ASIO, WASAPI, Direct Sound... is there any difference in sound quality?

[daftcombo]

Hi,

Some people say that ASIO gives the best sound, some WASAPI, some say Direct Sound is as good as it can get.
What's the truth?

I'm on Windows 10 and can't hear any difference, if I use my Topping D10 with Direct Sound or with its ASIO driver.

Are there any measurements out there?

Cheers.

 

[PierreV]

As long as the correct data reaches the DAC in a reasonable time frame, there will be exactly zero differences.

It's not really about SQ, but more about what the system sends (hidden resampling by windows for the non-technical user, non support of some data rates, etc), how often it is interrupted/disrupted (latency issues). That's more of an "IT" issue - people not actually being able to see what their system is doing at a "low" level - than an "audiophile" issue.

Early implementations of USB, with all their quirks, bandwidth-sharing, power supply sharing, clocking and, electrical issues did not help the general perception though...

 

[bennetng]

Hi,

Some people say that ASIO gives the best sound, some WASAPI, some say Direct Sound is as good as it can get.
What's the truth?

I'm on Windows 10 and can't hear any difference, if I use my Topping D10 with Direct Sound or with its ASIO driver.

Are there any measurements out there?

Cheers.

https://www.audiosciencereview.com/...ork-resolved-now-some-results.585/post-157095
https://www.audiosciencereview.com/...of-rme-adi-2-dac-with-windows.4072/post-95945
https://www.audiosciencereview.com/...lity-in-windows-using-wasapi.5272/post-116878

 

[Tup3x]

Hi,

Some people say that ASIO gives the best sound, some WASAPI, some say Direct Sound is as good as it can get.
What's the truth?

I'm on Windows 10 and can't hear any difference, if I use my Topping D10 with Direct Sound or with its ASIO driver.

Are there any measurements out there?

Cheers.

ASIO is meant for low latency playback so I wouldn't be surprised if some implementations sacrifice quality over latency. WASAPI can be bit perfect and even in shared mode it uses high quality resampler. Generally speaking though, when it comes to pure quality, it can't get any better than what exclusive WASAPI can offer. Not entirely sure how Direct Sound is handled in Windows 10 but it's either equal to WASAPI shared mode or worse.

 

[daftcombo]

Has anyone any measurements?

 

[bennetng]

Has anyone any measurements?

My #3 reply has linked to some measurements. Please read them.

 

[daftcombo]

My #3 reply has linked to some measurements. Please read them.

I saw measurements for the match Foobar VS HDPlayer in your first link (showing that HQPlayer measured better) but no charts & graphs about ASIO VS WASAPI VS DirectSound.

 

[bennetng]

I saw measurements for the match Foobar VS HDPlayer in your first link (showing that HQPlayer measured better) but no charts & graphs about ASIO VS WASAPI VS DirectSound.

Did you see my videos in my link? Watch the first video very carefully. I compared DS and WASAPI in the video.
https://www.audiosciencereview.com/...-of-software-volume-control.5922/#post-132450

Also read the link from Microsoft I posted:
https://blogs.msdn.microsoft.com/ma...audiometerinformation-and-full-scale-signals/

And read every link I posted in this thread. One of them contains official information from Microsoft as well.
https://www.audiosciencereview.com/...of-rme-adi-2-dac-with-windows.4072/post-95945

In general, don't generalize things and assume one API is always better or worse than the others. It depends on many factors. ASIO drivers, unless the generic ones like ASIO4all, are distributed by individual hardware vendors, you can't assume they work all the same. Some devices (especially pro audio interfaces) have additional onboard DSP and FPGA and they also require drivers to work, the drivers can override WASAPI behaviour as well.

So asking which API is better without looking into individual product is pretty meaningless.

 

[FrantzM]

Hi

I eternalized on these issues of bit perfect and ASIO vs WASAPI vs DS until I realized I and several audiophile friends with High Resolving systems (audiophile-speak for expensive) weren't able to reliably discern VBS mp3 from uncompressed ... on most music.

