Strange random dropouts with no detectable cause.

I have 2 desktops running Sonar (1 at a time), running the same up to date Windows 8.1.1, same Sonar X3e and Platinum Ipswich on both, same plugin versions for all 3rd-party effects and synths.
 
1 computer plays back a specific project, with no problems.  The other computer, playing the same project, has random dropouts of sound, but other than the loss of sound there is no discernible evidence of it.  (Meaning there is no taxing of CPU, memory, disk, and all effects are bypassed, and there is no indication in the Performance Module in the control bar, of anything amiss).
 
The 1st computer does have 32 GB of memory, and the problem computer has only 24 GB, but there does not seem to be any indication of the CPU not keeping up, or the memory being taxed, or there being any disk streaming issues.
 
Both computers have solid-state C: drives, another regular SATA III 7,200 rpm drive for Cakewalk Content and Projects, as well as another SATA III 7,200 rpm drive for sample libraries.
 
I just do not know where to look for why this is happening to only that 2nd computer.  (the 1st does have an Intel i7 2600k, while the 2nd computer has an AMD 8350 with 8-cores), but neither CPU is anywhere close to being stressed.
 
Both computers were running the exact same audio interface, with the exact same ASIO driver versions loaded, and the exact same sample rates and ASIO Buffer Sizes, along with matching Sonar Buffer sizes.
 
(Edited to add the following): By the way, I have the ASIO Buffer Size for both computers running at 1024, and the dropouts are occurring during simple playback, with all effects bypassed, and even just soloing on 1 audio track from a single midi track.
 
Anyone have any ideas?
 
Bob Bone
 
 

None of the drives show any fragmentation.
 
I DID just run LatencyMon, and it does indicate some issues may be present to cause dropouts, so I am in the process of looking at that.  I should have done so prior to even posting, but I really hadn't thought there would have been any issues - DOH!
 
Bob Bone
 

Thanks - I have upped the ASIO Buffer Size to 1024, none of the disks show fragmentation, there is very little CPU demand, Sonar Projects are on a separate drive, Sample Libraries are on a separate drive, and on the C:, there is only Windows, its Recovery Partition, and installed applications.  I altered the Location parameter for all user libraries (Downloads, Desktop, Pictures, Documents, etc.) to be on a different drive, as well, as the C: is a 240 GB SSD drive.
 
And yet, as the project plays along, even with all effects bypassed, it will have gaps in the playback sound, for roughly 1 second at a time, and at various times throughout the playback.  Additionally, I have 1 instance of Kontakt 5 loaded with a factory-supplied Violin Solo patch, and for some reason, in this project, there is a secondary distorted sound being played along with the regular violin sound.  This secondary distorted sound is an octave or so lower than the original violin pitch, but makes the whole sound quite unusable.  I do not know what is causing this to occur either,.
 
I am hoping that someone has faced some or all of the above, and can share some things for me to look at or try.
 
Thanks SOOOOO much, even for just reading through all of this - it is QUITE frustrating.
 
Bob Bone
 

Regrettably, both X3e and Platinum Ipswich have the same issues
 
Also all of the NI plugins are up to date, too.
 
Bob Bone
 

LatencyMon is a good guide, but another utility I find handy is Moo0 System Monitor (I use the portable version since there is no installation required). It is very rudimentary program which monitors system throughput and will flag chokepoints as they hit. The disk I/O portion of the program can be handy in seeing exactly what your hard drives are doing.

You forgot one, Craig... possible eminent hard drive failure.... I had this issue last year, then the drive went kaput, changed it and all's gravy... really sucked as it was my C drive which I hadn't backed up for awhile...a Samsung ssd to boot!

Hmm... It could be that Windows 10 "update". It is reported to pre-download the whole image of several Gb big, even if you did not opt-in for the upgrade. This seems to happen if you have automatic updates enabled. It could be that the associated disc/network activity is the cause of that drop-outs.
 
You can see if that is the case if there is an (hidden) directory on your hard drive named: $Windows.~BT. This directory contains the Windows 10 installation files and is between 3.5Gb and 6Gb in size.
 
If you have disabled automatic updates then there must be another reason for that drop-outs.

robert_e_bone
 
I found that there is a diagnostic mode on the UI for my Native-Instruments Kore 1 audio interface that I have been testing with, and is is showing repeating USB Data Errors, about 1 every 40 seconds, when the interface is being used by any Sonar project - either in X3 or Platinum.
 

This is sort of good news. Anything periodic can be tracked down. I would start with scheduled tasks and network activity, but also consider exotic causes. You could even be filling up some (other) buffer that's not apparent - like I remember crashing my digi mixer every 58 seconds until I figured it's a MIDI loop that silently fills the buffer until the whole unit decides to say good-bye ... always thought this was a USB problem ...
BTW, how stable is your power supply? Are you and your gang opening the fridge every 40 seconds for cold brew???

robert_e_bone
Well, I have more analytical work to do, on figuring out what is happening, but I have been working through it.
 
I found that there is a diagnostic mode on the UI for my Native-Instruments Kore 1 audio interface that I have been testing with, and is is showing repeating USB Data Errors, about 1 every 40 seconds, when the interface is being used by any Sonar project - either in X3 or Platinum.
 
