My guesses:
Sensitive USB components (either in the QU, the PC or both) combined with a weak USB cable?
Software - something that takes a mere few ms will be enough to kill the USB connections, and there are so many things nowadays that assume you have infinite CPU I/O and memory… If I was relying on it then I’d run a standalone PC…
Have you tried with a SanDisk Extreme in the QuDrive port?
Technical note: USB Audio frames are scheduled each 125µSec (1/8 mSec). A single missed frame will at least generate a click. A single driver in your system perfoming some sort of busy wait or defers execution of USB handling in combination with non-buffering hardware will break audio streams easily.
@VK
Have you tried the latency checker utilities referred to in our KB article? It would be interesting to see your results.
Thank you for that link, Nicola. I’ve dived into that a little further and remembered the Microsoft tool RATT and found this informative thread in the cocos (Reaper) forum. It basically describes how to track down DPC issues using RATT but even more interesting is, that onboard audio drivers more than once seem to be the culprit for these issues.
Yeah… so I thought my issue was fixed. As I had changed the video card to and AMD version it appeared to go away for a while but now its back like crazy. I too have to change my sampling rate again and then eventually have to reboot my system. There is enough users with issues now. Where is tech support with a solution?
@philt80 so you’ve already used the latency checker and posted results somewhere?
Yep! To tech support in my initial call log. Latency showed the usb driver giving all the latency issues (this kinda makes sense??). Since that its been about 8 months of my own troubleshooting, with no real aid from support other than to fool around with the latency settings. I’ve tried them all, changed hardware, daws, cables etc. I’m thinking that with all the others with the same issues on this forum its probably not some obscure piece of hardware. I’m using consumer grade products off the shelf, no modifications. No overclocking, no crazy hardware. No other board does this on the same computer when I try to record.
Pretty frustrating when you spend some significant bucks to purchase and ship. Totally not giving A&H a good name when I have to reboot or quit my DAW randomly within the hour when I’m trying to show clients, mix or record. And I purchased my QU24 because of the reputation the company had as a whole.
If its indeed the USB portion which cause the trouble on your system which also changed with a different graphics card, this sounds like bad hardware interrupt sharing for me. If that’s the case, there is probably nothing you (or A&H) can do on that platform to fix this issue.
Which other boards work well on that PC? Are these USB connected? Also streaming 32+30 channels as your Qu24 does?
Hi Andreas,
I have just returned from holiday and have been watching these posts whilst i have been away.
With the very greatest of respect, and whilst I understand you’re trying to help, I - probably like philt80 - have tried pretty much everything. And like him, I bought the desk because it was recommended to me as a great product without the need to purchase an internal soundcard.
I cannot keep taking my pc out of my rack. Aside from the fact its extremely heavy, it’s also time consuming and requires the help of another person to help lift it out. Like philt80, I have changed cables, settings, soundcards etc all to no avail.
What is apparent though is that like yourself, no-one has an actual solution to the problem. All i hear is supposition and theory which is equally as tiresome. My studio has now been out of operation for some 2-3 weeks preventing me from working, unless, and again like philt80, I spend another small fortune replacing my desk and purchasing another internal high end sound card for my pc. There is little point telling me “it could be this - it could be that - or that the problem is a big one which may have many reasons as to try and solve it.”
That’s no more helpful that sticking me in Leicester Square with a blindfold on - and telling me to find Marble Arch!!! “yeah - you’ve a problem, but if you walk this way or that way you’re bound to find it in the end”
It’s really not good enough at all - and the least A&H should do is acknowledge this problem that they can’t solve and take the desk back and swap it with one which has digital inputs and outputs as a measure of good will so that I can at least start work again.
I am unlikely to believe that this is the first time they have encountered this problem as so many of us seem to have it. There is also little point in me starting work again in my studio - until someone actually comes up with the solution, because as soon as I start programming, it takes about 5 minutes for the distortion to occur and less time thereafter for it to re-occur even after messing around with buffer and latency settings.
So please - I don’t wish to sound ungrateful at all - but none of your suggestions or theories have proved helpful as to sorting the problem out. I can only hope that the next time you or anyone posts on this thread, it will be with a solution that actually CURES the problem, and doesn’t have me running around in endless circles.
Yours very gratefully
vaughank1
VK, there is no reason to take out your frustration on people giving their time to attempt to help a problem. They are giving their time freely to help complete strangers with a problem, yet you are acting like you’ve paid them to give you answers. It’s a free forum where folks from around the globe can exchange ideas and communicate about these consoles. In my opinion your attempt to police this thread by demanding solutions from folks is not welcome nor in the spirit of these forums.
You are quite right. You have my profusest apologies and would like to state that I have tried to delete the post. There is absolutely no offence intended to anyone including Andreas and it was borne out of sheer frustration of having a 3 week shut down studio plus excess money spent to try to stop the problem.
I am writing this from my mobile and if my last post has not gone by Monday when I return home - I shall connect via PC to delete.
Again - accept my apologies.
Vaughank1
Hi all,
Please can I encourage anybody who is having ongoing issues to contact us directly on support@allen-heath.com. The email will create a support ticket where we’re happy to and work through these issues and try and help you resolve them.
You’ll need to include as much system info as possible (Computer, OS, driver version, firmware version etc) along with any troubleshooting already attempted.
I’m sure you can appreciate that on a PC there are many more factors which can affect USB audio performance but we’ll do out best to try and help troubleshoot.
Look forward to hearing from you.
Regards
Mike
USB audio isn’t simple to diagnose.
