new driver version

[dare]

I'm convinced that the shifts in ouputs and problems that many are experiencing are mainly due to variations in Intel chipsets and the drivers written for them. I also applaud Microsoft's efforts to require certification of all drivers. Printed circuit board layout, capacitor size and placement on the buss are other hardware variations that can cause timing problems (timing is everything). Without strict standards there is no way to account for or eliminate these problems. Neither MS nor MOTU can write drivers to accommodate all the variations.

Messed outputs are clearly driver issue as they stated in the tech response they sent to me. BTW - if any hardware does not work as expected (reacts differently to commands it recieves), then it is simply - broken. Other thing is if it recieves correct commands and calls from the CPU. If MOTU for example changes it's PCB layout then MOTU should accommodate this through firmware update. There is a real mess with the standards, but some of them still exist and need to be followed. And one other thing: Capacitors are fully described with their capacity, voltage and some other things like leaking current... Physical size doesnt have anything with their functionallity. That is an exact science, it is not abstract as you may think at all..

[yeti]

Messed outputs are clearly driver issue as they stated in the tech response they sent to me. BTW - if any hardware does not work as expected (reacts differently to commands it recieves), then it is simply - broken. Other thing is if it recieves correct commands and calls from the CPU. If MOTU for example changes it's PCB layout then MOTU should accommodate this through firmware update. There is a real mess with the standards, but some of them still exist and need to be followed. And one other thing: Capacitors are fully described with their capacity, voltage and some other things like leaking current... Physical size doesnt have anything with their functionallity. That is an exact science, it is not abstract as you may think at all..

I understand.... but beside all this knowledge... will motu be willing or be able to fix that?
All that I can see is, that their lack of support and driver update policity is not aceptabel.
For me in this case... if a product does not work as described, it has to be returned and money back!
I´m really angry about this endless betatesting and trying hours after hours and loosing time for my clients and my daily work.

[dare]

I understand.... but beside all this knowledge... will motu be willing or be able to fix that?
All that I can see is, that their lack of support and driver update policity is not aceptabel.
For me in this case... if a product does not work as described, it has to be returned and money back!
I´m really angry about this endless betatesting and trying hours after hours and loosing time for my clients and my daily work.

Exacly my point! I am not interested in the actual developement. I gave a good money for this and I want it working like it is described! I have a work to do... They shouldn't sell unfinished products.

[APADRecordings]

When I speak of capacitor size I'm referring to values. Even with that, though, an electrolytica can behave differently from a tantalum.

I neglected to address another issue - that of multiple looped units. Several years ago I started with two Travelers under MS Windows XP, SP2. I read the MOTU manual stating that as many as 3 units could be serially connected on the FW buss. I ran the two Travelers that way for about 3 years with no problems. I upgraded to XP SP3 and encountered no problems. I ran into a situation where I needed 12 microphone inputs so added an 8Pre and bogged down. So I used the 8Pre as a converter only running into the Traveler ADAT buss. However, when MOTU released the 1394 driver that changed and I was able to run all three on the FW buss.

As I write this I am playing a long file in Sound Forge. I have a Traveler MKIII, Traveler, and 8Pre looped and connected to a Unibrain FW adapter in the expansion slot of an IBM Thinkcentre 8171 (915V chipset) running XP SP3 with 1 GB of DDR memory. The file playing resides on a MS Home Server about 30 feet away. The server and Thinkcentre have Gigabit Ethernet out to a Gigabit Ethernet network.
There are no hiccups, hesitations, squeals or pops. It just works. The only "tweak" to XP was to set it up to give preference to programs running in the background and to run in "maximum 'erformance mode." So I'm playing the file across my network, running WordPad, and accessing the Web at the same time. Everything is stable.

When CueMix FX pops up I usually close out the 8Pre and Traveler versions while keeping the MKIII version up but depends on the job. I use Sony Vegas and Sound Forge for recording - again, depends on the job.

MOTU did extensive work on the timings in 1394 which is retained in 1400 and both work smoothly.

[APADRecordings]

I neglected to mention that the PCB layouts I'm referring to are the computer motherboards. Some boards work very, very well and others only marginally. G10 or FR4 material can be particularly problematic if it has been "touched up." When expoxy is heated to the glass state, as can happen with a rework, the dielectric constant in that particular spot changes and devices at that point operate very differently. I've been out of the business since the lead-free standards were adopted but the lead-free solders requiring even higher temperatures and more re-work. As speeds and complexity increase so do the problems.

Obviously MOTU checks its own work carefully and can easily since it is all manufactured in the US.

My avionics engineer son will tell you that if you want to use a computer for a particular task you'd better test it extensively for that task. The XP operating system is what it is and came before the present computers. It's Intel's problem, working in conjunction with Microsoft, to come up with working chipsets. Generally they work OK but my conversations with Intel indicate that if they can transfer pictures via FW that's their primary interest. FW is pretty far down on their list of interests. USB is way up there but it is not isochronous!

[APADRecordings]

My point is this: Regardless of the computer manufacturer, his reputation, etc. the computer and its implementation of the CPU and chipset are the great variables - not the operating system nor the converter manufacturers.

