The Arduino-Filters library only does the filtering, it doesn't have any functionality for audio input or output.
The filter objects are simply (stateful) functions that take one sample of the unfiltered signal as an input, and return the corresponding sample of the filtered signal as an output.
Where these samples come from or where they're going is up to you.
For audio, you'd usually use DMA and hardware peripherals for I2S for audio in- and output. This is specific to the board you're using. For Teensy, have a look at the Teensy Audio Library, for ESP32, look into the ESP-ADF.
I want to sample audio from a HF receiver - filter some of the noise out and re generate the filtered audio.
I did do most of it a while back read A0 and then output the 8bit value on a Port
with a R2R resistor network but I still need to find a way to filter the audio before the output stage.
I have Nanos and STM32F103C (for the extra speed and memory ) only needing the source code that will make it work.
I tested a moving avg filter once , but there sure are better ways.
The Nano is nowhere near powerful enough to run filters with floating point math at any decent rate. Even its ADC is too slow for decent audio quality. An integer-based moving average might work, but the quality won't be great at all.
An STM32 should be powerful enough,but you'll have to dive into the datasheet to find out how to set everything up for audio. You'll probably find some scattered information online, maybe even code or a library. This is a huge project.
For audio applications, I'd highly recommend getting a Teensy with an audio shield. The Audio library has its own BiQuad filter implementation: IIRC, it uses integer math and makes use of saturating integer instructions, so it might be faster than using Arduino-Filters with floats.