Sonar X2 experiencing a metronomic "pulse" when playback of softsynths or recorded audio

Ok, well the subject line expresses the problem in a nutshell. But first, if I have placed this message in the wrong forum would the moderator please move it?  This seemed to be the best fit for this message.
 
I've been experiencing a problem that is a rather hard "pulse" which causes the audio to drop out during the pulse duration, and this pulse is occurring at a regular metronomic rate of sometimes 120 beats per minute and sometimes 160 beats per minute. So far I haven't experienced it occurring at a rate other than these two.
 
My DAW is a home-built system, running a 3.4GHz AMD processor, 8 GB of RAM with a couple of Terabytes of drive storage.  Audio is handled and/or processed using an M-Audio Delta 66 PCI sound card with an OMNI box that attaches to the Delta via a 15-pin cable. The OMNI box amounts to a very full featured 4-channel mixer. Currently I have a guitar hooked up to IN 2, which is preamped, a Mackie mixer hooked up to INs 3 and 4, and I have audio out running from OUTs 1 and 2 to a set of powered studio monitor speakers. There is a headphones out jack that I have my computer's powered speakers hooked up to as well. I use that jack simply because it's easier to use than the monitor OUTs the OMNI has located at the rear of the unit, which would require that I put together a Y-cable to hook up the speakers. The Mackie mixer is not being used at this time.
 
When I first noticed this problem, I did not have my guitar running through the OMNI box. I was using an Alesis iO2 USB audio-digital interface.  At first I blamed the iO2, but later discovered that the problem existed without the guitar even being plugged in.
 
I mention the hardware just so you know. I don't think any of the above mentioned hardware is part of the problem, though. Mostly because this problem began about a week ago, and I've been using Sonar X2 for a variety of applications involving the soft synths and for audio for some months now,  and prior to about a week ago, everything was working fine. At first I was puzzled why the soft synth plug-ins and the audio were sharing this problem, and then I realized that, for the soft synths to work, I had to check the "First Synth Audio Out" box. So they had become audio as well.
 
I have some other music production software installed on this PC, including Band in a Box, Audacity, and an older copy of Reason. If I load the very same MIDI file into any of these products, and then configure the synths they use, some of which are plug-ins and some not, I don't get this pulse. (except Audacity, I can't figure out how to get midi files to play in Audacity) If I try to record audio in any of these programs, I don't get this pulse (including Audacity).  It's just within Sonar that this is occurring . . .  or so I thought until a short while ago. I've been struggling with trying to get Amplitube v3 to work with my system and finally got it to work just earlier this evening. And I'll be dinged if it isn't doing the same thing as Sonar.
 
In this case, it was the stand-alone version of Amplitube I was running, not the plug-in that I can load into Sonar. So, while I wouldn't have been surprised to find the Amplitube plug-in to have this problem, I guess I am a bit surprised that it is happening with the stand-alone version.  Just in case you aren't familiar with Amplitube, it is amp modelling software for guitarists.
 
So, have any of you run into a problem like this before?  Got any ideas?  I just want things to get back to where they were before last week. I'm not able to get any work done.

Thanks for your reply, Kalle.
 
Yes, I was using Amplitube in stand-alone mode. It was the only music-related application resident at the time I was using it.
 

Please confirm: Does the pulse get recorded? Does it show also in the audio waveform, or is the waveform normal, but output is pulsating? Is it equal in all output channels (loudspeakers, headphones), similar during recording and playback?
Does the pulse occur on "plain", single audio tracks with no copied double tracks or such that could cause phase cancelling?

 
Those are excellent questions and I had to check to confirm. No, the pulse is not being recorded. It appears in playback. Which makes sense, when you think about it, else, why would it also be showing up in MIDI playback when a plug-in synth is used (which is set for "First Synth Audio Out")?
 
Yes, the pulse is occurring on a plain, simple audio track with nothing else loaded. I specifically set up a test audio file with a single audio track, in which I then recorded my guitar, with no effects in the signal path. The quality of the audio between the pulses is excellent. But the pulses in the playback render it unusable.
 
As for routing and the output channels, well, this Delta-66 card comes with its own software mixer. 1 and 2 OUT run to a stereo system that powers my studio monitors. The 1 through 4 direct OUTs actually bypass the OMNI box, feeding the signal from the Delta 66 directly to whatever is connected to the OUTs. To hear the audio, I have to set the software mixer's output sources to either "mixer" or "sw rtn" (software return). The volume is slightly greater when I set the output source to sw rtn. Either way, however, I hear the pulse. This is true whether I'm listening to an audio recording or a "project" where I've set up plug-in soft synths to handle the midi voices. Again, if I leave the midi alone -- defaulted to the MS Wavetable Synth -- there is no pulse. Just when the synths are assigned to "first audio."
 

First thing I would try is record without the extras (Omnibox, Mackie) directly to Delta. You are using ASIO drivers?

 
Yes, I'm using the ASIO4ALL v2 drivers. Just now, I tried something else. With the MIDI project, I went into Edit>Preferences>Audio>Playback and Recording and changed the drivers to MME (32bit). I double checked the drivers in Devices, clicked on Apply, then closed it. When I hit Play, it sounded awful and I realized the buffers were set too low, so I went back in and increased the buffer sizes, then tried again. Wow. No more pulse! So next, I loaded in an audio project with these same MME drivers loaded and tried the audio. No such luck. Still the pulse.
 
