I have developed a simple (virtual) AVStream audio driver. It works, but there is a problem: my advertised audio formats are being ignored. I can support exactly one format at a time (for example, 22050-M-16), although that format will change from time to time, very similar to the avssamp sample. My thought was that I would advertise a variety of formats in my KSDATARANGES, and then refine things in the Intersect handler and the SetFormat handler. That's how I've done things for years with video capture drivers. But if I bring up graphedt, instantiate my device, and look at the formats on the output pin, my intersect handler is never called, nor does it look at my KSDATARANGES. The pin shows something like 50 different formats, including all the multiples of 11025 up to 44100, and all the multiples of 8000 up to 96000, mono and stereo, 8 and 16 bits. I've even changed things now so I only advertise one KSDATARANGE, and the list still contains 50 formats. When I first installed the driver, the control panel "Default Format" came up as 44100-S-16, and that's the format that the system always tried to use (in graphedt and with waveIn apps), even though that's a format I didn't support. I changed the control panel "Default Format" to 22050-M-16, and that worked. But if I now change my driver so that I only advertise 44100-S-8, it still insists on connecting at 22050-M-16. It is remembering the old information. I don't want to embed a sample rate converter in my driver, and I shouldn't have to. There's already one in Windows. How can I convince the audio system to negotiate with me? Where is the "Default Format" stored? -- Tim Roberts, timr@xxxxxxxxx Providenza & Boekelheide, Inc.