Reflect real-world use of mismatched synthesizers #63
Description
Currently
In my presentation I demonstrate how hard it is to understand a synthesizer mismatched to the written language, and make the case that we need more synthesizers.
Proposed
The demo and conclusion are still sound, but should also acknowledge that some screen reader users do manage to get by with a mismatched synthesizer.
One source (P.), a multilingual full-time screen reader user and technical expert, reported the following.
I asked:
Tell me again please, a couple of examples where users have used the "wrong" screen reader voice?
He responded:
Examples I've heard about:
- Italian for reading Romanian (Eloquence)
- The Czech Vocalizer voice Zuzana for Croatian, since Lana contains quite a few errors
- Czech Zuzana also for Lithuanian
- As already mentioned, various Finnish voices for Estonian
- For many unsupported languages, some blind people have learned to understand the English Eloquence. I know a couple of people from Poland who were able to do this; in Slovenia, it was apparently quite a phenomenon.
- I suspect the German voice outputs are also used for Luxembourgish, since no meaningful Luxembourgish voice output exists.
Of course, such an experience is suboptimal, but it shows how determined people are to have access to information.