thanks for this info Beakybird.
The problem with JPG specs is that it is a picture of text, not real text, so my voice synthesizer doesn't know it is text. It sees it as a picture only.