The disabling of the said HD Audio component as mentioned by Craig, actually frees one of the CPU cores from the so called "HD Audio work" hence the configured Windows System appears to have a lower latency.
 
The driver issue is mainly due to Windows Installers themselves. Since the beginning, Windows has always allowed itself to accept "any" additional component without minimum fuss... remembering the days of WinNT, plug in a new PCI card and Windows will display a message that a new device has been found. Then the drivers are installed via CD or Floppy.
 
FF to the days of "Plug and Play" - using generic devices work but without specific features as intended by the user. Again, drivers needed to be installed - however, this time it will be custom drivers being installed over generic. When WinME came out it was a major shambles perhaps the worse OS ever. With WinVista and beyond, clean installations became a lot easier... not really... just that most things are hidden. 

Even today, clean install a Windows computer and upon boot up, graphics, sound and other stuff are already working. Plug in a printer and some other stuff and they work... Going into the drivers area and you'll actually see the number of generic printer drivers installed by Windows is just madness.

It may appear that the competition is having better hardware especially in the pro audio World however, their software drivers have been designed to fit specific hardware to begin with. As for a purposefully built Windows system, a lot of time is dedicated to R&D on both the selected hardware and the sets of drivers to be installed or remove; hence the increased costs. Then on delivery to the customer, all he or she needs to do is hook up a power cord, power on and be good to go. Honestly as musicians and music people, we wouldnt want to spend out days tearing our hair out fretting over a new build or otherwise.

There are tools employed to select specific components before an install of Windows. However, to nitpick what goes in or out is seriously a nightmare in itself. With Win7 and Win8 its almost impossible because of the so called dependencies. Trust me.... i have given up. Rather i install everything MS wants and then work backwards.
 
For those interested, the keys here are "ks.inf" and "wdmaudio.inf". These files are always part of the code for every type of audio driver... there again there may be more....





  

johnnyV
Well Greg your discovery is very important for a lot of systems that will be crippled by that and it sure doesn't hurt to remind folks of this again and again. It is the best reason for this forum existing and all these little tips have been a great lifesaver for me.

So can you give me an opinion on whether this kind of convoluted fix is commonplace? Was I wrong to think that the context menu for the device in S, V, and G controllers should have a disable option, you'd select it, and carry on? It's the difficulty of disabling, and the fix for being able to do so, that I thought required detail above and beyond the usual "disable stuff you're not using."

keyzs
The disabling of the said HD Audio component as mentioned by Craig, actually frees one of the CPU cores from the so called "HD Audio work" hence the configured Windows System appears to have a lower latency.

 
Thanks for taking the time to write the informative post. You're one of the people who noticed the thread title posed a question
 
One engineer told me in private that nVidia drivers are notorious for being poorly written and hogging processes, and he wouldn't be surprised if AMD drivers were in the same ballpark. He had encountered one driver (I don't think it was nVidia) that actually stopped your hardware while it did its thing. I imagine that would have a negative effect on latency, LOL. He theorizes sloppy code that penalizes system performance is the culprit, and eliminating that process is what's providing the improvement.
 
That makes sense, because the system was perfectly capable of running at low latencies, but not consistently without having an eventual crash. So it seems logical to me that although the system could cope most of the time, sometimes the process would bog down the system so much it would just give up and crash. I assume increasing the latency covered up for those times when the system would otherwise be bogged down, but I don't know the nuts and bolts of how this stuff works...I'm experientially oriented.

johnnyV
Actually those custom builds, at least studio cats, ( not sure about the other one mentioned)  are not any more expensive than your typical off the shelf box. I've been shopping around and was blown away at what little you get in a HP, Dell, Lenovo etc pre built desktop. By the time you "build your own" with proper components your probably $$ over what the same Studio Cat's sell for.
 
