What I meant was, they chose a representation of the document consisting entirely of ASCII characters, to allow it to be stored with an ASCII encoding.

Of course, if all the characters in a document are ASCII, then the ASCII and UTF-8 encodings of the document are identical.