FIX: Running Motu FW devices in XP SP2, SP3

[derZorn]

anyone have a link to the SP1 drivers? Both my XP machines came with sp2. And I have not yet been able to find them on the net.

And why is there no fix from MOTU for this. SP2 has only been out since '03?

Here are the SP1 files I used - I got them somewhere else on these boards - but they didn't work for me.
Hope you have more luck.

http://rapidshare.de/files/41106646/fir ... 1.zip.html

[TheRealRoach]

Hi,
after a number of problems with getting my 896hd to work (which actually was a defective FW port in the 896hd) I'm finally running a MOTU/WinXP system.
I'm really happy with it..... Mostly.
From time to time (seemingly random) I get a number of squeals and dropouts and was hoping to find the answer to my problem in the answers found in this thread.
Well, so far I'm out of luck - the SP1 drivers won't stay installed on my sys for example, but one thing that is obvious to me is that it's
not the unit that has problems, but the system.

Be sure to check your steps very carefully. I've updated Post #1, step [1.5] to include a link to the sp1 files. Inside you will find abou 6 files related to fw drivers. Do a search of your whole system for these files remove (DON'T delete) them by placing them on a CD or usb memory stick. Also be sure that all of your SP1.cab, SP2.cab and SP3.cab files are also removed onto a temporary storage device not connected to the system. If the system will immediately replace these files in one shot (even without a reboot it would seem). The SP2/SP3 FW drivers are most definitely a contributing cause to your symptoms. You are running a very similar system to me so it will work out in the end.

When I connect an USB-Stick to my onboard USB ports I can provoke the high pitched squeal every time.
I tried that with a number of different sticks and ports on my ASUS M2N-E with Nvidia 570Ultra chipset.
The squeal is there for like a second while the device gets detected, then stops.
Sometimes audio sounds somewhat distorted after that.

So maybe our problems don't lie with the FW but somewhere in the USB hub ?
(

Funny you mention this because I've recently noticed the same phenomenon. Generally it isn't good practice to connect/disconnect drives while there are any programs running. I do believe that this squeal/momentary interruption is attributed to the system's plug-and-play scanning of the device as it is connected. Naturally as the internal data bussing of the system is re-oriented all other data streams would experience a momentary lag. I don't think this is a bug... however it would obviously be something to improve on the Windows platform in the future.

anyone have a link to the SP1 drivers? Both my XP machines came with sp2. And I have not yet been able to find them on the net.

And why is there no fix from MOTU for this. SP2 has only been out since '03?

I've just updated the original post to include a link to the SP1 files that I used. It's in step [1.5]. And yes, this FW bus issue is the same problem that has been around since SP2 was launched. In all fairness there is nothing technically wrong with the MOTU drivers, and there's not much that MOTU could really do to incorporate a fix into their drivers since the fix would require registry and backdating of Windows XP drivers... Windows XP FINALLY needs to fix this FW bus speed issue. There are "hotfixes" available, but they only perform one of the two major steps cited in my first post.... thereby having no affect.

What MOTU SHOULD have done and STILL SHOULD do, is post up a formal fix on motu.com similar to what I have posted here (which happens to be partially based on RME's fix that they posted on their site 4 years ago. I've even sent this fix to them to review, but they got back to me a couple weeks later with a "we think you should contact tech support with your problem." Before I came up with this fix for my system I was looking to sell off my MOTU's and go for something else, but nothing comes close at the same price-point.

For a long time this issue has been swept under the carpet because:
- the majority of consumers who use firewire use it for short bursts of data transfer (i.e. backup, extra storage) and wouldn't notice the affects (although they would technically be measureable
- the consumers who did notice the SP2 truncation of FW bus speed were those who used digital video cameras. They require a constant higher bandwidth data stream, so people had a few issues, and some of those were fixed with hotfixes.
- there aren't many people who use Windows XP machines with FW devices that require a very high bandwidth for continuous data streams. But we do. The majority of people who use FW devices for pro applications usually run MACs (where FW was born in the first place). You could own a Lamborghini, Ferrari, and Bugatti, but if you only every drive them at highway speeds then you'll never notice problems that occur as they are each respectively pushed hard.

[Mark S]

Edit: Real Roach answered all my questions. Thanks!!!. I won't have a chance to do this until next week but I'll report my results back, ( and on the Sonar forum too where this is a subject of conversation)

A comment: on top of all this, if you are using pre mkIII hardware you should be using the 3.6.7.4 drivers since the later drivers increase latency due to hardware calls used in the mkIII units, Be warned there too.

