[paranoia] New mobo == slower cdparanoia

[system at localhost.localdomain]

On Tue, 3 Feb 2004, Bill Davidsen wrote:

> On Mon, 2 Feb 2004, Ross Vandegrift wrote:
> 
> > On Mon, Feb 02, 2004 at 08:35:58PM -0500, Ross Vandegrift wrote:
> > > All CDs rip with lots of -/+, but now I'm even seeing ! and e.  I think
> > > I may try making the drives slaves on the VIA channels and seeing if
> > > things improve any.
> > 
> > Here's one for the archives, hope someone else can find it when they
> > need it...
> > 
> > The promise interfaces are just awful for CD ripping.  I moved my CD-ROM
> > onto a motherboard channel as a slave and got fast, solid ripping.  The
> > chipset is a VIA KT400, the promise card in question is an Ultra33.  The
> > CD-ROM doesn't even support UDMA33.
> > 
> > It sucks I can't have an extra channel just for ripping - it does hurt
> > disk performance this way.  But proper rips are totally worth it.
> 
> Actually, I seem to remember that there is a problem with Promise
> controllers and VIA chipsets. I can't remember the details, and it may be
> something else, but a check for the archives of this list and the
> linux-kernel list might be instructive.

All VIA southbridges have a poor implementation of the PCI bus. The PCI 
controller only allows any given device to have control of the bus for a 
very short amount of time. This problem was made infamous because the 
SoundBlaster Live! cards had a latency issue that confllicted with the VIA 
PCI implementation. Anyway, I never progressed to playing with the BIOS 
registers that control the PCI bus timing settings, but there was a man 
who wrote patches for windows to adjust these settings. Actually changing 
the timings should be easy under Linux, but knowing what to set them to 
may be a trick. In any case, if you lengthen the time a PCI device can 
have control of the bus, the Promise controller should work much better 
for streaming. For more info on this, check out 
http://www.tecchannel.de/hardware/817/1.html

TJ Harrell 

--- >8 ----
List archives:  http://www.xiph.org/archives/
Paranoia homepage: http://www.xiph.org/paranoia/
To unsubscribe from this list, send a message to 'paranoia-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.