Head-Fi.org › Forums › Equipment Forums › Sound Science › What's the point in using ASIO for audio playback?
New Posts  All Forums:Forum Nav:

What's the point in using ASIO for audio playback?

post #1 of 32
9/15/09 at 10:33am
Thread Starter 
What's the point in using ASIO for audio playback?

I've compared directsound and ASIO using RMAA and this is what I get as result:






Differences are within measurement error range I guess.

Properties of the playback device:






And what's bit-perfect supposed to mean anyway?
Reply
post #2 of 32
9/15/09 at 12:06pm
Beats me. I don't know ASIO from a hole in the ground.

k
Reply
post #3 of 32
9/15/09 at 1:41pm
Sake of mind?
Reply
post #4 of 32
9/15/09 at 1:54pm
I guess if you can't hear the .3dbA difference in dynamic headroom than you must not be a true audiophile. If you can't hear the housefly fart in the back of the studio than I guess it's time to hang up those headphones and find a less expensive addiction(see hobby).
Reply
post #5 of 32
9/15/09 at 2:04pm
The chief virtue of ASIO playback was to bypass Kmixer in Windows XP. Can I assume from the semi-transparent border of your RMAA test results window that you are using Vista? If so, that should account for your findings.

ASIO is still very useful for I/O, owing to its low latency. For playback only in Vista, however, I don't believe it's a necessity.
Reply
post #6 of 32
9/15/09 at 2:06pm
I guess its very soundcard dependant (some older ones did force resampling). With all DSP features off, if your soundcard passes the signal through to DAC without any forced things like resampling, using ASIO might indeed be pointless and bit perfect mode is there just for the sake of peace of mind?

And apparently this was more of a problem with Windows XP because kernel mixer resampled all signals passing through it to 48khz and did crappy job at it. If I understood correctly, Vistas mixer does no resampling, only digital volume adjusting is done, so ASIO and WASAPI are almost unnecessary.


*edit* Sherwood beated me.
Reply
post #7 of 32
9/15/09 at 2:15pm
Thread Starter 
Quote:
Originally Posted by Sherwood View Post
The chief virtue of ASIO playback was to bypass Kmixer in Windows XP. Can I assume from the semi-transparent border of your RMAA test results window that you are using Vista? If so, that should account for your findings.

ASIO is still very useful for I/O, owing to its low latency. For playback only in Vista, however, I don't believe it's a necessity.
Your assumptions are correct, it's Windows "Vista done right" 7.
I know that it's very useful for recording, especially if you try to play along with something the low delay is a blessing. And that's what it was made for.

Quote:
Originally Posted by MaZa View Post
I guess its very soundcard dependant (some older ones did force resampling). With all DSP features off, if your soundcard passes the signal through to DAC without any forced things like resampling, using ASIO might indeed be pointless and bit perfect mode is there just for the sake of peace of mind?

And apparently this was more of a problem with Windows XP because kernel mixer resampled all signals passing through it to 48khz and did crappy job at it. If I understood correctly, Vistas mixer does no resampling, only digital volume adjusting is done, so ASIO and WASAPI are almost unnecessary.


*edit* Sherwood beated me.
I have two Windows XP machines around, one with a very old c-media onboard soundcard. I'll repeat the test maybe on both machines.
Didn't know that the XP mixer forces resampling, but I'll see that soon enough for myself I guess.
Reply
post #8 of 32
9/15/09 at 2:22pm
Recently I got the Lynx AES16 adapter and had it installed in my Vista PC. My friend also told me to use the ASIO driver for Foobar2000. But on Vista it was the WASPI plug-in, and it worked very straight forward. The music sounds very good !

It happened that I had to move the Lynx card from the Vista machine to another machine which is XP. I just installed the card and the driver (ASIO featured). The first song I listened to sounded so weird ! It liked putting my head in a water tank and have an FM radio playing at the bottom of the tank !!

Later, I realized that the audio output was DS: Lynx Port 1. It wasn't ASIO one ! So, I downloaded Foobar2000's ASIO plug-in and managed the installation. This time the music sounded as good as I heard from my Vista machine.

My conclusion is that the KMixer or Direct Sound is not bit-perfect !! It also must do resampling or something.
Reply
post #9 of 32
9/15/09 at 3:34pm
in certain conditions, DS can be bit-perfect(volume knobs set to the max etc etc)...the biggest diff is that Vista's resampler works in 32float, instead of 16int for XP..
Reply
post #10 of 32
9/15/09 at 4:35pm
Thread Starter 
Volume wasn't set to the max and it didn't resample anything.

And I officially hate the stupid term 'bit-perfect' until I get a proper definition.
Reply
post #11 of 32
9/15/09 at 4:43pm
I don't hear a difference between ds and asio. I'm in xp. RMAA for xp would be nice to see since everyone complains about xp ds.

Note: I set volume to max and use foobar.
Reply
post #12 of 32
9/15/09 at 6:43pm
Quote:
Originally Posted by xnor View Post
Volume wasn't set to the max and it didn't resample anything.

And I officially hate the stupid term 'bit-perfect' until I get a proper definition.
AFAIK bit perfect simply means "sans-resampling" (by the OS) wrt playback from a computer. It is only a problem on Windows XP, but it's unlikely to be audible - it would be very easy to do a blind test with foobar if I remember correctly.
Reply
post #13 of 32
9/15/09 at 7:47pm
If you want an answer from the pov of someone who thinks software can affect sound quality, read this website cMP² | Main / HomePage .
Reply
post #14 of 32
9/15/09 at 7:47pm
Quote:
Originally Posted by AtomikPi View Post
I don't hear a difference between ds and asio. I'm in xp. RMAA for xp would be nice to see since everyone complains about xp ds.

Note: I set volume to max and use foobar.
a friend of mine ran tests w/ his LynxTwo-B soundcard using the built-in loopback, and he was able to measure the XP KMixer distortion.

@xnor: if you didn't set the volume to the max and were using DS, your audio was being resampled(not 1:1 bit-perfect)...but Vista/7 use a much improved resampling engine anyway, yet if you were to make a bit-perfect test w/ a DTS-CD or HDCD...your amp wouldn't recognize the stream(not bit-perfect).

and anyway, why not just use WASAPI on W7? it's a weapon of choice...supported by Reclock/foobar/MC/mediamonkey etc etc. it's prolly one of the best features(in WaveRT mode) of the post-XP OS, together w/ HPET(much more accurate system timer, XP wasn't meant for realtime)
Reply
post #15 of 32
9/16/09 at 7:15am
Thread Starter 
Quote:
Originally Posted by haloxt View Post
If you want an answer from the pov of someone who thinks software can affect sound quality, read this website cMP² | Main / HomePage .
lol thanks for that link. This guy clearly is out of his mind.
Never seen so much misinformation on a single website.
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › What's the point in using ASIO for audio playback?