kernelstreaming, recommended or not? |
Please read foobar2000 Tech Support Forum Rules before posting and comply with all the points.
Failure to provide all the information pointed out in the above document in your post is considered wasting other people's time and in extreme cases will lead to your topic getting locked without a reply.
See also: Hydrogenaudio Terms of Service.
![]() ![]() |
kernelstreaming, recommended or not? |
Apr 23 2003, 18:53
Post
#1
|
|
![]() Group: Members Posts: 1037 Joined: 23-May 02 From: DE Member No.: 2107 |
i used the search function and found interessting topics about kernel streaming but not if it is still just an experimental technology in foobar or if it should be used if it works. are there major bugs so that i should tay away from that feature or is it "officially" recommended
thanks, ben |
|
|
|
Apr 23 2003, 19:24
Post
#2
|
|
|
Group: Developer Posts: 717 Joined: 25-September 01 From: ... The Studio Member No.: 20 |
Honestly, with your level of knowledge regarding kernel streaming, you shouldn't really be using it. DirectSound gives "perfect" output most of the time, and while it's possible to get a few improvements by using DirectKS (kernel streaming), it's even easier to cause instability, inferior quality, and other negative issues if you aren't familiar with KS. Above all, please remember that DirectKS / Kernel Streaming is an experimental function, and it isn't a Foobar2000 only option; it's used in Cakewalk applications like Sonar.
I will quote a previous post of mine. I found it by using the Search function: QUOTE The differences between Kernel Streaming, DirectSound and waveOut have been discussed here in great detail and I couldn't summarize them without missing something. If you search you'll find what you are looking for. In short, if you have Win9X/NT you should use waveOut, otherwise use DirectSound unless you have a good reason not to. Kernel Streaming is technically called "Microsoft DirectKS". DirectKS works by opening a direct connection to an audio device and sending the data directly. This bypasses the Microsoft Windows Kernel Mixer, any device driver mixers or resamplers, etc. If you have bad device drivers that modify sound, experience one of the few known XP kMixer bugs, require (minimally) lower CPU usage, low latency output, or any other special case such as high samplerate output, then it's best to use KS. Known problems with KS include several issues, which vary wildly based on device drivers and hardware. For example, in some cases kMixer resamples your output, but it could also your audio device drivers, and possibly the hardware itself that is ultimately responsible for the resampling. KS bypasses the first two, so you can ONLY output a different sample rate via KS if the hardware itself resamples. If it does not you must use the DSP resampler. Volume control is a similar issue. If it isn't controlled by hardware then it will not work when using the KS function, but you can use Foobar2000's volume control (Now called Attenuation). Sometimes monoural output via KS will only go to one channel depending on hardware, and lots of other issues I haven't mentioned. If your hardware doesn't support multiple hardware channels then using KS will result in your hardware being used in a locked state -- no other applications can use audio while FB2K is playing. Depending on your hardware you might experience some, all, or none of these problems. Also beware that if FB2K crashes or you encounter a bug while using KS, you can crash your system with a bluescreen error or hard lockup. Edit: NOTE - Misconfiguring DirectKS can result in a loss of accuracy and quality. If you don't use the Resampler DSP and kMixer was resampling for you, it is possible that your audio hardware has lower quality resampling than kMixer. If, for example, most of your audio files are 44.1kHz and your hardware operates at 48kHz, expect lower quality. In other words: If you do not know what you are doing, do not enable KS. If you do enable it and you don't understand your hardware and software completely, prepare for system crashes, instability, and lower sound quality! It should also be noted that misconfiguring KS can also result in hardware lockups and Windows crashes, including kernel panics and "blue screen" errors. I hope the information above is correct. Despite what some may believe, I'm not the expert on these matters. -- Jeff This post has been edited by TrNSZ: Apr 23 2003, 19:28 |
|
|
|
Apr 24 2003, 06:34
Post
#3
|
|
![]() Group: Members Posts: 30 Joined: 24-April 03 Member No.: 6154 |
does the turtle beach santa cruz under windows xp supoort kernel streaming? and if so, what are the best settings?
btw, it lets me select it in the settings and plays sound, but i don't know if it actually works. no crashes though. |
|
|
|
Apr 24 2003, 06:46
Post
#4
|
|
![]() Group: Members Posts: 992 Joined: 17-March 03 Member No.: 5544 |
QUOTE (rubix @ Apr 23 2003 - 09:34 PM) does the turtle beach santa cruz under windows xp supoort kernel streaming? and if so, what are the best settings? btw, it lets me select it in the settings and plays sound, but i don't know if it actually works. no crashes though. Yes. 16-bit fixed point. (You can pass larger samples, but the drivers will just downsample) The Santa Cruz accepts 44.1Khz output with no problem. I've never tested 48Khz output. If it's playing, then it works. |
|
|
|
![]() ![]() |
|
Lo-Fi Version | Time is now: 23rd May 2013 - 01:54 |