 

sure johnnyV , I happen to agree with you 100% ….. and dare I say Mike ….
where only the best is good enough …that is the only way to go if you can swing it …
i have that end covered w my guitars , amps and pedals ….sure wish I could say that about my Windows computer .
 
regarding the general topic here 
I'm under the impression that SONAR itself is ahead of the loop as far as the hardware it runs on goes….
am I mistaken by thinking that it is the current hardware and layers of the windows OS that is holding up and creating the added latency issues ? it seems to me that the hardware needs to catch up to SONAR …Yes ? No ?
that appears to be what I'm getting from this discusion at this point 
anyway , I came to this topic to learn what i could , even tried out a few of the sugested configurations .
even on my off the shelf mid spec lap top …...there was a slightly noticeable difference…. 
 
Kenny
 

Anderton
 
Thanks for taking the time to write the informative post. You're one of the people who noticed the thread title posed a question
 

Hi Craig, thank you for your kind comments.
 
Like many of us here, all we request for is a system and platform which is stable enough to handle all our demands to produce the music which we so love. Along the way we learn much from others while we share what we can. Thank you again since the days of Ensoniq and Mr Mash ...
 
i did a little digging into this... apparently the HD Audio Controller not only releases the CPU Core as i once thought (post #61), there is also IRQ involved.  (Ref: https://social.technet.microsoft.com/Forums/windows/en-US/d8351bb8-2678-428c-bbb3-004c01f58cb7/microsoft-high-definition-audio-controller-conflict ) Looks like there is more than meets the eye here... 
 
i'll try to take this a little further whenever time permits and will keep this thread alive. On my current system, i have disable so much stuff i actually dont know where to being.  The most major of MAJOR issues that really bugs me is the way Windows integrates generic and legacy drivers during Windows first install. Things may be made worse when specific drivers and or updated drivers are installed over these generic ones.
 
In my most humble opinion, the key to improving Sonar's Audio Performance may be to redesign a brand new audio codec from scratch where it totally bypasses and or ignores anything which MS and ASIO have done in the past and or are doing either currently now without the baggage of legacy codex. To put it simply is to have Sonar "talk" directly to the Hardware Audio Interface bypassing even USB and or FW protocol. (That i dont know if its possible - a new protocol perhaps...???) (Another issue is how to bypass Windows' own media codecs...???)
 
The reason for this conclusion is that when ASIO first came out, it was not without issues, however it was the beginning of a dedicated audio streaming protocol that attempts to bypass the OS audio kernels, hence reducing full duplex audio delays. Although technology has improved tremendously since the first ASIO codec was released, the new and or old codes were never properly addressed to solve issues. Then came WASPI and WDM etc... These are just codes over codes which cover over even more codes.... Marketing has always dubbed these as new technologies... unfortunately not. Thats why there are still traces of legacy code to begin with (post #61)
 
The increase in speeds of modern CPUs and other hardware simply cover up of what's lurking below the actual code structures. These days, programmers will simply use a few lines of "new" code to cover up whats wrong rather than actually studying or modifying the original codes to get to the root of the problem. With today's hardware technologies, we should be recording and playing back at speeds close to the way transistors handle current, unfortunately we arent...
 
Personally i have used CWPA since the days of old and Twelve Tone will always have a special place in my heart. 
 
Sincerely i look forward to the day when Cakewalk will be the one DAW that rules them all . Where we can all proudly say if you arent using Cakewalk, you arent "industry standard"...  sounds familiar ha... 
 
cheers all 

Excellent stuff here!

Sanderxpander
Ugh I really envy OSX people sometimes. I hate parts of it, and really wouldn't want to switch away from Sonar, but the whole audio/midi part is just soooooo much better done.

My next os is Linux. I see no practical or moral benefits to Apple products.

fwiw, the HDMI audio on my lappy is labelled as "Intel(R) Display Audio" as it has an intel chipset, and "High Definition Audio Device" refers to the usual onboard soundcard

gswitz

Sanderxpander
Ugh I really envy OSX people sometimes. I hate parts of it, and really wouldn't want to switch away from Sonar, but the whole audio/midi part is just soooooo much better done.

My next os is Linux. I see no practical or moral benefits to Apple products.

but... but... don't they say something about your lifestyle? ;-)

The main practical benefits I see are Logic X, which is a great program, plus all the other software, hardware and plugins that I use that will work without any issue if I were to make the switch. I haven't really tried Linux seriously but it seems too tweaky/fiddly for a professional DAW.