.nbdj.com/Logic - Logic Wish List

Joeri Vankeirsbilck wrote:
> 
> From: Joeri Vankeirsbilck <joeri@...>
> 
> Hi,
> 
> The last part of my tests: using different audio files. Well, as I
> thought: this shows the bottleneck of IDE vs. SCSI: not using any DSP
> plugins already causes lots of CPU load when using the ATA66 drive.
> 
> I got 24 MONO tracks, all using different 16bit 44.1 kHz files. The
> entire system (OS, apps and audio files) are on the same ATA66 harddisk.
> When adding a 25th audio file, I got an overload message. Very strange:
> the metering is not correct: the audio load showed 40% and the disk I/O
> showed 25% and still an overload when adding just one more file. I guess
> that's a bug in the disk I/O metering. So, 24 tracks is the maximum
> (adding a Fat EQ to every track got the CPU load to 50%, so there's
> still room for some DSP stuff) using only this one harddisk.
> 
> Another thing I've noticed is that the disk I/O metering sometimes
> doesn't work after an overload.
> 

Hi Joeri 

I think you kept the buffer length constant during the different file
playback 

what about doubling the buffer length , and than check the maxfiles ? 

suggestion : stay at 24 files : double buffer length : check cpu load 
Increase files till overload 

why ? with a bigger buffer , the cpu deal with bigger chunks , which
will increase efficiency 

I do not know how much time you can spare for this .. 

regards

frank