Description
Tested on 6.1.0 stable and 6.2.0-beta.2. I ran into this trying to write up a sound-reactive LED example for the BLM Badge. I distilled the code down to what's necessary to reproduce. It fails the same way on the Circuit Playground Express. This code works properly on Circuit Playground Bluefruit.
When the code runs, mic.record only records 32 samples, the rest of the 160 are 0. Below is an example. (All of the examples I can find of using sound to get levels uses 160 samples. That is where I got that number.)
array('H', [33445, 33360, 33397, 33451, 33380, 33401, 33402, 33420, 33466, 33399, 33382, 33373, 33410, 33440, 33427, 33334, 33398, 33402, 33329, 33335, 33420, 33384, 33466, 33378, 33411, 33440, 33361, 33376, 33436, 33444, 33386, 33321, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0])
Here's the catch. If I update the file and save, right before the board resets, it fills the array. So without a time.sleep(5) in it, the last few arrays are filled. With the time.sleep(5) in it, sometimes it fills one array, but often misses it. Below is an example from right before the board resets. I changed it to time.sleep(0.5) and it filled the last array before reset.
array('H', [33384, 33238, 33257, 33224, 33226, 33252, 33233, 33290, 33212, 33225, 33133, 33172, 33192, 33190, 33243, 33230, 33241, 33285, 33235, 33255, 33332, 33173, 33251, 33289, 33256, 33294, 33195, 33311, 33261, 33263, 33272, 33193, 33249, 33259, 33235, 33290, 33262, 33302, 33196, 33251, 33202, 33231, 33266, 33157, 33172, 33205, 33281, 33225, 33222, 33205, 33215, 33255, 33298, 33261, 33208, 33230, 33293, 33222, 33257, 33245, 33264, 33202, 33244, 33273, 33219, 33198, 33209, 33278, 33267, 33278, 33117, 33201, 33228, 33252, 33347, 33296, 33264, 33239, 33270, 33277, 33237, 33220, 33269, 33293, 33253, 33199, 33291, 33238, 33256, 33207, 33208, 33203, 33292, 33262, 33212, 33221, 33189, 33171, 33241, 33269, 33194, 33247, 33255, 33179, 33307, 33225, 33217, 33299, 33210, 33196, 33296, 33232, 33263, 33183, 33263, 33172, 33141, 33194, 33202, 33232, 33240, 33237, 33280, 33192, 33161, 33234, 33229, 33263, 33216, 33222, 33224, 33231, 33233, 33282, 33223, 33180, 33192, 33257, 33219, 33288, 33243, 33252, 33250, 33321, 33226, 33222, 33310, 33253, 33242, 33246, 33248, 33263, 33202, 33286, 33162, 33271, 33230, 33268, 33234, 33227])
Here is the code to reproduce.
import time
import array
import math
import board
import audiobusio
mic = audiobusio.PDMIn(board.MICROPHONE_CLOCK, board.MICROPHONE_DATA, sample_rate=16000, bit_depth=16)
while True:
samples = array.array("H", [0] * 160)
mic.record(samples, len(samples))
print(samples)
time.sleep(5)