What's the difference between Directsound and waveout?

[trains are bad]

In foobar2000, can someone please explain the differences?

 

[imzjustplayin]

It's really a different way to produce the sound. Direct sound is part of the microsoft API Direct X and is essentially a universal way for all devices to work more properly with the windows operating system. By using Direct Sound, you're more likely to be guaranteed an audio connection because of the fact that DirectX creates a universal place for multimedia devices to consult to in order to do what they do like synth MIDI, Play audio or video. This is assuming of course that your device is direct X compatible and if it's not, you won't get anysound. Personally, I use kernel streaming when ever I can, one benefit I see with it is the fact that adjusting the "WAV" option in sound controls doesn't change how loud it is hence the option of not being "wav".

 

[MikoLayer]

I have noticed Directosound consistently has lower cpu utilization compared to wavout. Then again, I have seen some complaints here about direcsound sounding bright. Personally, I think directsound is just fine for dirt cheap non-critical listening environment. On my secondary rig with AV-710, kernel streaming is my ouput method of choice. The only problem is if something else tries to access the soundcard while music is playing, it stops the music

 

[Revliskciuq]

If you have an ASIO capable device, then you should use ASIO over kernal streaming.

Alot of the time you'll hear us preaching about "bit-pefect" output through kernal streaming or ASIO, and while there is a difference, it's nothing to get bent out of shape about. Waveout always works, and doesn't sound bad by any means.

 

[MikoLayer]

Quote:

Originally Posted by Revliskciuq If you have an ASIO capable device, then you should use ASIO over kernal streaming.

and why is that? I thought latency really doesnt matter for us music listening crowd. If its that KS's inability to handle other sound creeping into the stream and stopping the music, then it does make sense tho

 

[eboomer]

Quote:

Originally Posted by MikoLayer and why is that? I thought latency really doesnt matter for us music listening crowd. If its that KS's inability to handle other sound creeping into the stream and stopping the music, then it does make sense tho

I think the only advantages are such user-friendlyness related ones. As I understand it, kernal streaming is a newer streaming method and still has some bugs to be worked out.

 

[Revliskciuq]

Quote:

Originally Posted by MikoLayer If its that KS's inability to handle other sound creeping into the stream and stopping the music, then it does make sense tho

You got it. Both output methods are bit-perfect, but generally speaking I find that if you have an ASIO capable device it's better to use ASIO over kernel streaming, which isn't as well implemented and causes problems like you described.

 

[Asmo]

Quote:

Originally Posted by MikoLayer and why is that? I thought latency really doesnt matter for us music listening crowd. If its that KS's inability to handle other sound creeping into the stream and stopping the music, then it does make sense tho

Simple, ASIO is a well known standard, kernel streaming is basically a hack.

 

[thomaspf]

Hm, I never thought of ASIO until win2000 introduced kmixer.

kmixer is gone from Vista and that will probably mean that ASIO willbe gone form my machine as well.

Cheers

Thomas