Scope PCI 64bit testing

[the19thbear]

Please post any cons/pros/thoughts about the new 64bit pci update - what has been fixed, what needs to be fixed, what has changed etc etc

thanks alot!

[delron22]

What update - can't see anything on SC page?

[PaulSh]

It's a Beta version apparently - there's a (somewhat derailed) thread in the Announcements section.

[delron22]

Cheers,

I did pick that up and have downloaded it now. Is it worth taking the plunge with the Beta or should I wait before i rebuild with 64 bit?


Del

[PaulSh]

I don't have Scope 5.0, which means I can't try out the 5.1 Beta and so I can't really offer an opinion. I'm really looking forward to the release though because I use VSL and I keep on maxing out my memory in XP. I'm still on 4.x BTW because the only reason I would need to upgrade would be for 64-bit support, everything else is fine for my needs.

EDIT...

I now have the 5.1 beta working well on my Pulsar 2 card under Win7-64 with Sonar 8.5.3-64, although it's fair to say that I've not given it anything very complicated to do so far. The only niggle I have is about WASAPI, so I've posted over in the 5.1 thread in the problem solving section.

EDIT 2...

Given up on WASAPI and using the new fixed ASIO drivers.

[MMC]

I thought, the Scope-Software will also be rewritten...? When I look in the Systems->Programs-Folder, the SW is 32 Bit (x86) and with the Task Monitor you can see the Scope is running as *32 (bit).
I think they have only changed the Drivers...
Will be a 64-bit-Version of the Scope SW released?

[PaulSh]

As far as I understand it, the only reason why it would matter whether the Scope software is running in x86 or x64 mode is if you wanted to use the Scope STS samplers with multiple gigabytes of samples. The Scope system itself is running native on the DSP's, so with the exception of things like the samplers that make extensive use of the PC's memory, all that the x86 part of the software is doing is managing the interface to the DSP system.

[MMC]

This is not the right Answer to my Question!
Are they programming a 64bit-Scope Software or have they only changed the Drivers.
Then I ask myself, why does the Software have a new version number and why is this a full Package?

[astroman]

it IS the right answer, but you put the wrong question...
and you have no idea what the software is about
otherwise you wouldn't have asked at all
in fact it looks like you don't even know what 64 bit stands for...
so why bother yourself (and the rest of the world) ?

the extended adressing has not a single advantage for what the SFP system does.
NOT EVEN ONE
nobody would 'rewrite' anything in this context - the stuff would just be recompiled.
Which they (seem to) avoid THAT as much as possible because compilers HAVE bugs. All of them.
Unnecessary recompile of such a complex system has a significant potential of introducing additional errors.

Btw you're complaining about 'rewriting' a > 200 man years (developement time) piece of software. LOL

cheers, Tom

[siriusbliss]

I thought, the Scope-Software will also be rewritten...? When I look in the Systems->Programs-Folder, the SW is 32 Bit (x86) and with the Task Monitor you can see the Scope is running as *32 (bit).
I think they have only changed the Drivers...
Will be a 64-bit-Version of the Scope SW released?

64-bit drivers first, then the full 64-bit version is coming.

Greg

[siriusbliss]

This is not the right Answer to my Question!
Are they programming a 64bit-Scope Software or have they only changed the Drivers.
Then I ask myself, why does the Software have a new version number and why is this a full Package?

Because they've also done some updates to the devices and DSP handling - especially for Xite.

Greg

[johnnycage]

I just submitted a bug to bugs@soniccore.de for the x64 SONIC_CORE_SCOPE_PCI_v5.1.2707-BETA_EN.EXE / Latest ftp ASIO/MME drivers.

I am experiencing crackling/glitching when using the ASIO2/ASIO1 drivers, even with ULLI set to maximum, Scope card is master clock, and the Scope PCI card being the only thing on its IRQ (no conflicts). Tried swapping PCI slots as well just to do it, to no avail.

Glitching doesn't happen with the MME driver.

I have an i7 975X @ 3.33Ghz / 12GB 1600Mhz ram, ASUS P6T mobo with latest BIOS, running Win7 ULT x64.

I thought I saw something like it in an older thread, but that person seemed to have had an IRQ conflict.

Anyone else having a similar problem with the latest ASIO/MME beta drivers?

