Hi Eugene, It's not the kernel per se. There are a lot of device drivers fighting for CPU time. Some of them may be less than optimally designed, to put it mildly. You can not prioritize the device interrupts which would enable you to boost the priority of your audio interface for instance. But I don't think the situation is that bad. With ASIO I can work reliably with a buffersize of 128 samples which gives a latency of about 4 milliseconds. For my purposes that is good enough. -Evert -----Original Message----- From: wdmaudiodev-bounce@xxxxxxxxxxxxx [mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx]On Behalf Of Eugene Muzychenko Sent: Friday, February 09, 2007 5:16 AM To: Evert van der Poll Subject: [wdmaudiodev] Re: The latest in acheiveable latencies using Windows? Hello Evert, > I agree with you that 64 samples is probably the best latency you can > achieve with audio devices and PCI hardware is the most likely candidate to > achieve it with. I'm wondering why Windows audio is SO slow. 1 GHz Pentium processor performs near a billion of instructions during one millisecond. Due to memory, let's say it will be 700 thousands. WHAT could Windows kernel routines do to spend so much time? Regards, Eugene ****************** WDMAUDIODEV addresses: Post message: mailto:wdmaudiodev@xxxxxxxxxxxxx Subscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=subscribe Unsubscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=unsubscribe Moderator: mailto:wdmaudiodev-moderators@xxxxxxxxxxxxx URL to WDMAUDIODEV page: http://www.wdmaudiodev.com/ ****************** WDMAUDIODEV addresses: Post message: mailto:wdmaudiodev@xxxxxxxxxxxxx Subscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=subscribe Unsubscribe: mailto:wdmaudiodev-request@xxxxxxxxxxxxx?subject=unsubscribe Moderator: mailto:wdmaudiodev-moderators@xxxxxxxxxxxxx URL to WDMAUDIODEV page: http://www.wdmaudiodev.com/