The operating system is what it is.

Folks like MOTU build their converters to a spec from Microsoft and they work. I have 8 computers here that work beautifully with the Traveler MKIII, the Traveler, 8Pre, 896MKIII and the 828MKIII and/or any combination of them with up to 3 on the 1394a buss. They do what they do very well and, I think, are more versatile than anything else out there at any price. They are a superb value!

I also have 2 other computers that don't work reliably and 1 that cannot and will not ever work because the Dell I/O system is seriously compromised. There was no way to know this in advance. The manufacturers all have good reputations and reviews.

For streaming audio over the FW buss the problem is not MOTU and not Microsoft. The problem comes down to the computer motherboard and chipset implementation.

[dare]

Uh... Thanks for the extensive info... Ok I agree - even the different manufacturer makes slightly different parts that should be the same... That is why there is a tolerance printed onto them When the board is made it undergoes a series of tests to confirm it complies with the demands. If for example some board doesnt return correct values - then it can't be considered functional. Of course there are a lot of factors that affect some piece of hardware, but they are all taken into acount.

But we could go like this forever... The point is: I understand that problematic chips may cause errors like blue screens, dropouts, etc... but wrong routing??? You have already told us that you have the same messed routing issue... So there are no exceptions.

Besides that, everything works fine with me too, but that routing thing... And what about their answer that they are aware of the problem ? They would simply reply me with - go and get another FW card, but not this time. I use ASIO on CUBASE 5 - so it is a way more demanding job than just recording in vegas or forge... A lot of VSTs, Powercore and UADs - which are extremly timing critical - all works fine - but on MAIN OUT I have ANALOG 1+2 and *that* is hurting me a lot! You can't tell for sure which part of the drivers-hardware-software chain is not working as it should (you can't blame the chips as I can't blame the drivers) I simply refer on what they post as an answer to my question... They shurely have done some testing before their reply...

[TheRealRoach]

For streaming audio over the FW buss the problem is not MOTU and not Microsoft. The problem comes down to the computer motherboard and chipset implementation.

This is a terribly inaccurate statement. It is well documented by many users on this forum that a replacing various stock MS-supplied drivers will remedy FW bus issues.

[APADRecordings]

I'll have to take issue you with RealRoach. The fact that Dare and I both have systems that work fine with the standard OS indicates that "it's not broken." You, yourself, indicate that alternative drivers will "remedy" FW issues. Sure, but they are merely compensating for problems created by poor implementation of FW by the motherboard/chipset designers and/or those writing drivers for the chipsets. I've looked through this and other forums and those "fixes" only work for some and actually make it worse for others - another indication they are merely compensation for problems introduced by others. For example, I downloaded the latest greatest driver for the chipset on one of my computers and immediately experienced serious FW problems. I could have tried replacing the "stock MS-supplied drivers" and gone on a goose chase. Instead, I rolled back to an earlier Intel driver that had been certified by IBM and the problems disappeared.

I'm glad they may work but I still maintain many folks are blaming the wrong people. They are going under the assumption that their computer is OK, the standard if you will, and it's up to everyone else to make drivers that make it work right!

No manufacturer could possibly test for all the variations out there. They have to build to a specification and that specification happens to be supplied by MS. Eight of my computers here work just fine. Two are not reliable and one will never work (that's how crummy the motherboard designer was).

[dare]

I have to add one more thing: IF some systems are working and some not, then it is clearly something wrong with the systems in question (bad chipsets, etc.), but if an error is present on every system under every OS then it is probably driver's fault. So, in the case of crackles and pops - I solved that by going on Windows 7 - maybie different driver and system structure solved it... Or compensated for the errors on my FW bus, doesn't matter - MOTU is not guilty on the matter I personally think that M$ is. Their issues with the FW standard are documented by themselfs - FW is simply not fully standardised. Maybie some FW adapters are really crappy and chipsets too, but in my case - changing system was enough. I would like to hear more reports on Win7 and MOTUs that drop auido... And to add - Motu isnt the only one which has that issue... TC for example.

[yeti]

halllo... want to ask again:

... will motu be willing or be able to fix that?

what do you think...
addendum: WHEN?

[TheRealRoach]

It is a wild goose chase indeed.

I cannot explain why your stock setups work and but I am curious to find out any common thread that is continuous for your eight functioning systems, and likewise the three that are hopeless. You have mentioned motherboard design and their driver sets being the key, but I would seriously be surprised if all eight systems are considerably different in brand (Asus, Gigabyte, etc), age, format (P4, dual-core, etc). There must be some observable successful trait that these stock systems all share. Even if they are all of identical brand, age, format, etc, I would still be interested to know what the common thread is. For example: All of the functioning systems are Asus-based, or all of the functioning systems are XP SP3, or all use the same Unibrain FW cards, or all use onboard FW ports... etc.

