dimelives1
Thanks for the responses, guys.
 
Yeah, I was sure to only install the driver itself and leave out the 3D, PhysX, and GeForce Experience add-ons.

I just ran back through this thread and caught this. You actually do want the PhysX drivers. Try running the installer in custom mode again and "perform clean installation" at the bottom. Only the Display drivers, and PhysX drivers should be checked though.
 
Also watch during that installation. I forget if Nvidia slips in other things (I believe it re-enables the HDMI audio outputs in Device Manager as well, which should be disabled); but fortunately, they are all discrete entries in the Windows uninstall dialogs for "Software and Programs" Do a quick check of programs by installation date in Windows after the install just to check those.
 
IIRC, every driver update requires the HDMI audio outs disabled in the Device Manager.

mettelus
I just ran back through this thread and caught this. You actually do want the PhysX drivers. Try running the installer in custom mode again and "perform clean installation" at the bottom. Only the Display drivers, and PhysX drivers should be checked though.
 

 
Thanks for the suggestion, mettelus. I gave it a try, after doing a total wipe of NVIDIA from my system as recommended in an email from NVIDIA support, but it made no difference. The behavior has remained.

It sounds a lot like a driver issue, but hard to tell. Have you checked the Device Manager and disabled the HDMI audio?
 
Have you tried running Resplendence LatencyMon (it is in the free downloads on the left on that page)? For most machines, audio buffers in the 256 sample range will usually play fine for most things, but not the best for tracking. As far as drivers specific to the Nvidia card, I can only think of the HDMI audio if you didn't load the bloatware.
 
LatencyMon will give a little better insight on what is running on the machine.

Try disabling Aero Peek.  You said disabling the NVidia driver has that effect and maybe that's the problem.

I just read through this entire thread again and one thing I couldn't find, mettelus, was what you're using for audio hardware. It was the post about shared IRQs that got me thinking about this.
 
Assuming you're using outboard audio hardware connected via USB... have you tried plugging your audio hardware into a different USB port? Sometimes it matters, too, if you've got a USB 2.0 device plugged into either a USB 1.0 (or 1.1) or USB 3.0 port. They're all supposed to be backwards/forwards compatible, but it's really kind of a crap shoot. You'll have to consult your MB manual to find out which USB ports are 2.0, 1.0 and 3.0 although they are color-coded so you could just try it in a port that's a different color than the one you have it in now.
 
And if it's a PCI or PCIe card, do you have another slot you can move it to?
 
Be warned, however, that switching slots or ports could necessitate reinstalling drivers depending on which version of Windows you're using.
 
If it's an IRQ issue, shaking things up like this can make a difference (even though that whole FF thing still makes no sense in this context and seems more like the memory allocation issue that was being bandied about earlier).

Hi, guys. Still experiencing the issue, but I've stumbled across another thing that I can use to make it go away besides having a Firefox window open: having an image open in Windows Picture Viewer... So... make of that what you will...
 
 

mettelus
It sounds a lot like a driver issue, but hard to tell. Have you checked the Device Manager and disabled the HDMI audio?

 
Yes, I've disabled anything and everything related to high definition audio in the DM.
 
 

mettelus
Have you tried running Resplendence LatencyMon (it is in the free downloads on the left on that page)? For most machines, audio buffers in the 256 sample range will usually play fine for most things, but not the best for tracking. As far as drivers specific to the Nvidia card, I can only think of the HDMI audio if you didn't load the bloatware.
 
LatencyMon will give a little better insight on what is running on the machine.

 
I'll have to give that a try. I'll report back with any relevant results from that after I can get the time to run it.
 
 

BRainbow
Try disabling Aero Peek.  You said disabling the NVidia driver has that effect and maybe that's the problem.

 
Thanks for the suggestions, BRainbow, but I've already tried disabling all Aero effects. It didn't make any change in the behavior.
 
 

rontarrant
I just read through this entire thread again and one thing I couldn't find, mettelus, was what you're using for audio hardware. It was the post about shared IRQs that got me thinking about this.
 
Assuming you're using outboard audio hardware connected via USB... have you tried plugging your audio hardware into a different USB port? Sometimes it matters, too, if you've got a USB 2.0 device plugged into either a USB 1.0 (or 1.1) or USB 3.0 port. They're all supposed to be backwards/forwards compatible, but it's really kind of a crap shoot. You'll have to consult your MB manual to find out which USB ports are 2.0, 1.0 and 3.0 although they are color-coded so you could just try it in a port that's a different color than the one you have it in now.

 
Hi, Ron. I think your questions were intended for me (I'm the OP).
 
I never experienced any issues like this with my audio hardware setup in the years that it's been configured exactly the same way. The problem only cropped up after installing the new graphics card.
 
 

rontarrant
And if it's a PCI or PCIe card, do you have another slot you can move it to?

 
Unfortunately, I've only got the one slot for it. My mobo is pretty basic.

dimelives1
I'm stumped. What can be gleaned from this? The only thing I can think of is maybe it's some kind of memory issue, like the video card for some reason needs more to do in able to run smoothly in Sonar?... But how would that make any sense?
Opening a browser window is even easier than disabling the driver as a workaround, but I'd love to know WHY it's having the effect that it is.
Any theories?... 

