FIX: Running Motu FW devices in XP SP2, SP3

[TheRealRoach]

Bridgy,

It is possible that there are multiple copies of the SP2/SP3.cab on your system. They all need to be temporarily moved since XP is "smart" like that. Again, moving the "cab" files doesn't really affect the system since they don't contain active drivers. The active drivers are located in system32/drivers.

Majdid,

SP2 is just as flawed as SP3 unfortunately. But of course, every case is different and for your purposes it may work ok. It when you start to get into higher bandwidth requirements (i.e. daisy chained FW devices) that more problems occur.

[Bridgy]

Hi Roach,
Thanks for your continued interest. It is not the multiple copies of the cab files giving me a problem. There are multiple instances of the two 1394 files themselves all over the place. I guess I could take the approach that once I have secured the cab files I should just be able to delete all other instances of those two files and hope that does the trick.

On the issue of the "found new hardware":- When you ask windows to look in a specific place for drivers, doesn't it look for a .inf file when installing a driver? How can the install work if we are tring to install from a folder that contains only those two 1394 files and no .inf file? Surely, the install dialog will not let you proceed unless it sees a .inf file in the folder you specify?

Finally, wasn't SP3 supposed to include the hotfix that addresses the SP2-introduced firewire-speed-throttling? So am I right in saying that, with SP3, part 2 of your fix (the registry hack) should be redundant or am I missing something?

Thanks again for all your help, it is much appreciated.

Bridgy

[TheRealRoach]

Hi Roach,
Thanks for your continued interest. It is not the multiple copies of the cab files giving me a problem. There are multiple instances of the two 1394 files themselves all over the place. I guess I could take the approach that once I have secured the cab files I should just be able to delete all other instances of those two files and hope that does the trick.

That's the right idea. You just have to hide everything from Windows so that it is forced to ask YOU which files to install. You may want to delete ALL file names that coincide with the SP1 versions that I link to below: (and again, all of these files are located in your "cab"s so as long as your cab's exist somewhere you're fine... Worst comes to worst and they're still on your XP install discs. Let me know if deleting this group of files makes the difference and i can add it into the steps.

On the issue of the "found new hardware":- When you ask windows to look in a specific place for drivers, doesn't it look for a .inf file when installing a driver? How can the install work if we are tring to install from a folder that contains only those two 1394 files and no .inf file? Surely, the install dialog will not let you proceed unless it sees a .inf file in the folder you specify?

You may be missing a few files... I may have accidentally omitted them from my steps.. still a work in progress
Go to http://www.roachmouthrecords.com/firewire_sp1.zipand download. All the necessary files are in this file.

Finally, wasn't SP3 supposed to include the hotfix that addresses the SP2-introduced firewire-speed-throttling? So am I right in saying that, with SP3, part 2 of your fix (the registry hack) should be redundant or am I missing something?

It sure was and you can imagine how pissed i was when I got my new SP3 system and it exhibited the exact same problems as more than 5 years previous when I fixed the same problem by rolling back my SP2 update back to SP1. At the time I didn't know that it was the FW drivers specifically... Something is very much broken with the post SP1 FW drivers for Windows and how they interact with MOTU stuff.

Thanks again for all your help, it is much appreciated.

Bridgy

No prob. MOTU tech support is totally useless so hopefully through this trial and error I can continue to edit my "FIX" instructions so that it works for everyone. So far about 1 in 2 have luck with it it seems.

[Bridgy]

Roach

Well I finally managed to get the SP1 1394 drivers in and aaaaaaagggggghhhhhhh 10 times worse !!! The crackling was even happening on LOADING the patches !! What a dissaster.
I have gone back to the SP3 drivers now and it is back to how I was and at least now the patches load quietly even if the start to pop crack and squeal when I try to play them.

The patch the really exhibits the problem the most is the Nouveau Pard 1 and (much worse) Nouveau Pad 2. They are just unuseable !!

I have even tried and Firewire card with TI chipset, still the same problem. Looks like my problem could lie elsewhere. Absolutely no idea where to go now. Seems like I have tried everything I can think of.

Oh well, at least some of the patches are useable I guess.

Regards
Bridgy

[TheRealRoach]

Bridgy,

Well that is a huge surprise that SP1 actually made it worse. That pretty much exhausted all of my ideas unfortunately. Just as a last ditch attempt, could you list all of your system specs? Operating system and DAW software?

[Bridgy]

Hi Roach,

I have come to the conclusion the my machine is just not man enough to run the big Omnisphere patches and I will have to throw more money at it at some point.
My current spec is:-

ASUS P4PE Main Board
P4 2.8GHz CPU
2Gb PC2700 Ram
1 IDE CDRW
1 IDE DVDRW
2 * 120Gb 7200rpm IDE HDDs :-
Disc 0 Partition 1 - Windows, Partition 2 - Cubase 4.5.2 Application
Disc 1 Cubase Audio data
1 * Velociraptor 10000rpm SATA drive (configured as a single RAID 0 drive)
Omnisphere STEAM samples
1 TI based Firewire card
2 * UAD-1 DSP cards
Matrox PCI QUAD Graphics card + 4 * 17" flat panels

Might try and get hold of a bigger CPU (can only go as high as a P4 3.2GHz on this mobo) or go the whole hog and upgrade the main board which will give me better performance upgrade options as I have just about maxed out my current P4PE.

Anyway, I thing I have just about exhausted all the things I can try now... unless you know differen that is !