[Mark S]

I'll have to take issue you with RealRoach. The fact that Dare and I both have systems that work fine with the standard OS indicates that "it's not broken." You, yourself, indicate that alternative drivers will "remedy" FW issues. Sure, but they are merely compensating for problems created by poor implementation of FW by the motherboard/chipset designers and/or those writing drivers for the chipsets. I've looked through this and other forums and those "fixes" only work for some and actually make it worse for others - another indication they are merely compensation for problems introduced by others. For example, I downloaded the latest greatest driver for the chipset on one of my computers and immediately experienced serious FW problems. I could have tried replacing the "stock MS-supplied drivers" and gone on a goose chase. Instead, I rolled back to an earlier Intel driver that had been certified by IBM and the problems disappeared.

I'm glad they may work but I still maintain many folks are blaming the wrong people. They are going under the assumption that their computer is OK, the standard if you will, and it's up to everyone else to make drivers that make it work right!

No manufacturer could possibly test for all the variations out there. They have to build to a specification and that specification happens to be supplied by MS. Eight of my computers here work just fine. Two are not reliable and one will never work (that's how crummy the motherboard designer was).

FWIW: the firewire standard is a MAC standard. A number of us who have had to dig into this have found from sources way more in the position to know that the standard as written is by MAC inspecific in a crucial area. This allows for performance variances between manufacturers that can still be considered within the standard. It also allows for Microsoft to write a 1394 driver that is 'in the standard' but allows for variances of results. XP SP1 was good but didn't work with multi proc machines. SP2 won't work with daisy chained set-ups. My 896hd works well on every machine I own (via and nforce) but I add the 8pre, which also works well on it's own, and neither works until I change to the AVT driver which is written to the more robust end of the standard.

Having gone through all of this those of us who've had to work to get our systems working have had the benefit of advice from some of the best DAW builders in the country. It is from them I've found that a TI, Agere/Lucent chipset for firewire controller is highly advised. Nforce2 and some Via work. (lucky for me on both but I did my homework before I bought them too), and In Microsoft Windows systems past XP SP1, you will not get two units daisy chained. Everyone on any forum, for any maker of audio unit, who has done this is: 1) using sp1 driver version for the 1394 driver, 2) connecting optically, or 3) using 3rd party drivers. I have had luck with all three ways but the most solid success with #3. This was the biggest issue RR was dealing with.

To make the blanket statement about streaming not being a MS problem is a statement that doesn't stand up to the experience of many qualified people who have had to work this out, RME who first published the fix, many professional DAW makers who implement this fix as SOP, of AVT who were unable to get their high end firewire cameras to work with and version of Windows past sp1 including Vista and were forced to write their own driver,...

Though many may have a finger in the pie on this one Microsoft has the potential for biggest impact just by fixing their driver. Maybe W7 will fix it,...

[APADRecordings]

Appreciate your comments and questions.

Last night I was at the Bass Performance Hall in Fort Worth recording the Youth Orchestra of Greater Fort Worth. I had, daisy-chained, a Traveler MKIII, Traveler, and 8Pre going to an IBM Thinkcentre S51 (8171) with a Unibrain 3 port FW adapter in the expansion slot. The IBM has XP Pro SP3 with no changes other than to give priority to background operations and set for maximum performance. Daisy chaining 2 MOTUs was no problem with SP2 or SP3. When MOTU posted driver 1394 and now 1400 it became possible for me to daisy-chain three devices.

Machines that work well for me:
(1) ASUS Z7100 notebook, 865 chipset, with cardbus adapter using TI chip;
(2) IBM Netvista S42, 845G/GL chipset with Unibrain FW adapter in one machine and a Belkin Adapter using an Agere chipset in the other
(1) IBM Thinkcentre S50 (8183), 865GV chipset, Unibrain FW adapter;
(4) IBM Thinkcentre S51 (8171), 915GV chipset, Unibrain FW adapters

Machines that are unreliable:
(1) ASUS P4C800E Deluxe, ATX, 875P chipset and on-board FW controller [don't know what it is]. However works with 2 MOTUs using Unibrain FW Adapter installed in expansion slot
(1) Intel DG631PR [Core 2 Duo with G31 Express chipset) Ended up installing Windows Home Server on this machine. Great job in that application

Total Disaster with severlely compromised I/O subsystem:
(1) Del D830 notebook, GM965 chipset, either cardbus or Expresscard adapter

I also tested an IBM Thinkcentre M55e which I believe used a 945GV chipset. It also was a disaster for Firewire. Superb in every other respect. I have not found a Core2 Duo machine that was usable. I'm sure there are some but I personally haven't found one.

[APADRecordings]

I purchased two Travlers in the spring of 2005 because I wanted 8 microphone inputs. I didn't know there was supposed to be a problem with daisy-chaining two or more devices so I did and they worked fine. FW was definitely more "touch" than my older PCI system but it worked, and in fact, the Travelers enabled me to come up with a system to record wirelessly. http://www.apadrecordings.com/wireless_recording.htm

I have another SKB rack case with an 896MKIII and another Traveler daisy-chained and it also works great with the IBM Netvistas and Thinkcentres and the ASUS Z7100 notebook.

I use Microsoft Live! One Care on my computers for virus/malware/adware protection. It's well worth the 50 bucks a year since MS maintains it and I never have to fool with it.