hello, Neutron.
yes, i believe you do need a driver to run ACHI which is installed when windows is installed(i could be wrong). and no, i don't think that it's really a big advantage when IDE mode already handles plenty of data with less trouble.....
Masterverb Test Thread
Neutron, yes you need to install intel SATA driver in Windows before switching in bios (you may install intel matrix storage manager app which auto installs sata driver or install only the driver on the floppy or download from intel site).
and yes, without AHCI no NCQ nor hot plug. (...I suppose other low level SATA protocol optimizations should be missing)
cheers
Fede
and yes, without AHCI no NCQ nor hot plug. (...I suppose other low level SATA protocol optimizations should be missing)
cheers
Fede
-
luigicampala
- Posts: 112
- Joined: Sun Dec 16, 2007 3:40 am
- Location: Germany
- Contact:
thats very good. do you have zillions of USB controllers showing up when you boot, using lots of IRQs?
i can get 12 on mine so far. same board, pulsar 2 + 2 scope pro
everyone else, I think ill have a go with the drivers later. maybe when i change drives after Vlociraptor is avaliable.
I have a lot of other chores in the studio, and scope PC seems to be pretty good atm.
i can get 12 on mine so far. same board, pulsar 2 + 2 scope pro
everyone else, I think ill have a go with the drivers later. maybe when i change drives after Vlociraptor is avaliable.
I have a lot of other chores in the studio, and scope PC seems to be pretty good atm.-
luigicampala
- Posts: 112
- Joined: Sun Dec 16, 2007 3:40 am
- Location: Germany
- Contact:
i had zillions, but i had to disable usb as i had that crackling problem (i don´t know the exact reason for it by now, i think either the internal soundcard or the usb-keyboard/mouse). recently i enabled usb again but disabled the usb port sharing the irq with my pulsar card. currently there is only a printer connected via usb. no crackling so far, by the way... 
-
luigicampala
- Posts: 112
- Joined: Sun Dec 16, 2007 3:40 am
- Location: Germany
- Contact:
a another thing: what bios are you using? there used to be troubles with the performance
you can read, what i found out in this topic
http://www.planetz.com/forums/viewtopic ... highlight=
it started which version f05.
more details can be found here
http://www.tomshardware.com/forum/24721 ... oblem-ds3r
you can read, what i found out in this topic
http://www.planetz.com/forums/viewtopic ... highlight=
it started which version f05.
more details can be found here
http://www.tomshardware.com/forum/24721 ... oblem-ds3r
That latency tester is a great weapon to add to your test arsenal!
I just tried it on my workstation at work. its a dual dualcore xeon. the results were all below 100us. now if dell can do it surely anyone can.
I cant revert to a lower BIOS because i have wolfdale CPU
I just tried it on my workstation at work. its a dual dualcore xeon. the results were all below 100us. now if dell can do it surely anyone can.
I cant revert to a lower BIOS because i have wolfdale CPU
Last edited by Neutron on Thu Apr 24, 2008 8:04 am, edited 2 times in total.
-
luigicampala
- Posts: 112
- Joined: Sun Dec 16, 2007 3:40 am
- Location: Germany
- Contact:
when i installed virus TI on a different PC it offered to turn off "selective suspend" which I never even heard of. they said it adversley effects PC and usb performance.
anyways here is how to turn it off yourself
HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\USB
Entry: DisableSelectiveSuspend
Type: DWORD
Value: 1 disables selective suspend. 0 enables selective suspend
Note If this registry key is not present, you must create it.
This setting affects all USB host controller drivers in the system. If the DisableSelectiveSuspend value is set to 1, the selective suspend feature is turned off and the Allow the computer to turn off this device to save power check box does not appear on the Power Management tab for the USB Root Hub
See if that helps the Spikes.
anyways here is how to turn it off yourself
HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\USB
Entry: DisableSelectiveSuspend
Type: DWORD
Value: 1 disables selective suspend. 0 enables selective suspend
Note If this registry key is not present, you must create it.
This setting affects all USB host controller drivers in the system. If the DisableSelectiveSuspend value is set to 1, the selective suspend feature is turned off and the Allow the computer to turn off this device to save power check box does not appear on the Power Management tab for the USB Root Hub
See if that helps the Spikes.
Are you sure it's correct? I can't find a USB directory.....the only "usb" related ones are "usbheci", "usbhub" "USBSTOR" and "usbhuci".Neutron wrote: HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\USB
.
?
Ha...maybe that's already a key? ok done
I have some spikes with the samplers in my new core 2 duo machine....I have also my trustworthy PIII machine with a P2 and a Luna who doesn't have this problem and it's pipelined to the main Scope PC
I tried this setting, it seems to have gotten rid of "crash after coming back to scope after a while"
I checked my computer with the latency tester. it sits at 2-3us with a spike to 3-500 every 4 or 5 intervals and up to over 1000 if i open a program such as ableton (scope didnt do much) you can also bring it up to about 200 by grabing a window and dragging it all about.
I got a BSOD, but im not sure why i forgot to turn off auto reboot. i think its all those USB controllers. I have ordered a PCI-E usb controller in hopes that i can disable the onboard ones, and the PCI-E one will use different IRQs.
I am rather baffled why Gigabyte has made the ICH9R chip use the same IRQs for the USB and PCI, while Supermicro had them completely isolated. (it was ICH9)
and why you can not at least turn off some of the usb controllers instead of all or nothing.