[derZorn]

My system been's finally running a whole day without any serious squeals or droputs !
I had to go a number of roads though, and I'm not yet convinced I'm through this for all time.
Means, I'm using my 896hd unit for simple mixing and not in the field recordings for now - I won't yet risk screwing a session for technical probs.

Anyway, what i did to do on top of installing the SP1 drivers as described in here was
- update my mainboards bios (asus m2n-e with nforce 570ultra chipset)
- disable onboard sound on bios-level
- move my FW card to a different PCI slot
- make sure the FW is on a IRQ on it's own and not sharing it with another component
- rollback drivers to 3.6.7.0 (I had problems with WinXP not overwriting newer files here, so the audio part of it is still 3.6.8.1380)
- increase buffer levels to 256 - better 512 ( I really don't like that part - bet it's part of above driver prob)
- I even integrated the sp1 drivers in the sp2.cab and drivers.cab to make sure they NEVER EVER come back

I'm not sure which of these finally made the problem go away (for now), so I listed all I did in the end.
Anybody with similar persistent problems should try above, too.

I'd like to backdate the audio driver to 3.6.7.4 / 3.6.7.0, but I'm not sure which files remain in the system that need to be
deleted manually after deinstalling the 3.6.8.1360/1380 drivers.
Anybody got a list of files that remain in the system after deinstalling the drivers ?

[TheRealRoach]

Anyway, what i did to do on top of installing the SP1 drivers as described in here was
- update my mainboards bios (asus m2n-e with nforce 570ultra chipset)
- disable onboard sound on bios-level
- move my FW card to a different PCI slot
- make sure the FW is on a IRQ on it's own and not sharing it with another component
- rollback drivers to 3.6.7.0 (I had problems with WinXP not overwriting newer files here, so the audio part of it is still 3.6.8.1380)
- increase buffer levels to 256 - better 512 ( I really don't like that part - bet it's part of above driver prob)
- I even integrated the sp1 drivers in the sp2.cab and drivers.cab to make sure they NEVER EVER come back

All though many of these things are suggested changes to improve performance, the SP1 drivers without a doubt are the main reason. Years ago (when Sp2 first dropped), I spend hundreds of dollars in hardware upgrades and time to make my dual 896s work together. Once I rolled back to SP1 everything worked great. Even with my crappy single processor laptop with very little ram and tons of background applications. The MOTU stuff is surprisingly sturdy and stable when the proper XP FW drivers are being used.

Edit... You should be able to rock much lower buffer setting. My system runs at 64 sample buffer, and if it's struggling, 96. Even my old P4 2.4 system ran at 128 pretty comfortably.

[reeper]

yeah.. well I have no problems at 64buffers with Asus p5q-e and xp64 sp2... cmon just install the proper hardware and you will be free of these workarounds.

[TheRealRoach]

yeah.. well I have no problems at 64buffers with Asus p5q-e and xp64 sp2... cmon just install the proper hardware and you will be free of these workarounds.

Care to provide some details or a link?

[Mark S]

Update trying to link 896hd and 8pre:

Driver 3.6.7.4
MSI Neo2 platinum with 3800 dual operton
xp pro
Sonar 7.2 PE
2.7ms latency @ 256

SP2 drivers would not allow syncing. Was told that the only way it would ever link was via adat which would effectively kill 4 channels in this setup.
Once I successfully got it all set up with sp1 drivers I was able to get it synced. It's been just a bit since this success so I'll work it out and see how it works in the long run. At this point I'll see how this works. I still get a hard bsod when i turn off the slaved system but I guess I'll quit doing that. The host controller is a VIA controller, so I'm working with a less than recommended hardware item. I'll probably try a pci card with a TI chip and see if there are any gains with that since my pentium M laptop also has the same situation to deal with, and a via controller too.

A couple suggestions to the process you've laid out:

(maybe I missed it) uninstall motu drivers as the first step.

I had to isolate the driver.cab with the sp2.cab because the drivers mysteriously reappeared when I deleted them.

for me, moving the sp2 cab back to it's location was not optional, as the instructions suggested, since there were drivers I needed in there to install the Motu

BOLD the shift/del in the instructions for getting rid of the old files. Took me a few tries before I gave myself the dope slap.

[TheRealRoach]

Hey Mark, thanks for the feedback. Just out of curiosity, are you measuring round trip latency (input AND output latency?) when you mention your 2.7ms at 256 sample buffer? Thanks for the suggestions on improving clarity, I'll be sure to make some edits when I get a chance.

