I'm trying to add a simple audio effects processor to my project. (mono in - stereo out - chorus effect)
Audio input via the onboard 12bit ADC, output via a stereo 12bit SPI DAC.
Sampling rate needs to be at least 32kHz, so there's about 31µs time for AD-conversion, processing and DA conversion.
Pseudo code:
loop @ 32kHz
{
adc_reading
audio processing
dac_writing
wait until 1/32000s are over
}
HSPI and VSPI are already in use for other things, so I tried to bitbang data to the DAC chip, which works fine and only takes about 6µs for sending out a set of stereo samples. The audio effect algorithm only takes a few µs either. So there's about 15µs left for ADC sampling.
My first approach was using a timer at 32kHz and to read single samples with "adc1_get_raw". However "adc1_get_raw" takes 40µs for one sample, which is too long unfortunately.
I am now looking into I2S and DMA, but I haven't found out how to make it work for my specific application.
I can make it work by accessing the DMA buffer, but since my loop and the i2s sampling rate are not perfectly in sync, I am having terrible jitter noise. I don't know how to sync my loop to the I2S sampling rate. I need some kind of trigger/interrupt for every input sample.
Ideally I'd want to use i2s_read to grab a single sample and let it controll the timing of my loop, but since I cannot set the buffer size to 1 sample, the only way to get samples is in packets of 8 or more.
Clearly I am lacking knowledge here
Any kind of help is much appreciated!
Thank you very much