I changed USB cables, and it made no difference, and changing to a different USB 2.0 port made no difference either.
 
So, I will try using a Focusrite Scarlet 2i2 and see what happens there.
 
Thanks to all for helping - I am pretty sure we'll get this figured out.
 
Bob Bone
 

What video cards are you using for each of the systems? 

I ask because I was getting pretty consistent drop outs on my FX-8350, 16GB, 2 SSD's, 1 HHD, M-Audio R-Bus rig after upgrading to a EVGA 760 GTX. LatencyMon indicated that the Nvidia driver was causing some interrupts in streaming. Performance was okay for very light projects, but anything more complicated dropped out constantly.
 
I had to switch the video card from Line-Based to Managed Signal-Based interrupt. This thread goes into it in more detail: http://forums.guru3d.com/showthread.php?t=378044
 
Make sure you're not getting any faults with LacenyMon first: http://www.resplendence.com/latencymon
This was the utility I used: http://www.mediafire.com/download/hrxw4rsb9xnw1ei/MSI_util.zip

 
Make sure you have a backup to roll back to! In my case I didn't experience any issues switching all of my devices to MSI, though I know other users have experienced instant blue screens on the switch. 

OK, well I have been making progress with all of this, though am not finished analyzing and testing methods of elimination of audio dropouts in a specific project.
 
This block of testing has been being done with the Native-Instruments Kore 1 Audio Interface, 
With regard to this project, and results of running with the Native-Instruments Kore 1 Audio Interface, I found a regular repeating number of ASIO Data Errors show on the Kore 1 Audio Interface, whenever the interface was connected to the computer, and ANY Sonar X3 or Sonar Platinum project was loaded.  Please note that these reported ASIO DATA ERRORS were still being capture and reported by the Kore 1 audio interface - even if the project in Sonar was NOT playing.  And this situation occurred whether or not it was X3e x64 or Platinum Ipswich x64.  Swapping out the USB cable for a brand new one made no difference, and swapping out the current USB 2.0 port for another USB  2.0 port also made no difference  and the Native Instruments Kore 1 Audio Interface reported ASIO Data Errors appeared with regularity still - even though the Sonar project(s) were not playing, simply loladed into Sonar Xe3 x64 or Sonar Platinum Ipswich x64).
 
Note: With the Native Instruments Kore 1 Audio Interface connected through USB 2.0 port to the computer and selected as the audio interface with ASIO driver Mode and Sample Rate of 48000 and 24-bit Recording Bit-Depth, to a brand new Sonar project from the Normal Template, the Diagnostics Tab on the Kore 1 interface began showing a small number of ASIO Data Errors, then a few seconds 2, then after some more time 3 3, and after a little - 4, etc after around 10-20 seconds it would add another.   Swapping out for a brand new USB cable made no difference - it would still continue to aggregate these errors, so they were still being produced.
 
So, I am about to suck down a couple of quite large cups of coffee, and will continue to test and document here, the same sort of testing to be performed in the same manner, with the exception that the Native Instruments Kor1 Audio Interface will be swapped out for a FocusRite Scarlet 2i2 USB 2.0 Audio Interface, and the same testing will be done again.
 
I will include testing results from the swapped out interface in an additional post to this thread.
 
Bob Bone
 

OK - I seem to have stabilized the goings on with the hangups and dropouts with this particular project on this particular computer.
 
I am not finished testing and trying to reduce things that I changed, but I thought I would at least indicate that it now seems to play back from start to finish with no dropouts or crashes or any of that.
 
I am, however, to bleary-brained to continue testing at the moment, so will take a big giant nap and see what that does.  :)
 
By the way, in addition to making sure all was patch current and all of that sort of thing, the other changes I have recently done were:
 
1. Added 8 GB memory (bringing total to 24 GB)
2. Cleaned up temporary files/folders all drives
3. Tested with Kore 1 audio interface
4. Switched to FocusRite Scarlet 2i2
5. Tested with all effects bypassed
6. Bunch of other tests various configurations
7. Switched USB Cable to interface
8. Switched USB 2.0 port for interface
9. Disabled FIOSTV entries in Software Devices
10. Disabled all HD audio entries in Sound, Video, and Game Controllers
11. Adjusted Registry Paramater for AMD as follows, to simulate core parking:
 
      AMD Core Parking Equivalent
      [HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Power\PowerSettings\54533251=82be-      4824-96c1-47b60b740d00\0cc5b647-c1df-4637-891a-dec35c318583]
     Attributes = dword: set to 00000000
     Reboot System
 
12. Enabled a single HD Audio to allow Windows sound and HDTV connection
 
 
The song has thus far played back 4 times completely through with no weirdness at all.  Tomorrow, I will add additional tasks, like adding new tracks, other edits, etc., to see if things still hold up.
 
Bob Bone
 

If your SSD is a Samsung you probably need a magician...

http://www.samsung.com/gl...support/downloads.html

Be sure that drive is not set up to index on. The indexing file can be large with default settings and shouldn't be put on an SSD anyway.

Because of a reinstall I had 300GB of index files on my data HDD that were no longer accessible. I had to move files and format the partition to get rid of it.