So now I find this interesting. With the ASIO drivers, both the plug-in synths in the MIDI project, and the audio only file are exhibiting the pulse, whereas with the MME drivers (and the same soft synths active in the MIDI project), only the audio file is exhibiting the pulse. I don't understand this.  What, perhaps MME drivers don't see the soft synths as using audio, even though I had to click on "First Synth Audio Out" for them to be heard?  Certainly the MIDI project doesn't sound any different, using the MME drivers as opposed to the ASIO ones, once I dialed in a sufficient number of buffers.
 

If no FX whatsoever is included, one suspect to me would be the routing. Could there be phase cancelling caused somehow
with those particular tempi or something? Like the signal going two routes and clashing at output causing phase problem at only some tempi.

 
Yes, I would also suspect routing. But given that this problem first appeared when I was running my guitar through the Alesis iO2 which was connected to a USB port, and then later it appeared when I ran my guitar into 1 IN on the OMNI box, it causes me to believe that the source of the problem lies elsewhere. I used two entirely different paths, yet got the same results with each.  And please recall what I wrote in my first post -- how, when I first discovered this problem, by unplugging and plugging back in the iO2 I was getting shorter and shorter times before the signal became corrupted. That in itself is strange behavior to me.
 
Well, I tried an easy way to rule out the Delta 66 and/or its OMNI box and/or routing to/from/within the OMNI box being the problem. I also have an old Sound Blaster Audigy card installed in the system. I've had it deactivated until just now. After activating it, I set Sonar up to use its drivers and then I had to run the cable for the speakers to the SB card so I could hear anything. I tried the audo file just now, and there is no change. With the SB card, I'm still hearing that hard pulse.  So I think I've pretty much ruled out any external hardware as a possible cause (and internal, pertaining specifically to the Delta PCI card).
 
I don't recall making any changes, software-wise that could be causing this. Especially, I must remind myself, since it is not restricted to Sonar. Amplitube is "infected" also, whereas several other apps on this machine are not.
 
I'm trying something else right now. I went to Windows' uninstall utility and I'm looking at the software by date installed.  There area a couple of pieces of software that were installed around the time I first noticed this problem: a couple of toolbar apps that I don't recall giving permission for them to be there. This is really annoying. The first app won't uninstall. The uninstaller window has been just sitting there, saying "preparing to remove" for the past 15 minutes. Time to get physical.
 
I'll check back in once I've removed some suspect pieces of software, see if that helped any. I'm also gonna pull the Soundblaster card from the case, since I'm not using it, and uninstall all its related software.
 

But I've been using ASIO4ALL for months, and it's just the past week or a bit more now that I'm having this problem.
 
I tell you what, it wasn't an easy process getting rid of those two pesky toolbar apps. They were installed by an entity called Spigot, which is apparently a legit concern. Spigot installs these apps for other companies. And it makes sure that they can't be uninstalled. They don't show up in the Windows "Programs and Features" place where you normally do the uninstalls. The apps don't appear in the Start menu, so they don't have uninstall apps associated with them, and you can't easily find their folders because they're hidden. It took me a couple of hours of gloves-off wrangling to get rid of most of the crap and the few odd pieces that refused to be deleted? You know how when you try to delete a file, sometimes you'll get a popup telling you that it's in use by another application and won't let you delete it?  The few pesky files that were left that refused to be deleted generated popups that said they were being used by another application -- and those other applications were themselves!  I was feeling pretty ornery by that point, found out that I might not be able to delete 'em but I could rename 'em, so I renamed them to long strings of expletives.  Made me feel better and I reckon they've been rendered harmless.
 
Sadly, however, all that adventure was for naught, when it comes to my pulse problem. It's still there. Not calling it quits yet, not by a long shot, I tried a different OS. Yeah, this machine dual boots between Win7 and XP. I kept an XP partition on the machine because I have a few pieces of legacy hardware that I still use that Win7 doesn't support.  I have Sonar X1 set up in the XP partition.  And I've got XP drivers for the M-Audio Delta card as well. So I went to trying things out and, how about that -- neither MIDI files with their plug-ins nor audio files exhibited the pulse.  Regarding audio, I was having a lot of trouble with the WDM drivers (I think that's what they were called?).  They have no headroom at all. If the signal even touches the red area in the VU meters, the audio emits a sharp crackling buzz -- the worst sort of clipping imaginable.  Figuring the ASIO drivers would have more headroom, I set X1 up with them instead and could not get any response. No matter what I did, I could not get the program to recognize any of the ports.  Now that was frustrating.  But it gave me an idea -- I wasn't holding out a lot of hope, but it was something I hadn't tried yet. I hadn't tried Sonar X1 in Win7 to see if the pulse problem would occur with it. So back I went, this time just to test audio. Nope. X1 exhibits the pulse as well.
 
It's getting late. Tomorrow, I'm gonna start paring away at things -- removing anything that has electricity running through it from its vicinity. I'll do this item by item so that if there is a culprit, I'll know which it is.