WASAPI vs ASIO vs Windows Volume Control
Sep 5, 2018 at 12:01 AM Post #16 of 31

smodtactical

1000+ Head-Fier
Joined
Nov 11, 2012
Posts
1,309
Likes
553
Your description of the difference sounds more like placebo than an actual difference in sound reproduction. Perhaps if you use criteria that refer to specific aspects of sound, like response, dynamics, distortion, etc. A controlled comparison test would be good too.

Ok I took a short segment of well mastered flac to analyze. You can listen too if you wish. Its the first 15 seconds of Nicholas Gunn - Earth story (sacred gun). When it starts out the Indian call sound and with the subsequent bell like sounds what I really notice is the clarity of the vibrating echos. Its just clearly more distinct, audible and separated from the 'whooshing' wind sound that comes in over top. This is made even more clear after the wind/whoosh passes and you just hear the exaggerated string sound being plucked sound... the echo is just so nice and clear in wasapi where in DS its there but smeared and not as clear. The layering is also somewhat collapsed in DS probably contributing to the smearing.

The whoosh sound also is more nuanced where you can hear internal detail of it in wasapi whereas in DS it just feels very monotonal and doesn't let you hear the underlying indian call sound.

At around 20-25 seconds when the drum comes in, in wasapi you can hear a double or triple sort of beat with each drum stroke quite distinctly where in DS it sounds again smeared almost like 1 weird sounding stroke instead.

Everything as I said just sounds slightly more separated and thus detail is more easily picked out. Almost like turning up the 'contrast' or increasing the dynamics of the sound slightly.
 
Sep 5, 2018 at 8:03 AM Post #17 of 31

439598

100+ Head-Fier
Joined
Nov 25, 2015
Posts
332
Likes
76
Ok I took a short segment of well mastered flac to analyze. You can listen too if you wish. Its the first 15 seconds of Nicholas Gunn - Earth story (sacred gun). When it starts out the Indian call sound and with the subsequent bell like sounds what I really notice is the clarity of the vibrating echos. Its just clearly more distinct, audible and separated from the 'whooshing' wind sound that comes in over top. This is made even more clear after the wind/whoosh passes and you just hear the exaggerated string sound being plucked sound... the echo is just so nice and clear in wasapi where in DS its there but smeared and not as clear. The layering is also somewhat collapsed in DS probably contributing to the smearing.

The whoosh sound also is more nuanced where you can hear internal detail of it in wasapi whereas in DS it just feels very monotonal and doesn't let you hear the underlying indian call sound.

At around 20-25 seconds when the drum comes in, in wasapi you can hear a double or triple sort of beat with each drum stroke quite distinctly where in DS it sounds again smeared almost like 1 weird sounding stroke instead.

Everything as I said just sounds slightly more separated and thus detail is more easily picked out. Almost like turning up the 'contrast' or increasing the dynamics of the sound slightly.
acknowledging a difference here is the first step towards a new level of SQ on your PC, congrats.
if you are interested you can try out the free trial of HQ Player, the difference between HQ Player and foobar was far greater than DS and WASAPI for me.
just set PCM sample rate to the highest multiple of 44.1kHz that your DAC supports and set buffer time as low as possible without stuttering. This software is slow and awkward to use compared to foobar so im very reluctant to buy the full version but the sound is good

BTW do you have DS set to a different sample rate to the files you're playing? this will cause problems particularly if its not a multiple (48/96/192 kHz) and may explain the difference you are hearing. as explained already DS is not very good at resampling
 
Last edited:
Sep 5, 2018 at 8:46 AM Post #18 of 31

smodtactical

1000+ Head-Fier
Joined
Nov 11, 2012
Posts
1,309
Likes
553
acknowledging a difference here is the first step towards a new level of SQ on your PC, congrats.
if you are interested you can try out the free trial of HQ Player, the difference between HQ Player and foobar was far greater than DS and WASAPI for me.
just set PCM sample rate to the highest multiple of 44.1kHz that your DAC supports and set buffer time as low as possible without stuttering. This software is slow and awkward to use compared to foobar so im very reluctant to buy the full version but the sound is good

BTW do you have DS set to a different sample rate to the files you're playing? this will cause problems particularly if its not a multiple (48/96/192 kHz) and may explain the difference you are hearing. as explained already DS is not very good at resampling

Thanks I'll check it out! I have everything set to 192 khz in windows.
 
Sep 5, 2018 at 1:44 PM Post #19 of 31

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Posts
22,232
Likes
4,158
Location
Hollywood USA
At around 20-25 seconds when the drum comes in, in wasapi you can hear a double or triple sort of beat with each drum stroke quite distinctly where in DS it sounds again smeared almost like 1 weird sounding stroke instead.