There arent any other boards i can find with 3 adjacent PCI slots and support wolfdale.
If scope crashed in supermicro it would never BSOD, just application error. but if it crashes in Gigabyte. its reboot time. (so far)
I checked my computer with the latency tester. it sits at 2-3us with a spike to 3-500 every 4 or 5 intervals and up to over 1000 if i open a program such as ableton (scope didnt do much) you can also bring it up to about 200 by grabing a window and dragging it all about.
I got a BSOD, but im not sure why i forgot to turn off auto reboot. i think its all those USB controllers. I have ordered a PCI-E usb controller in hopes that i can disable the onboard ones, and the PCI-E one will use different IRQs.
I am rather baffled why Gigabyte has made the ICH9R chip use the same IRQs for the USB and PCI, while Supermicro had them completely isolated. (it was ICH9)
and why you can not at least turn off some of the usb controllers instead of all or nothing.
There arent any other boards i can find with 3 adjacent PCI slots and support wolfdale.
If scope crashed in supermicro it would never BSOD, just application error. but if it crashes in Gigabyte. its reboot time. (so far)
To make CPU seen as single processor from XP:
Run---msconfig---BOOT.INI tab---advanced options, tick /NUMPROC and select one. Reboot. That is reversible in any moment, worth a try.
It solved my STS 5000 problems. If you don't have an heavy use of vst's you won't even notice the CPU power loss. it should be around the 25% less power. I don't use vst's at all, so I'm quite fine.
Despite what's a common route today, I have the sensation that the current Scope software is not optimally working with the core2duo and multicore systems. Another impression that I have is that the PCI bus architecture on these new PCIe mobos is somehow less good and quite secondary in the resources assignement of the mobos. An old 815e chipset- pentium3 system I have wins hands down the competition as for PCI bandwidth. If it didn't have the 512 Mb limit for the ram it would still be my main machine.
Run---msconfig---BOOT.INI tab---advanced options, tick /NUMPROC and select one. Reboot. That is reversible in any moment, worth a try.
It solved my STS 5000 problems. If you don't have an heavy use of vst's you won't even notice the CPU power loss. it should be around the 25% less power. I don't use vst's at all, so I'm quite fine.
Despite what's a common route today, I have the sensation that the current Scope software is not optimally working with the core2duo and multicore systems. Another impression that I have is that the PCI bus architecture on these new PCIe mobos is somehow less good and quite secondary in the resources assignement of the mobos. An old 815e chipset- pentium3 system I have wins hands down the competition as for PCI bandwidth. If it didn't have the 512 Mb limit for the ram it would still be my main machine.
DFI Infinity Dark P965-S with Q6600
2Gb Corsair 6400C4DHX Twinx
Radeon HD3850
Win XP Home
Pulsar 1 + Pulsar 2 cards (10 DSP totals).
After changing PCI slot to the Pulsar 2 (so it's not sharing anymore IRQ with LAN and Video Card too), and putting Pulsar 1 as main board
in the cset.ini file I get:
9 Masterverb PRO (reaching DSP limit loading a 10th one)
8 Masterverb (reaching PCI capacity limits loading the 9th one)
but if I hit "cancel" I am able to add 4 more Masterverb (so 12 in total)
Having Pulsar2 as main card I was getting PCI capacity limit after the 2nd Masterverb PRO and at 8th Masterverb.
Strange?
2Gb Corsair 6400C4DHX Twinx
Radeon HD3850
Win XP Home
Pulsar 1 + Pulsar 2 cards (10 DSP totals).
After changing PCI slot to the Pulsar 2 (so it's not sharing anymore IRQ with LAN and Video Card too), and putting Pulsar 1 as main board
in the cset.ini file I get:
9 Masterverb PRO (reaching DSP limit loading a 10th one)
8 Masterverb (reaching PCI capacity limits loading the 9th one)
but if I hit "cancel" I am able to add 4 more Masterverb (so 12 in total)
Having Pulsar2 as main card I was getting PCI capacity limit after the 2nd Masterverb PRO and at 8th Masterverb.
Strange?
you should leave out the MV Pro in this test - it's built differently and not comparable at all.
the 'regular' MasterVerb is only choosen as a 'unit' that almost every Scope user has, not for it's audio capabilities.
if you have a very good MV performance, then It may be interesting (as a sidenote) how many instances of the STW A100 are possible (in case you have access to it), as that's probably the most PCI-demanding device in Scope.
cheers, Tom
the 'regular' MasterVerb is only choosen as a 'unit' that almost every Scope user has, not for it's audio capabilities.
if you have a very good MV performance, then It may be interesting (as a sidenote) how many instances of the STW A100 are possible (in case you have access to it), as that's probably the most PCI-demanding device in Scope.
cheers, Tom
Ok, good to know, thanks.
With old DAW (Asus P4P800 deluxe and XP in PC standard mode and NOT ACPI mode I reached 10 MV without any error pop-up).
Anyway in some of my songs I used more than one MV Pro, so these settings seems to be better for me to work again with those kind of Scope projects on the new DAW.
With old DAW (Asus P4P800 deluxe and XP in PC standard mode and NOT ACPI mode I reached 10 MV without any error pop-up).
Anyway in some of my songs I used more than one MV Pro, so these settings seems to be better for me to work again with those kind of Scope projects on the new DAW.
-
Music Manic
- Posts: 1743
- Joined: Wed May 15, 2002 4:00 pm
- Contact:
Re: Masterverb Test Thread
bump this..please test new boards