My experience has been that the proximity to audible words is inversely correlated with how well I’m engaging with the subject. If I’m overtly reading aloud for anything other than an audience or to taste how a particularly elegant construction feels coming off the tongue something has gone wrong.
If I can I’m reading it to myself, silently but still conscious of the text as a series of spoken words even if not actually voiced, I’m bogged down and making poor headway.
When things are going well the awareness of the text as having a phonetic counterpart just sort of fades into flow. Not sure how that works; but you just lose the time and find yourself with a knowledge of the intervening text despite having no memory of having passed over it in anything but silence. Hard to describe.