Hi, Henrik, Look at the AVSSAMP sample driver in the DDK. Remove all the Video code and you should have something very close to what you want. You will just need to change the driver so that instead of reading from a file on the hard disk, it accepts audio data from a program. There are a couple of different ways to transfer data from user mode to kernel mode. Considering that this sample is available what you want to do is relatively easy. Having said that, I should warn you that if you have not written an NT Kernel Driver before this will not be an easy task. So you should make sure you have adequate time allocated. Good luck and don't hesitate to contact me if you have any questions. _____ From: wdmaudiodev-bounce@xxxxxxxxxxxxx [mailto:wdmaudiodev-bounce@xxxxxxxxxxxxx] On Behalf Of Henrik Thystrup Sent: Tuesday, June 03, 2008 10:40 AM To: wdmaudiodev@xxxxxxxxxxxxx Subject: [wdmaudiodev] Virtual Mic driver Hello, I've been assigned a task to develop a virtual mic driver. The sound data is collected from a shared memory created in an application and the idea is that I develop a driver that when selected from e.g. skype og MSN get the data from this shared block and pumps it in the application and the user at the other end hopefully can hear the data, When reading info I concluded that I should develop this as an user driver so I dont get into the kernel/user mode memory problem. My question is if this is the right design and also where I can get som example off how to do do it. I guess I need to use the WaveIn class but how ? I've installed the DDK but to be hornest it a huge task to read and understand all that information. Any help would appreciated. - henrik