microphone¶
This serves as the audio driver folder
Microphone Configuration¶
microphone_model: currently only support
RESPEAKER.chunk: number of frames per buffer from hardware to memory each time.
record_seconds: number of seconds each time.
channel_type
May be ASR (Audio Speech Recognition), RAW (Raw Audio Data) or Playback.
sample_rate
Number of frames that are recorded per second.
sample_width
Number of bytes per sample (1, 2, 3, or 4).
For example, if there are 6 channels with 16000 Hz rate and 2 bytes width, then 4 second recording is
Number of frames: 16000
Number of “samples”: 16000 * 6
I use the term “samples” here just to make it easy for understanding, which is seldom used in this context.
Total size: 6 × 16,000 x 4 × 2 = 768,000 bytes.
You might see other metrics elsewhere as follows:
BIT DEPTH same to sample_width except that the unit is bit.
BIT RATE number of bits encoded per second (kbps or k) – for compressed format like mp3.
Output¶
Raw microphone data (cyber channel
apollo/sensor/microphone).