[garyb]

it's possible you don't have the lastest version...
i believe there was a new driver a couple of days ago, which other users claim has fixed crackles on their systems.

[johnnycage]

Nah, I grabbed the latest drivers this morning, along with the latest scope build (v5.1.2707-BETA_EN) as soon as they were uploaded. (9/10/2010 5:20AM CST [GMT -6:00]).

[tjsound]

I have ASUS P6T and XITE. All worked fine, no glitches, cracles, etc. ULLI - 3ms Win7 x64, Cubase, Reaper, Ableton. (reaperx64 - the best)
Did you tune BIOS for audio work?

[astroman]

I thought, the Scope-Software will also be rewritten...? When I look in the Systems->Programs-Folder, the SW is 32 Bit (x86) and with the Task Monitor you can see the Scope is running as *32 (bit).
I think they have only changed the Drivers...
Will be a 64-bit-Version of the Scope SW released?

64-bit drivers first, then the full 64-bit version is coming

Greg, there's still no need to rewrite anything for 64 bit adressing because the Scope application simply doesn't need such memory space.
The only device that could use it is a sampler (as mentioned) and that does it's access via disk and mobo memory in native coding, which could be handled by a vxd like device or something similiar.
The IO part of the system will certainly be adapted to the new adressing scheme for communication with devices and 'foreign' applications, but that's about it imho.

cheers, Tom
( do people really consider 64 bit better because it's a bigger number ?)

[ARCADIOS]

ram

[siriusbliss]

I thought, the Scope-Software will also be rewritten...? When I look in the Systems->Programs-Folder, the SW is 32 Bit (x86) and with the Task Monitor you can see the Scope is running as *32 (bit).
I think they have only changed the Drivers...
Will be a 64-bit-Version of the Scope SW released?

64-bit drivers first, then the full 64-bit version is coming

Greg, there's still no need to rewrite anything for 64 bit adressing because the Scope application simply doesn't need such memory space.
The only device that could use it is a sampler (as mentioned) and that does it's access via disk and mobo memory in native coding, which could be handled by a vxd like device or something similiar.
The IO part of the system will certainly be adapted to the new adressing scheme for communication with devices and 'foreign' applications, but that's about it imho.

cheers, Tom
( do people really consider 64 bit better because it's a bigger number ?)

You don't have to tell me - tell SC.

I've already mentioned many times elsewhere that it's not a real necessity for Scope to go 64-bit (other than drivers).

But if SC needs to gain advantage for VDAT and STS, (or maybe more onboard RAM handling on Xite) then so be it.

Greg

[johnnycage]

I have ASUS P6T and XITE. All worked fine, no glitches, cracles, etc. ULLI - 3ms Win7 x64, Cubase, Reaper, Ableton. (reaperx64 - the best)
Did you tune BIOS for audio work?

Yes, disabled all unnecessary things (regular ide controller, onboard sound, onboard 1394, even un-used usb ports in Device Manager, etc.), and there are not any irq conflicts for my scope card. I'm pretty experienced with having to tweak Windows/PCs to make stuff act right.

I have Scope pro 15dsp PCI card btw, not XITE. (I wish ) I'm sure that you've got much more bandwidth with that XITE PCI-e rig.

I tried without my a-16 hooked up via z-link as well. I am going to try other mixers and perhaps the straight-24 bit ASIO modules in scope, just to see if that helps. Maybe raising the priority of scope.exe or the old trick for setting windows for background services priority might help too...

I just don't see it being a horsepower issue with my rig. The Scope card is the only other card in my PC besides my Video card, which is PCIe.

I will keep tweaking stuff I guess, and maybe wait for the next driver/scope update.

Thanks for the replies and suggestions.

[PaulSh]

ram

So, which part of the Scope system do you think needs multiple gigabytes of RAM? Certainly not the DSP code, you'll run out of DSP resources or PCI bandwidth long before you run out of code memory. DSP code isn't bloatware like current x86/x64 code, it's very small and tight - even the largest Scope devices are typically barely more than a megabyte. With the single exception of the Scope STS samplers, Scope should run quite happily in under 100MB. If you're an extensive user of the samplers then I might agree with you about the need to be able to use more RAM, but otherwise, no.