A PCM input splitter device is a capture-only device that you can enable using the audio configuration file. By default, it's disabled. The PCM input splitter allows multiple processes and threads to capture from the same hardware device concurrently. For example, you could use different processes to capture audio for both voice recognition and file recording at the same time. The PCM input splitter supports up to 10 concurrently attached subchannels.
A PCM input splitter appears as a new PCM capture device that supports multiple subchannels, which you configure in the audio configuration file. The PCM input splitter is specifically attached to a single PCM hardware capture device. This one-to-one mapping allows an API call to identify the PCM input splitting device associated with its hardware device. This PCM device is created immediately after the software mixer device and has a device index that is one greater. If there isn't a PCM software mixer for the parent PCM hardware device, then the device index of the splitter is one greater than the parent PCM hardware device.
Both the PCM input splitter and PCM software mixer must operate at the same data fragment size, which you set via the sw_mixer_ms option. Because you cannot use a samples value to specify the fragment size the PCM input splitter uses, you cannot size the PCM software mixer fragments using samples (the sw_mixer_samples option) when you use the input splitter.
For more information about using an input splitter, see description for the input_splitter_enable key in the "Audio Configuration file" section of "io-audio" in the QNX Neutrino Utilities Guide.