The timing required by a USB device such as the QU series isn’t really catered for by the USB standard.
Most of the time it will be fine, because the USB bandwidth is provided in excess, but guaranteed timing isn’t the forte of USB.
I presume there is a USB version of wireshark that might help?
Hello All
In second and probably most humble of apologies to Andreas, and also other forum users regarding this thread, I really have been to town to try to discover exactly what the problem is with the desk - the pc and the distortion noise.
One thing that was evident, is that the problem doesn’t seem to be lying with the desk. It’s almost definitely my PC.
So again, thanks to Andreas - You were right!!!
I have plugged the desk into my laptop which also runs windows 7 and there is not an issue at all with it. I have yet to load cubase onto the laptop and if there are no further issues, then my main PC is surely the problem… Which is a shame, because I have had the main PC actually tested by Microsoft (a friend of mine who works and writes for them) and they have said that I have no conflicting software or hardware issues, bad drivers or anything. So trying to ascertain exactly what component my Desk doesn’t like inside my pc is now practically impossible - which makes any single component impossible to replace now.
It seems like I’ll have to ditch this pc and get another.
Unless there is any other possible suggestion that anyone can give for me to try with the PC before it goes.
All in all - it’s been a shambolic waste of £2k (the PC that is…) - but sadly it doesn’t ease any of the frustration of a closed down studio and a complete stoppage of work.
Andreas, My profusest apologies, which I hope you’ll accept as to all the forum readers who were offended by me from my last but one post.
Wth best regards
Vaughan
@Vaughan, it takes much more to really offend me. 
VK I feel you pain and I dont think you mentioned anything that wasn’t true or said anything to get people all emotional. When one buys a product for a decent amount of money they expect support. I have not seen a solution to this anywhere on this site nor from tech support. Buying another PC is not a solution because I can guarantee that the laptop if you continue to use it will still buffer out as mine did. I on the other hand am not going to spend another 2k to purchase another PC to get this fixed. I’ve had to get my QU24 shipped (125$ CAD) and I’m not willing to pay to ship it back to be “fixed” then have it shipped yet again to find out it was not really fixed.
Mike thanks for your post but I was talking to you, I did send everything that I could and I have sent (even posted on other threads), specs, driver versions, sound clips etc but the most that I got from you was to adjust my buffer settings. It looks like everyone else is fighting with buffer settings and not winning. I have not heard from you since and am not even sure if my call is still open.
IMO there is lots of products streaming USB, many of them without this issue. I don’t believe its a hardware “interrupt” issue but I could be wrong. Nothing that I can see is chewing my processor or crushing my drive except for the A&H driver.
Also I’m not ungrateful of all the posts, we are all trying to help/work here but I agree with VK, there are lots of suggestions without solutions and right now I just need a solution.
Thanks for your time!
philt80
I have this problem too. And here is an audiofile to listen.
Thanks for the screenshot of Latency Monitor. Two points immediately come to mind (and I should do an analysis on my machine too). On first the DPC execution time of the USB driver itself. About 84µSec is rather large knowing that USB interrupts likely occur each 125µSec.
But I’m much more concerned about the 55µSec interrupt time from the DX driver. ISRs are executed directly on behalf of the hardware interrupt and will delay anything else in the system, particularly DPC routines. Situation may be relaxed on a multicore system, depending on implementation of the ISR. 55µSec sounds like nothing, but if this one delays the execution of the USB DPC, that one continues to run with the occurrence of the next USB interrupt, since 55+84=138µSec > 125µSec (isochronous frame rate).
This may explain that changing the graphics card “fixed” the issue for some users (at least for some time).
What other applications are running concurrently to the DAW in that situation? Anything using Video (like web browser, skype etc.)?
Wondering that that DX ISR is intended for, maybe VSync?
Well, that’s guessing around, I’ll do a LatencyMon run to get something to compare with.
Ok, here we are. LatencyMon really is a neat tool to gather some internal timing information from drivers and internal mechanism. However, it needs a decent amount of interpretation.
Most interesting thing I’ve found is, that my two systems both running Win10/64 Bit but expose rather different results in terms of driver names and use, which I can’t explain yet.
My rather old Dell (Core2Duo @ 2GHz) uses USBPORT.sys as low level USB interrupt driver, which seems to feed DXGKRNL.sys (named DirectX Graphics Kernel) with seems to consolidate audio frames into sample clusters and in turn calls the QuUSBDriver_x64.sys on a rather low (=uncritical) DPC rate.
My MacBookPro (Core i5 @ 2.5GHz) in contrast obviously uses Wdf01000.sys (Kernelmode Framework Driver) to take interrupts, consolidate sample clusters and feed QuUSBDriver_x64.sys. Both USBPORT.sys and DXGKRNL.sys are both present but operating on a much lower rate.
My CoreDuo was still running the plain old V2.20 driver which I’ve updated to V3.34, didn’t make a big difference except the 3.34 Qu driver takes significantly less cpu time.
On the MacBook I’ve tested both V3.20 and V3.34 drivers. The 3.34 still is an improvement over 3.20 in terms of cpu time, but less significant as the step from V2.20 to V3.34, obviously.
Using the old CoreDuo I can easily produce frame drops when activating the WiFi driver. Its fun listening to the Explorer operating on a network drive. But I never could reproduce that robotic noise as described here.
And, finally, my LatencyMon results are much worse than those showed in Slowhand’s screenshot, particularly on the CoreDuo system which successfully did hours of recording before I’ve switched to the MacBook.
Still not smarter.
