I've spent time debugging this and it looks like the problem is with the BT chip firmware that potentially contains not ideal parameters.
I've also captured an audio sample directly from pcm and the sound is clear at the beginning however the quality decreases over time with the noise overtaking the voice eventually [like the noise volume increases step by step]. At the same time the voice through BT headset is polluted with the noise from the beginning.
I have also used a different firmware for the BT chip and as it produced no audio at all through the BT it has also made the pcm record perfectly clean from the beginning till the end.
I've spent time debugging this and it looks like the problem is with the BT chip firmware that potentially contains not ideal parameters.
I've also captured an audio sample directly from pcm and the sound is clear at the beginning however the quality decreases over time with the noise overtaking the voice eventually [like the noise volume increases step by step]. At the same time the voice through BT headset is polluted with the noise from the beginning.
I have also used a different firmware for the BT chip and as it produced no audio at all through the BT it has also made the pcm record perfectly clean from the beginning till the end.