Many thanks for all your help
Regards
Bridgy

[TheRealRoach]

Bridgy,

Just out of curiosity, were you still experiencing the squeals once you had installed SP1 drivers? Or were you only experiencing buffer-related clicks and interruptions in playback?

With regards to your system, I just upgraded from a computer similarly spec'd to yours up to a quad 3ghz machine and the difference is night and day, especially when running samples like the Spectrasonics stuff.

One cautious word though, be prepared to potentially encounter the same driver/squeal/Windows/Motu weirdness even with the new system... It was the continued problem with my new system that spawned this whole "fix"!

Good luck anyway. If you make and breakthroughs post them back here.

[Bridgy]

Hi Roach,

I did a bit more testing yesterday and found a couple of strange things.

When I have a Cubase project containing one track with the Omnisphere plugin lodaded with its patch browser open, my cpu is sitting at around 65% (there is one recorded vocal track and one recorded guitar track as well) The patch that is loaded is Nouveau Pad 1 and Cubase is stopped i.e. I have not pressed play yet. If I then close the patch browser (leaving just the main Omnisphere plugin interface open) the cpu drops to around 29%!!!! If I then play just a three note chord on the MIDI keyboard the CPU max's out at 100% with click and pops all over the place.

I am hoping this info may give someone more experienced in these matters some kind of clue as to what could be the cause of my problem.

By the way, a friend of mine has a less powerful P4 than mine (2.6Ghz) but has no problems with this patch in Omnisphere or indeed with the somewhat larger Nouveau Pad 2 so I am convinced I should be able to sort mine out.

Please rack your brain and see if there is anything you can suggest that might be the cause.

Neil

[TheRealRoach]

Hi Bridgy,

This almost seems more like a glitch in on the Omnisphere/Spectrasonics side of things more than anything else. With regards to you original post a few weeks ago about pops and clicks, did you only experience these problems with Omnisphere? Did the system run fine as long as Omnisphere wasn't loaded?

I would start checking the Cubase.net forums, and search "Omnisphere" as there have already been quite a few threads about it.

The CPU spikes that you describe aren't really a symptom of the SP1 driver, et al., problems that are described in this thread.

From my experience with RMX, Atmosphere and Trilogy, they are all quite CPU friendly. I used to run RMX, a couple Trilogies and a few Atmosphere's on my old system which was a P4 2.8, and it dealt with that fine, and at a low buffer setting too.

Best of luck.

[FelisenumJaap]

Hi and first thanks!
I would like you to know that this also fixes things for my Tascam fw1804. I had no sound issues, but about half of the time the firewire connection wouldn´t work when booting. Switching it off and on sometimes cured it, but other times caused a blue screen.
This seems now completely cured, so it´s definitely a XP post SP1 issue.
Now let´s hope it will also cure the problems my friend Jaap has with his new ultralite MKIII. That´s why I was here anyway.....
Thanks again
Vincent
Edit, OK the above is true, but the device is practically unusable. I can´t change latency at all and once there is an asio overload, it looses sync and can´t recover. So SP1 drivers are no go for Tascam FW1804.

[portnoval]

First post! I just got an 896mkIII and have a small amount of clicks and pops. It may be that I'm running a VIA firewire chipset although I'd like to try this fix before I get a new card. I've previously done the registry hack fix but XP PRO was pre-installed with SP2 so I don't have the sp1.cab file. Could someone please post (or email me) the SP1 versions of 1394bus.sys and ohci1394.sys? My email is portnoval AT yahoo DOT ca. cheers

[portnoval]

Sorry for the noob post, I hadn't realized there was a second page to this thread before I posted... I installed the sp1 drivers along with a few system tweaks like uninstalling the built-in firewire and soundcard. I mixed in Sonar for about an hour and had no clicks or pops whatsoever.

A note about the driver install: When I tried manually deleting the .sys files they kept reappearing even though I had renamed all my driver.cab, sp2.cab and sp3.cab files (in multiple folders). So I just went in to the firewire driver details and said "update driver" and pointed to the new drivers. It gave me some warnings that the 1394bus.sys and ohci1394.sys files weren't digitally signed and warned that they were older versions than the current drivers. Regardless, it let me change them and when I went back to driver details it showed the proper sp1 versions. Is there any reason this approach can't be used by others? This would save the trouble of removing all the .cab files.

[TheRealRoach]

Hi portnoval,

Yup the "Update Driver" method you mentioned works as well. There have been varied reports about either one of them working, and not working as well. For me, the Update Driver method didn't work for some reason... But thanks for posting the alternate method.

[casedawg]

Tried this as a last ditch effort to fix audio dropouts and high pitched noises on my Ultralite mk3. No luck. I had to use a slightly different method than listed in the first post because even after uninstalling the 1394 controller in device manager, renaming the cab files and RARing them up, and deleting all the 1394bus.sys and ohci1394.sys files, the .sys files kept recreating themselves in Windows\system32\drivers. They would also automatically revert back to the newer version when I replaced them with the older version. Anyway, I ended up using the device manager update driver function along with the INI posted in this thread to get the SP1 drivers installed without them reverting back to the newer versions.

[TheRealRoach]

Hi casedawg,

Could you list your full system specs, software, and relevant driver specs?

Are you sure you followed the steps properly? including the checks?

EDIT: I just found your specs in another post. I don't run any Vista machines so any guarantees of this fix working are pretty much out the window..... As mentioned before though, the audio interruptions that you experienced are caused by a bandwidth limitation as caused by system drivers.