It would be nice to be able to force Audacity to record whatever the sound card captured in its original format to avoid sample rate conversion and data format changes. The default values (or project values) can still be used to set the sound card up, but the user is able to figure out what sampling rate and format the sound card driver and hardware actually decided to use. This is especially useful for recordings that are going to be analyzed by a computer, rather than listened to by a human. Examples include sound location and separation estimated by microphone arrays, and human speech and animal vocalization recognition.
One way to do this for the sample rate:
under “Real-time Conversion” please add the option “None” to the “Sample Rate Converter” drop down menu.
Another way to do this for both sample rate and format is to
put a check box “Do not convert sound card outputs”