Wow! That is time smearing on a massive scale. I'm not a PC guy, so I can't replicate what you have there. I was under the impression that digital audio had all the bugs worked out on PC. Perhaps that isn't the case. I'm happy with my Macs that do audio perfect out of the box.
 
Sep 5, 2018 at 11:30 PM Post #20 of 31

smodtactical

1000+ Head-Fier
Joined
Nov 11, 2012
Posts
1,309
Likes
553
Wow! That is time smearing on a massive scale. I'm not a PC guy, so I can't replicate what you have there. I was under the impression that digital audio had all the bugs worked out on PC. Perhaps that isn't the case. I'm happy with my Macs that do audio perfect out of the box.

Well I am a hardcore gamer and love VR... so a MAC is pretty useless for me. :wink:
 
Sep 5, 2018 at 11:35 PM Post #21 of 31

theveterans

Headphoneus Supremus
Joined
Sep 2, 2014
Posts
2,345
Likes
1,009
Location
Santa Rosa, Laguna, Philippines
Though I never pay much attention to ASIO / WASAPI vs DirectSound for sound quality, I definitely use them for a different purpose: I don't want any other audio stream ruining my audio playback so I use WASAPI + ASIO to separate my music stream (dedicated stream to my DAC) while all of the system sounds and all other programs on my default Intel HD audio sound (that is muted).
 
Sep 5, 2018 at 11:40 PM Post #22 of 31

smodtactical

1000+ Head-Fier
Joined
Nov 11, 2012
Posts
1,309
Likes
553
Though I never pay much attention to ASIO / WASAPI vs DirectSound for sound quality, I definitely use them for a different purpose: I don't want any other audio stream ruining my audio playback so I use WASAPI + ASIO to separate my music stream (dedicated stream to my DAC) while all of the system sounds and all other programs on my default Intel HD audio sound (that is muted).

A clear benefit of wasapi/asio.
 
Sep 5, 2018 at 11:51 PM Post #23 of 31

theveterans

Headphoneus Supremus
Joined
Sep 2, 2014
Posts
2,345
Likes
1,009
Location
Santa Rosa, Laguna, Philippines
A clear benefit of wasapi/asio.

Some programs such as VLC media player that don't have WASAPI and ASIO can simply reroute the audio to another device, your dedicated DAC for example (doesn't matter if the stream goes to the windows mixer since you can set that device's bit and sample rate to match the audio stream if you don't want resampling) and have the same benefit of uninterrupted audio stream as WASAPI or ASIO. Wish Spotify would implement this too since I can set my default audio to on-board audio (where all of the system sounds / other program sounds are being routed ) while Spotify plays on my DAC, giving it exclusive access to the DAC (goes through its own separate windows mixer instead of WASAPI or ASIO)

Untitled.jpg
 
Sep 6, 2018 at 12:21 AM Post #24 of 31

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Posts
22,232
Likes
4,158
Location
Hollywood USA
Well I am a hardcore gamer and love VR... so a MAC is pretty useless for me. :wink:

Yeah. Macs don't make great gaming computers. But the Mac Mini is a pretty sweet all in one media server. Are you looking to improve the sound of video games specifically or recorded music?
 
Sep 6, 2018 at 12:58 AM Post #25 of 31

smodtactical

1000+ Head-Fier
Joined
Nov 11, 2012
Posts
1,309
Likes
553
Yeah. Macs don't make great gaming computers. But the Mac Mini is a pretty sweet all in one media server. Are you looking to improve the sound of video games specifically or recorded music?

Not really looking to enhance video games, just music. I find in general high end amps and dacs don't make a massive difference in gaming... the pay off is really in the music. But not interested in buying a mac to facilitate this at all.
 
Sep 6, 2018 at 1:41 AM Post #26 of 31

gregorio

Headphoneus Supremus
Joined
Feb 14, 2008
Posts
4,058
Likes
2,162
Wow! That is time smearing on a massive scale. I'm not a PC guy, so I can't replicate what you have there. I was under the impression that digital audio had all the bugs worked out on PC. Perhaps that isn't the case. I'm happy with my Macs that do audio perfect out of the box.

In theory PC's have all the bugs worked with digital audio too, the problem isn't so much with the PC's themselves but with third party software and the combination of different software. The combination of the programs, the drivers, Windows itself and some of the settings users choose (v. high sample rates for example). The software architecture of Macs is more tightly controlled by Apple, simpler and therefore there's generally less potential for foul ups but it can still happen:

I had a client who requested an additional master made specifically for Youtube. When testing using Safari to stream the video from Youtube the mix sounded how I expected but using Firefox (for Mac) it didn't, there was quite a noticeable amount of what sounded rather similar to compression artefacts. Just to make sure it wasn't my ears playing tricks on me, I loop-back recorded the output from both Safari and Firefox and then Null tested them. Sure enough they didn't null well at all, the difference file containing data all the way up to around -3dBFS! I don't know exactly what Firefox was doing or why, it's possible there was some incorrect setting selected but if so it was buried beyond a cursory look through the settings. This was about 4 years ago and Firefox has had numerous updates since then, so it's possible that issue no longer exists.

