I'm confused, adrianf, as you seem to be contradicting yourself within this same thread.
That's definitely a loose connection somewhere in the chain. ...
Has nothing to do with a "loose" or dirty connection. ...
My understanding is that a poorly-powered mic rapidly loses dynamic range and soon distorts (most noticeably) on a loud source and yet the sample is of a very quiet recording, no music, only Ranaldo speaking, with no apparent change in level even though the noise disappears towards the end of the sample.
I'm not disagreeing with you, only trying to understand more, or better what you may mean. Thanks.