Hey everybody,
 
So I'm reading that Windows 10 has better audio.  Anything in particular I should be doing to get that improvements or is it behind the scenes?  I'm finding the ASIO4ALL drivers to work way better on my interface than the manufacturer's, but still curious about this.
 
Donny

doncolga
Hey everybody,
 
So I'm reading that Windows 10 has better audio.  Anything in particular I should be doing to get that improvements or is it behind the scenes?  I'm finding the ASIO4ALL drivers to work way better on my interface than the manufacturer's, but still curious about this.
 
Donny

What audio interface is this?

ASIO4ALL is just a WDM driver wrapper and is very good at conflicting with other ASIO drivers
In pretty much all cases you are better off fully uninstalling it and using the WDM driver if the latest ASIO driver and firmware for your interface does not do a great job.

ASIO is still the preferred driver, avoid ASIO4ALL.

This is the xr18. Should it be able to work with drivers other than asio?

MS is pushing hard to make WASAPI API performance on par with ASIO (which completely bypasses the Windows audio stack) and that's where the audio improvements can be found in 10. If you use ASIO then you're not really touching the code that's been overhauled so if you see improvements in performance it's likely die to overall streamilining of the OS itself. WASAPI performance in 10 is now only slightly behind ASIO and the effort will continue with new releases of the OS.
 
There's a completely new MIDI API that looks pretty good but apps have to write to it to get the benefits, but there have been improvements to the existing API as well so if you do much in the way of MIDI you will leverage those.
 
So to answer your question, if you're using ASIOForAll you're using the Windows audio stack so you should be seeing improved performance now.
 
Check out the video here to get the lowdown: http://pcaudiolabs.com/windows-10-for-pro-audio/

That's good to hear and very interesting. On asio4all my cpu usage dropped a lot, RT latency was 10 ms and Sonar felt steady on asio4all. I was happy with this on my current hardware which is 7 years old, but was curious if I could still improve.

Should the xr18 be able to work directly with wdm drivers directly? If yes and considering 10's improvements, should that be still better? Is it even correct to call wdm a driver? Seems like it's more like Mac core audio...Am I off base there?

In general you don't want to use ASIO unless you have software that just doesn't support WDM/KS and your hardware doesn't have good ASIO drivers. Since Sonar does support WDM/KS, I'd use that, or use the hardware's own ASIO drivers. As to which to use depends on the card. For pro cards I usually try ASIO first as they often have better ASIO drivers than they do WDM drivers.  For a consumer card, they probably only have WDM drivers so that is the way to go.
 
You can try both and see what works best for your setup. For me right now it is ASIO, since Focusrite has sub-par WDM support. For your interface I dunno, as I've never encountered one. Just have to play with it and see.
 
In general give ASIO4ALL a miss though, as it is just an ASIO to WDM bridge, which means that you can just skip the bridge part and talk WDM directly.

I've connected back my Multiface II just to experiment more with it and to compare with the XR18.  It behaves very well under ASIO and WDM and I will say that it feels more stable.  You can open and add plug ins easier while its running and its smoother.  ASIO feels stable at 8 ms round trip latency; CPU's working reasonably hard.  Figure that's not bad on my hardware.  WDM is reporting 5 ms latency...would that be round trip also?  Another observation...on the Sonar CPU meter, the first one is working alot more than the others?...that normal?

Be aware if you are testing without ASIO4ALL you need to fully remove it as it's likely to conflict .
Sounds like you want to be looking into loopback tests for actual roundtrip:
 
http://forum.cakewalk.com/Loopback-Test-for-Latency-m3051266.aspx
 
http://forum.cakewalk.com...r-offset-m3178396.aspx
 
First CPU core generally gets hit the most. Normal.
Make sure you windows advanced power settings have min/Max cpu set for 100.
Make sure Sonar preferences is set for a multiple CPU's.
 

Doktor Avalanche
Be aware if you are testing without ASIO4ALL you need to fully remove it as it's likely to conflict .
Sounds like you want to be looking into loopback tests for actual roundtrip:
 
http://forum.cakewalk.com/Loopback-Test-for-Latency-m3051266.aspx
 
http://forum.cakewalk.com...r-offset-m3178396.aspx
 
First CPU core generally gets hit the most. Normal.
Make sure you windows advanced power settings have min/Max cpu set for 100.
Make sure Sonar preferences is set for a multiple CPU's.
 

Yes I did take that off totally before I tested back with the Multiface.  I will reboot and test again to be sure and also double check those Windows and Sonar settings.  Thanks very much!