G
 
Sep 6, 2018 at 12:52 PM Post #27 of 31

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Posts
22,232
Likes
4,158
Location
Hollywood USA
That's interesting. I don't do programming myself, but I think with the Mac most software throws it back to the machine to handle audio. I'm curious why that would happen. Perhaps it was applying additional buffering and compression. Was there a longer delay before the file started playing? It might have been a bit of code that didn't apply to Macs that got inadvertently ported over to the Mac version. I'll do some googling on that.

Edit: I just googled and found people saying that having Hardware Acceleration turned on in Firefox could mute the audio in certain cases. That might have been interfering. I also found that Firefox has built in playback for Ogg, WebM and WAV, but it uses the platform for other varieties of compressed audio.
 
Last edited:
Sep 7, 2018 at 3:31 AM Post #28 of 31

gregorio

Headphoneus Supremus
Joined
Feb 14, 2008
Posts
4,058
Likes
2,162
That's interesting. I don't do programming myself, but I think with the Mac most software throws it back to the machine to handle audio. I'm curious why that would happen. Perhaps it was applying additional buffering and compression. Was there a longer delay before the file started playing?

Edit: I just googled and found people saying that having Hardware Acceleration turned on in Firefox could mute the audio in certain cases. That might have been interfering. I also found that Firefox has built in playback for Ogg, WebM and WAV, but it uses the platform for other varieties of compressed audio.

With a Mac, generally the audio is passed to the OS's CoreAudio but of course the player (within the browser) can alter the audio before sending it CoreAudio, volume being the most obvious example. Buffering would not make a difference, obviously I time aligned the loop-back recordings before null testing. It could be that Firefox was adding some compression but if so I couldn't find a setting for it in the user settings. There is a way to get much deeper into Firefox's settings but I didn't have time to get into that at the time.

I believe I checked both with Hardware Acceleration enabled and disabled, although it was about 4 years ago and I don't remember specifically everything I checked. Firefox was my default browser, so when I checked the Youtube stream it was initially with Firefox. Hearing that the master wasn't as I'd intended, I went through all the settings I could find/think of which might have some affect. Nothing made any difference and I was just about to go back and alter the master when I decided to check with Safari just in case, even though I assumed it would be the same. Hence why I did the null test, even though I heard the difference I didn't believe there should be one. I should mention that it was an obvious difference to me but then I was very finely tuned to what had been a rather tricky and time consuming mastering job. I believe it would've had an effect on critical listeners and probably even to a small degree on casual listeners but it wasn't some massive night/day difference. The result left me with a dilemma, should I change the master for Firefox or leave it so it sounded how I wanted when played with Safari? And, what about Firefox, Chrome and other browsers on PC? In the end there simply wasn't time to check what would happen with every browser and the artist and I decided to leave the master as it was.

G
 
Last edited:
Sep 7, 2018 at 12:17 PM Post #29 of 31

bigshot

Headphoneus Supremus
Joined
Nov 16, 2004
Posts
22,232
Likes
4,158
Location
Hollywood USA
Seems like it was some sort of bug in Firefox. They've probably fixed it by now. Interesting!
 
Jan 13, 2020 at 12:31 PM Post #30 of 31

ForSerious

New Head-Fier
Joined
Apr 1, 2015
Posts
38
Likes
11
Location
Virginia USA
Wow! That is time smearing on a massive scale.

I'm not sure if this is common knowledge but sometime ago, Windows (Both 10 and 7) added a feature aimed at "improving" sound for laptop speakers and cheap earbuds (in my opinion). Anyway, by default in the sound device configuration, the full range speaker option is not checked making everything sound different (worse). To make matters worse, this option periodically will get reset to the default value with random updates.

I wonder if that option was left to default in smodtactica's tests.

Either way. You can consider yourself lucky to have not been an unknowing recipient of such executive decisions. I would not be surprised to learn that over 90% of people who use Windows and listen to music don't even know that that option exists.

I'm not a PC guy... I'm happy with my Macs...
How are Macs not in the category of personal computer? Haha just kidding. That's a whole nother debate for another forum. We can probably blame Apple's advertising campaign from 2006 and on for that confusion.
 

Users who are viewing this thread

Top