Could well be the driver service routines behave differently. You'd probably see something interesting running Resplendence LatencyMon when the glitching is happening.

Keith

dimelives1I never experienced any issues like this with my audio hardware setup in the years that it's been configured exactly the same way. The problem only cropped up after installing the new graphics card.

 
Yup, and with Windows being the unpredictable beast it is, you may have to compensate in what seem like strange ways to get everything working together again.
 

dimelives1Unfortunately, I've only got the one slot for it. My mobo is pretty basic.

 
Okay, can you borrow a USB audio device from someone and try that? You'd have to remove your PCI audio card to make it a fair test.
 
Just curious: what MB do you have (make/model)?

OK, I finally go the chance to run Resplendence LatencyMon to see if it yielded any helpful results. I tested it with nothing running but Sonar (and LatencyMon, of course), where the stuttering playback occurs, as a control test. Then I ran the same stretch of audio with a blank Firefox window open, where the stuttering does not occur. Here are the most notable discrepancies between the two sets of data; I'll leave it to those more savvy than I with these kinds of stats to say what may or may not be relevant:
 
MEASURED INTERRUPT TO USER PROCESS LATENCIES
 
Highest measured interrupt to process latency (µs)
Control:  2151.312516
w/Firefox:  845.879520
 
Highest measured interrupt to DPC latency (µs)
Control:  1664.384298
w/Firefox:  600.875581
 
REPORTED ISRs
 
Highest ISR routine execution time (µs)
Control:  1897.498997
w/Firefox:  949.337567
 
Driver with highest ISR routine execution time
Control:  dxgkrnl.sys - DirectX Graphics Kernel, Microsoft Corporation
w/Firefox:  dxgkrnl.sys - DirectX Graphics Kernel, Microsoft Corporation
 
ISR count (execution time 1000-1999 µs)
Control:  51
w/Firefox:  0
 
REPORTED DPCs
 
Highest DPC routine execution time (µs)
Control:  1868.751337
w/Firefox:  876.690842
 
Driver with highest DPC routine execution time
Control:  nvlddmkm.sys - NVIDIA Windows Kernel Mode Driver, Version 372.54 , NVIDIA Corporation
w/Firefox:  USBPORT.SYS - USB 1.1 & 2.0 Port Driver, Microsoft Corporation
 
DPC count (execution time 500-999 µs)
Control:  22
w/Firefox:  33
 
DPC count (execution time 1000-1999 µs)
Control:  5
w/Firefox:  0
 
PER CPU DATA
 
CPU 0 ISR highest execution time (µs)
Control:  1879.471925
w/Firefox:  921.260695
 
CPU 1 ISR highest execution time (µs)
Control:  1755.911430
w/Firefox:  653.038436
 
CPU 1 DPC highest execution time (µs)
Control:  1313.464572
w/Firefox:  284.248997
 
CPU 2 ISR highest execution time (µs)
Control:  1835.136698
w/Firefox:  897.509358
 
CPU 2 DPC highest execution time (µs)
Control:  1868.751337
w/Firefox:  345.708890
 
CPU 3 ISR highest execution time (µs)
Control:  1897.498997
w/Firefox:  949.337567
 
CPU 3 DPC highest execution time (µs)
Control:  1806.969586
w/Firefox:  269.467580
 
CONCLUSION
 
Control:  Your system appears to be having trouble handling real-time audio and other tasks. You are likely to experience buffer underruns appearing as drop outs, clicks or pops. One or more DPC routines that belong to a driver running in your system appear to be executing for too long. Also one or more ISR routines that belong to a driver running in your system appear to be executing for too long. One problem may be related to power management, disable CPU throttling settings in Control Panel and BIOS setup. Check for BIOS updates. LatencyMon has been analyzing your system for  0:01:12  (h:mm:ss) on all processors.
 
w/Firefox:  Your system appears to be suitable for handling real-time audio and other tasks without dropouts. LatencyMon has been analyzing your system for  0:01:13  (h:mm:ss) on all processors.
 
 
(I also ran a test with Windows Picture Viewer open, which stops the stuttering from happening too. The results were virtually identical to those of having Firefox open.)
 
--------------------------------------------------
 
So, is there anything that can be determined from any of that? Keith would appear to me to have been on the right track with the driver routines. But are we able to figure out a "why" or "what can be done about it"?
 
 

rontarrant
Okay, can you borrow a USB audio device from someone and try that? You'd have to remove your PCI audio card to make it a fair test.

 
I apologize, Ron, I mistook you before; I thought you were referring to the video card.
 
My audio interface is a USB type, the Roland Cakewalk UA-1G (which some may scoff at, but it has served me well for my needs, up to installing the new video card).
 

rontarrant
Just curious: what MB do you have (make/model)?

 
It's a 0K83V0 board, from a Dell Inspiron 560s. (Hey, I'm on a tight budget over here)
 

the nvidia control panel has settings-per-app, it's worth checking those for sonar if you haven't already