The TI versus VIA chip set thing seems to be a bit of a red herring as far as this thread is concerned. Once the Windows FW drivers are sorted out I can break all sorts of MOTU's rules and suggestions and things still run great. I can run my two 896s off my 4-pin built in VIA FW port on my single processor M 1.73ghz laptop without a problem!, and it's a hunk of crap. I've never noticed any performance differences between VIA and TI over the past few years.

[llamaza]

Hi all

First of all, the greatest post.

I am a proud owner of a second hand Motu Ultralite MKII since 6 months. Since i installed 3.6.7.0 drivers and everything seems to be reasonable good (not able to make lower the latency (512 buffer size was the minimum) and the record and playback volume was lower than it was supposed so I left this post aside. But this week i have tried 3.6.8.1380. It has not been a good idea because I begun to experience some high pitched noise whenever the system starts the driver ( i think so, but i am not sure). The noises were present at the hardware metering in all channels during 2 or 3 seconds. I was terrified because nothing of it was happening before upgrading the motu soft to the new cuemix fx and the new drivers 3.6.8.1380.

So I have started the fix today at 11pm and I have already finished the whole process.

Need to mention

- It was hard to downgrade the firewire drivers from sp2 to sp1. In my setup, I have made the upgrade drivers options directly form Device manager, using the files mentioned above.

- It was equal to me.

Rollback drivers to 3.6.7.0 (I had problems with WinXP not overwriting newer files here, so the audio part of it is still 3.6.8.1380)

- Registry edit was already done.

Everything sounds correctly again, and the regedit improvment is definitely worth of it. Last week I was able to finish a project (20 tracks, lot of glaceverbs (free vst reverb)) and it overloads my computer sometimes using 2048 buffer side
Now I am able to playback it with 256 buffer side at the same time browsing motunation.com via firefox and neither dropouts nor noises are heared.

Thank you very much.

I´ve got a question
I have installed lastest drivers from Motu.com and it says 3.6.8.1380, but it appears 3.6.8.1359 in the Motu Audio Console. Why? Do you think it depends on the MKII or MKIII hardware?

My setup
Dell Inspiron 6000 (Centrino 1.73, 2Gb ram, PCMCIA Conceptronic (Chipset Via), Ati X300)
Ableton Live
Motu Ultralite (3.6.8.1380)

[TheRealRoach]

Hi llamaza,

I'm happy the post improved things for you.

Driver 3.6.8.* really screwed things up for me, but "reeper" has posted a mod for the driver that you may want to look into. I haven't investigated it yet as I've been a little busy. You'll notice that i'm still using 3.6.7.0 and it's been working quite well.

The unusual driver name "3.6.8.1359" has been noted by others and is likely just a "typo". Take care, happy holidays

[Mark S]

something that was dicovered by a DAW builder on the Sonar forum is that the new drivers (3.6.8.x) have a hardware call to the dsp chips in the new Mk III units that increases latency (doubles) in the previous units. He recommends the 3.6.7.4 driver which is the one before the new units. I was able to find a link to it in another thread here that is linked to MOTU but it does not contain the ASIO drivers, which is unfortunate since they work better for me.

Realroach: the 2.7 ms latency is simply what was reported by Sonat 7 PE. I do not know if that is round trip, had assumed it was. I'll check more. That is the WDM latency since, as I mentioned above, the download from MOTU did not have the ASIO drivers in them.

Anyone have a link to the whole package? It should be around 24 meg,...

[mr1]

Running XP SP2 with an 896HD and had huuuge problems with clicks/pops/squeals, even after a clean and stripped down install of Windows.
After various BIOS upgrades and Windows tweaks, it seems the thing has made the most difference has been rolling back to the SP1 firewire drivers and rolling back the Motu drivers to 3.6.7.0. Can now get through whole sessions at 'modest' buffer settings with no issues. So +1 to rolling back those drivers!

On the other hand, no luck with Vista (both in my workstation and my laptop). That works best with the latest Motu drivers and its 'ok' when running in ASIO with a high buffer setting, but just playing music through WDM is impossible - sounds like a badly tuned FM radio.

[llamaza]

Yesterday, after 2 hours of Live project work, everything goes to a horrible snow/hiss noise in the main outputs of my Motu Ultralite. I´ve just downgraded the drivers to 3.6.7.0. Everything seems to be correcto today.
Hope it solves my problem.

[studioguy457]

Yesterday, after 2 hours of Live project work, everything goes to a horrible snow/hiss noise in the main outputs of my Motu Ultralite. I´ve just downgraded the drivers to 3.6.7.0. Everything seems to be correcto today.
Hope it solves my problem.

Where do I find those drivers? I only have the new ones.