or Connect
Head-Fi.org › Forums › Equipment Forums › Computer Audio › ASIO vs. Kernel Streaming
New Posts  All Forums:Forum Nav:

ASIO vs. Kernel Streaming - Page 2

post #16 of 57
As Glassman pointed out above with buffer sizes of 1024 or 2048 samlples the Juli@ works pop free for any 16/44.1-48 or even 24/96 content you play.

If you still see issues you might want to look at the PCI delays for the other components in your machine. I had some unrelated studdering issues and found that reducing the PCI delay of my Radeon card to 96 cycles helped making ASIO playback very solid.

Cheers

Thomas
post #17 of 57
Thread Starter 
Quote:
Originally Posted by thomaspf
As Glassman pointed out above with buffer sizes of 1024 or 2048 samlples the Juli@ works pop free for any 16/44.1-48 or even 24/96 content you play.

If you still see issues you might want to look at the PCI delays for the other components in your machine. I had some unrelated studdering issues and found that reducing the PCI delay of my Radeon card to 96 cycles helped making ASIO playback very solid.

Cheers

Thomas
Now ASIO has worked fine when I changed the latency to 1024. Hope it keeps on working.
post #18 of 57
Quote:
Originally Posted by Patu
This forum is full of people who use computer as their main source.

Tell us your opinions! Which one is better?
If you haven't got anything else resembling a half-decent source (all I had before my Aureon was a ~1992 entry-level Onkyo DX-6920/DX-703 CDP, already a big improvement over the one I had before), a PC sound card is an inexpensive way of enjoying good sound without spending a fortune. A PC-based setup also gives you almost infinite flexibility in terms of available material and signal processing (you can do high-quality, precise EQ in software that would probably cost a fortune in hardware). I'm now routing the signal from my FM tuner through the PC even when not recording, so I can use my EQ all the time. (Got some cables [1 2] today, so I no longer have to plug things around all the time.)
post #19 of 57
Thread Starter 
Quote:
Originally Posted by sgrossklass
...
Well my point was that which one is better, ASIO or KS.

I know that PC is best bang for the buck when it comes to source.
post #20 of 57
Quote:
Originally Posted by Patu
Well my point was that which one is better, ASIO or KS.

I know that PC is best bang for the buck when it comes to source.
Oops - guess I forgot that. Well, then, given both are implemented equally well driver wise, the better of the two is the one which is better supported (or works better) in your applications. Foobar2000 users will generally be happy with KS (but may try ASIO if sound interruption on system sounds occurs), while Winamp addicts have hardly any choice besides ASIO.
post #21 of 57
Quote:
Originally Posted by Patu
Well my point was that which one is better, ASIO or KS.

I know that PC is best bang for the buck when it comes to source.
You didn't make that very clear! LOL

Well ASIO is a standard way to talk directly to the hardware, whereas KS is something built into windows 2k/xp to do a similar job. In the end it's a case of YMMV, but I prefer to use ASIO.
post #22 of 57
Thread Starter 
Quote:
Originally Posted by lordgibbness
You didn't make that very clear! LOL

Well ASIO is a standard way to talk directly to the hardware, whereas KS is something built into windows 2k/xp to do a similar job. In the end it's a case of YMMV, but I prefer to use ASIO.
At least with ASIO music keeps playing even if there's some other sounds on the background.

Now ASIO has worked fine with the 1024 latency.
post #23 of 57
If KS playback cuts out when other stuff is played, I'd try setting the sound card acceleration to no more than basic (one notch from left), which will disable hardware mixing. Shouldn't make a difference for Envy24HT(-S)/MT/GT cards (the chip can't do it anyway), but who knows.
post #24 of 57
I have experianced the same cracking that Patu discribed. I just got my Juli@ installed a few mins ago. Do you think this is a problem that the Juli@ is more prone to over other cards?
post #25 of 57
I use XFi <coax> Headroom Desktop DAC/AMP > HD 650. I tried both the KS from Foobar2K and ASIO4all. I experienced the same "crackling" you describe. It reminded me of handheld AM radio sound to some degree. I get intrusion of XP sounds in my music stream using KS, but dont have actual stopping of the music.

