WASAPI - ASIO-Kernal which is better

I am using IFI ZEN, WASAPI - ASIO-Kernal which is better? Currently I am on kernal.
Any advantages of one over the other?

In theory, Kernel Streaming is superior, then ASIO then WASAPI.

You seem to have the same DAC as I do but Kernel Streaming does not work for me at all. Are you able to get the full range of resolutions and can you play DSD with KS?


I’ve tried ASIO and Kernel Streaming with my Singxer SDA-6 Pro and i think Kernel Streaming is the most open and spacious sounding method.

I just tried ASIO, yes I can get DSD to 256 as max. With Kernal I cant upsample to DSD.
If I set up DSD upsample in ASIO then switch to Kernal, it seems to stay on DSD with kernal.

I do notice that Kernal sox sounds much better than DSD upsample. More defined low end.

Does up sampling matter though? I am still trying to tell if I hear a difference.

That’s not true. "ASIO and WASAPI, like Kernal Streaming have direct access to the buffers. (Also, Microsoft abandoned Kernel Streaming long ago)

Kernal 100/100
Asio 95/100
WASAPI 90/100

1 Like

What are those numbers?

This is my rating
I listened to one song 50 times. Kernel is best.

This discussion is great but did anyone have any issues getting Kernel Streaming to work and managed to overcome them? It simply locks up my PC if I try to use it!

1 Like

The only experience I have with Kernel streaming is with Jplay and JRiver.
Didn’t do extensive testing with JRiver but felt kernel streaming was a bit better forme than Asio or Wasapi.
Jplay in kernel streaming definitely checked all boxes for me, best I ever heard but no comparison with the other 2 here in the same setup.

I wished I could test it in AS but I cannot even find my usb dac in the menu when chosing kernel streaming. Only the ethernet out is available. No pb with Asio or wasapi.

1 Like

Had this too with Schiit DACs, try playing a 16/44,1 Song only.

It seems to me, that these numbers match my experience too… but I can’t tell if its a placebo :wink: … can’t really A/B with my setup, because switching takes time and resets the playhead… My question would be: What about Apple’s Core Audio in comparasion?

Is certainly a matter of taste honestly.
I find Asio resolves a little better and is a little relaxed when listening and more refined.
Wasapi I found him in the treble range a bit sharper and not quite crisp in the midrange than Asio, Asio is a bit more subtle.

Can certainly change from Dac to Dac again I guess.
But basic tones are certainly somewhat different.

Honestly try both what one likes and also play times at the resolution.
Personally, I prefer 32/96, although DSD should be better I miss a bit.
But certainly also conditioned by preference.

The software offers relatively many possibilities to try it. Even Audirvana 3.5 is there enough to play and test.

1 Like

Hi Everybody.

I did not know there was a community. So, I am happy to be here.

After many tests, Kernal Streaming gives more details. ASIO is slightly more musical on some tracks, but it eats a little of the details. WASAPI is also nice, but I feel some artifacts like extra bass.

So, the best overall is kernel streaming. The question is what kernel optimization to use. Even though the recommendation is extreme, sometimes I hear harshness and shift back to standard or high. Now playing extreme, and it seems amazing—supper details.

I use Windows with some isolation connected to my Chord Mscaler through USB. MScaler is powered by a battery for isolation. Then the DAC is Chord Dave. The experience is amazing. You hear parts in the music that were hidden from you for years.

So, what is your say on the optimization of the kernel streaming? I would like also to know more about kernel streaming in general, like APIs, development, why it is different, and how? Why other companies do not do the same? Has anybody tested MacOS vs. Windows?



There are thousands of answers to your questions on the net and everyone tells a different story. Kernel streaming bypasses windows sound settings completely. For windows definitely the best option. The mac community is united in saying that mac is better than windows pc for audio. The windows community says that windows pc with optimizations, for example fidelizer, is just as good for audio as a mac. I tested both variants over years and can not tell any difference. Others told me that there is a difference - may be my ears are not audiophile…
Think you have to do your own choice. When it sounds good to you, then it sounds good.


Excellent writing, Steff. It is even confusing as the number of combinations for connections is massive, and each one sounds different (At least for an audiophile with sensitive hearing). The first conclusion is that PCM streaming to a DAC is not a mature technology like LAN communications, for example.

I tried so many configurations with different protocols and equipment. My key learning is that once your DAC is power isolated, you start to have consistency in the results. We call this controlled experience in research.

Regarding Windows vs. Mac, as Mac may sound slightly better out of the box, Windows has more potential. One basic configuration for Windows is to run with a fast processor clocked constantly at a high frequency so that Windows as OS does not take more than 5% of CPU at any time while playing music. This helps produce quicker and more reliable USB streaming. Other factors like power, brand of Motherboard, and USB port manufacturer are all important. I am not a fan of port isolators as they may create additions overhead/jitter.

All of the above is because computers are not designed for audio output, and in my opinion, most audio streamers in the market have the same issue explaining why each streamer brand sounds diferenet.

1 Like

WASAPI cuts out the terrible windows default sound addon, kernel streaming cuts out all software before your player except for the smallest part of the OS, but ASIO skips the kernel, and outputs directly to your hardware. It’s the purest output signal. Bliss.
Microsoft has to learn to use ASIO 4 Everything, not to sound like geeks.

WASAPI is the newest Windows driver, and ASIO is developed by Steinberg a long time ago. Do you know what ASIO do? Obviously you don’t. It just borrows Windows driver and makes it easier to manage for DAWs, like reducing buffer size so you can record with lower latency. ASIO4All is essentially a “driver wrapper” it packages the native Windows audio drivers to appear as an ASIO driver, thus enabling it to bypass the additional software layers and connect directly to the onboard soundcard or audio interface. There are some audio interfaces that have their own ASIO drivers. Usually establish brands.

1 Like