[arrowsmith]

That's really interesting. I wonder what other audio cues can be used other than the simple content of the words.

Voice, maybe? Humans are great at distinguishing lots of different voices. Are there any screen reader tools that use a mixture of different people's voices to add additional fidelity to the information?

[ali_m]

Yes, some screen readers such as JAWS can be configured to use a different voice to represent bold text, for example. They can also use non-speech sounds to represent things such as HTML elements.

One of the things that really blew my mind is just how fast a proficient screen reader user can crank up the speed, e.g. https://soundcloud.com/freecodecamp/zersiaxs-screen-reader. It's totally unintelligible to me, and doesn't even sound much like human speech.

If you're no longer constrained by the speed and sound of normal speech all kinds of other interesting audio representations become possible.