Even if the crackling were stopped with a higher buffer size, there was still a harsher quality to the music, which was not present in KS, so I am sticking with KS for now.

Taking a break before trying ASIO with a USB setup on my rig.
post #26 of 57
Quote:
Originally Posted by Dave_M
I have experianced the same cracking that Patu discribed. I just got my Juli@ installed a few mins ago. Do you think this is a problem that the Juli@ is more prone to over other cards?
Well, anything optimized for low latency will be more finicky here. Look up the PCI INT# routing in your motherboard manual, maybe you can find a slot that does not share an interrupt with much else (particularly storage or USB devices or other bandwidth/interrupt intensive stuff).
post #27 of 57
Regarding the KS vs ASIO vs DS debate: on paper ASIO is the best option as KS, for some reason, in certain cases, still gets resampled, but in practise they probably have inaudible differences.

I had an argument about it at hydrogen audio.

I'd be interested to see if it's possible to AB them... other than having a friend with me to click the different output settings, I don't know how to test.
post #28 of 57
Quote:
Originally Posted by Bones13
I use XFi <coax> Headroom Desktop DAC/AMP > HD 650. I tried both the KS from Foobar2K and ASIO4all. I experienced the same "crackling" you describe. It reminded me of handheld AM radio sound to some degree. I get intrusion of XP sounds in my music stream using KS, but dont have actual stopping of the music.

Even if the crackling were stopped with a higher buffer size, there was still a harsher quality to the music, which was not present in KS, so I am sticking with KS for now.

Taking a break before trying ASIO with a USB setup on my rig.
Why Asio4All. Don't X-Fi work with it's own Creative ASIO?
Is Foobar's "Kernel Streaming" plugin still on experimental stage?

jiitee
post #29 of 57
Quote:
Originally Posted by HiFiRE
Regarding the KS vs ASIO vs DS debate: on paper ASIO is the best option as KS, for some reason, in certain cases, still gets resampled, but in practise they probably have inaudible differences.

I had an argument about it at hydrogen audio.

I'd be interested to see if it's possible to AB them... other than having a friend with me to click the different output settings, I don't know how to test.
With a good speaker system, it's fairly easy to tell the sound quality difference between ASIO and KS using Foobar. The difference is less obvious with my K1000/K340 setup but can still be heard.

I know other people hear it, too, b/c Bert Doppenberg (or the Oris hornspeaker fame) recommends using KS instead of ASIO b/c KS is more (paraphrase) "natural" sounding to him. He obviously uses SET/oris horn speaker system, not headphones.

I hear this difference, too, but for the same reasons I prefer ASIO. To my ears, KS is a bit more liquid, laid-back, and has slightly softer sounding leading edges, leading to the "natural" sound.

ASIO to me has crisper leading edges, more forwardness and clarity, and better bass definition. ASIO is less forgiving of bad recordings, which will sound less "natural" and more digital. However, on good recordings, I believe ASIO to have better resolution and fidelity.

Before someone posts the usual "ASIO and KS should sound the same" statement, I highly recommend carefully comparing them using music you're intimately familiar with, preferably with revealing speaker systems.
post #30 of 57
I have carefully compared them ever since ks was in beta and have not found to be any difference at all.

In fact in many cases it does not really matter which front end API the application calls to send bits to the sound card. The driver that hands identical bits to the sound card is often one and the same.

So if there are differences these can not be attributed to diferences in ASIO or KS in general but are probably effects of the drivers for a particular card. The ASIO to KS filters that exist are a good example demonstrating that point.

Vista will actually ship with a new API that provides direct and exclusive access to a sound card similar to ASIO and KS but will work on a much broader range of audio devices.

Cheers

Thomas
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Computer Audio
Head-Fi.org › Forums › Equipment Forums › Computer Audio › ASIO vs. Kernel Streaming