 

[Odysseus]

So, is there any point in installing the ASIO drivers for, say, SMSL SU-8 if the WASAPI exclusive (using MusicBee) works fine without installing any drivers?

 

[noel_fs]

So, is there any point in installing the ASIO drivers for, say, SMSL SU-8 if the WASAPI exclusive (using MusicBee) works fine without installing any drivers?

If the drivers are native from manufacter you get an smoother usage. Sound quality? Probably depends but i would say that asio is usually silghtly better to the point of almost not being audible

 

[Odysseus]

If the drivers are native from manufacter you get an smoother usage. Sound quality? Probably depends but i would say that asio is usually silghtly better to the point of almost not being audible

But if ASIO and WASAPI exclusive are both bit perfect, why would it make any difference to the sound quality?
(I understand that ASIO has lower latency but this is relevant to live performance rather than music listening)

 

[noel_fs]

if they were both bit perfect there wont be a difference but i guess they arent from my experience. Even a player sounds different using same output wasapi/asio

 

[Pluto]

if ASIO and WASAPI exclusive are both bit perfect, why would it make any difference to the sound quality?

It would not. The quest for ASIO to have as low a latency as possible (which is its primary purpose) might, in some circumstances, cause the playback to be less stable than is desirable but any problems in this respect are not subtle.

It's worth noting that most standard devices can be driven in WASAPI mode by most Windows machines without much ado as WASAPI is part of the OS (and is mature, having been around since Windows 7) and nobody seriously disputes its stability or accuracy. All that is required is that the application supports WASAPI which, of course, most decent audio applications do.

The provision of ASIO is down to the audio hardware manufacturer and, as such, is a bit more subject to the quality of the implementation.

 

[noel_fs]

It would not. The quest for ASIO to have as low a latency as possible (which is its primary purpose) might, in some circumstances, cause the playback to be less stable than is desirable but any problems in this respect are not subtle.

It's worth noting that most standard devices can be driven in WASAPI mode by most Windows machines without much ado as WASAPI is part of the OS (and is mature, having been around since Windows 7) and nobody seriously disputes its stability or accuracy. All that is required is that the application supports WASAPI which, of course, most decent audio applications do.

The provision of ASIO is down to the audio hardware manufacturer and, as such, is a bit more subject to the quality of the implementation.

XD
Wasapi has been around since windows vista release

thats all im gonna say

 

[Odysseus]

if they were both bit perfect there wont be a difference but i guess they arent from my experience. Even a player sounds different using same output wasapi/asio

What do you mean? If I play an mp3/flac/wav/DSD file with Foobar or MusicBee and send it using WASAPI exclusive to my external DAC, I should hear exactly the same output. They are both sending the same digital signal to the DAC.

 

[Odysseus]

It would not. The quest for ASIO to have as low a latency as possible (which is its primary purpose) might, in some circumstances, cause the playback to be less stable than is desirable but any problems in this respect are not subtle.

It's worth noting that most standard devices can be driven in WASAPI mode by most Windows machines without much ado as WASAPI is part of the OS (and is mature, having been around since Windows 7) and nobody seriously disputes its stability or accuracy. All that is required is that the application supports WASAPI which, of course, most decent audio applications do.

The provision of ASIO is down to the audio hardware manufacturer and, as such, is a bit more subject to the quality of the implementation.

That's my understanding too. ASIO is low latency so it is good to use for live performances; if there are any problems with the ASIO implementation in that case it will be more than obvious, not just an issue of quality. WASAPI exclusive should send the same bits to the DAC as a good ASIO implementation. In any case, WASAPI exclusive works, ASIO will work if the implementation is good, otherwise the ASIO connection will break down and music will stutter.

 

[noel_fs]

What do you mean? If I play an mp3/flac/wav/DSD file with Foobar or MusicBee and send it using WASAPI exclusive to my external DAC, I should hear exactly the same output. They are both sending the same digital signal to the DAC.

Thats the theory we all know

 

[Odysseus]

Thats the theory we all know

How can a player change the bits it's just transmitting?

 

[JohnYang1997]

from measurement
wasapi and asio same
direct shit
audible？ barely