SOLVED: First project with Win 8 - i7 machine, pops?

Hi,

 I've been using Cakewalk / Sonar since Cakewalk for DOS 2.0 (1988) and have tweaked my various machines along the way for optimum DAW use. I just installed Sonar Producer 8.5.3 (64 bit) on a new i7 machine with Win8-64. I played with a project this weekend that had a couple of audio tracks and Session drummer midi/audio tracks. And also had one instrument playing in Kontakt 3.5 (64 bit). With ASIO buffers at 128, CPU was at 4% and Disk Usage was at zero (below 1%), but yet I still heard random pops/clicks during playback. 

 I moved my ASIO buffers from 128 to 256 to 1024. It seemed to get better at 1024, but I don't understand why I experienced any pops with the CPU only at 4%. I'm going to next try individually turning off the few plugins I was using, and thought that there may be some compatibility issues with various plug-ins and 64 bit, or vs. buffer sizes.

 I know that the mastering plugins like LP-64 stuff and Perfectspace need more buffers, but I wasn't using them. Also have heard of bit-bridge issues, but this is my first experience with a 64 bit OS.
I was using: Kontact 3.5 (64 bit); Session Drummer; a Sonitus Reverb; Tube Leveler; Vintage Channel 64. 


 So a few questions:
 - does this list have any potential issues at 64-bit, 128 buffers? 
 - Is there a master list in a forum sticky somewhere showing buffer requirements per plugin? 
 - Other thoughts on pop/click gremlins in Windows 8 - 64? 


 Also, I have set the windows priority to background processing and disabled antivirus during Sonar use. 


 Thanks, 
     Dave 


 Machine: Windows 8 64; i7-3770; 8G RAM; two 1TB hard drives (one for SW & OS; one for projects and their audio data); Audiofire 8; Radeon 1GB 7570 graphic card; Wireless disabled; Wired USB mouse and keyboard.

Kristoffer,

Thanks.  Two good questions.  I will run LatencyMon tonight and see what I can learn.  Haven't done that before on any machine, the last PC I used was set-up around 8-9 years ago.

I've read about the firewire legacy drivers, but also have read threads indicating it may have been a short term issue and that newer drivers / Windows 8 might not have the problem.  The latest Echo Audiofire driver's readme files don't mention rolling back.

Echo also has a firewire analyser tool that I still need to run too.  Should have done those things before posting.  I'll post the results once I learn more...

Also, seeing some threads that indicate maybe VC-64 needs some minimum number of buffers (still haven't found any exact requirements though). 

Dave 

LatencyMon helped tremendously.  My system was sitting around 3000us when I first fired it up.  I quickly learned that the WiFi driver was the main culprit and since I don't plan to use it, keeping it disabled is no issue.  Doing that dropped my peaks down to around 300us.  Also interesting that disabling the WiFi via network adapters in control panel didn't solve it, had to disable via device manager.

Then I learned that disabling the wired ethernet connection dropped the system down to only around 30-60us consistently. 

Those were my main problems.  However, I still thought I was hearing a few pops now and then.  Maybe they are something I was imagining and maybe they were real, very subtle.

 I turned off everything in my Sonar project except a single Kontakt 3.5 instrument (congas), and found out that it still has some little random clicks. Perhaps they won't be there when I render, I'll play with it more.  I also thought it may be best to download the latest Kontakt free player (ver 5) and use it with my full version 3.5 library.  I'll try that later.  I've never used Kontakt for anything other than a player in Sonar anyway.

Guess I'm still not 100% sure that Kontakt itself is the remaining issue, need more time with this new machine and some headphones.  I haven't played enough yet with other tracks/synths to totally convince myself that everything else is fine. 

I ran LatencyMon in the background while doing that isolation exercise and the peak was still only 60us, this while the Sonar reported CPU stayed below 4% and disk usage was <1%.  So the cause of those Kontakt clicks is likely not do to any latency issues, but something completely different (?).
  
Thanks Kristoffer for pointing me to the latency tools.


Dave

That's true about trying WDM drivers. Calkwalk recommends this in the knowledge base information on audio drivers. Lots of people have better luck with them it would seem. 
But the other thing to look into will be your firewire set up. There seems to be a lot of threads and they all point at having a Texas Instrument Firewire chipset for some reason. 

Ya it's pretty interesting playing with the DPCLAT meter. For me on my Laptop the battery management cuts it way down. WI Fi always is a biggy. down under 50ms is about as good as it can get so that's no longer your issue. 

 Audio Interface drivers are the next thing to look into. 

Thanks everyone.

My ASIO buffer setting was 128 (with 2.9ms latency as a result) in the initial testing.  I've cranked it up to 256 buffers (5.8ms latency) with no change in the random clicks.

I'll check and see if WDM performs better and post the results.  Will also revert back to program priority (thanks for that pointer scook).

Since the clicks were heard with only one Kontakt track running, there were no recordings involved other than the original sample recordings.  But the clicks are random, and not repeatable, so that rules out anything in the recorded domain.

This weekend will involve more detailed listening, will post results.

Dave

TI FW is a must for DAW :)