In the Voice Dream Reader app, you can select a mispronounced word, type in the correct pronunciation, hit a play button to check how what you typed is pronounced by the reader, and then save that pronunciation. This would be a useful feature in Audemic.
This is absolutely critical for so many fields. TTS struggles quite a bit to pronounce taxonomic nomenclature correctly, so the ability to manually do this would be fantastic.
I agree. I really like the idea of this app, but it has some issues with pronunciation. For example, if the app comes across Ca2+ (calcium ion), the Audemic speech engine will interpret it as "CAAH".
Literal pronunciation of technical words often defeats the purpose of having the computer read papers to me in the first place. If the users could correct the pronunciation of common technical words, Audemic could compile a nice library of properly pronounced words for their user base.
Yes, this is a necessity. I have only tried this out on one paper, and within the first section, I found that I needed to keep going back and listen to the same passage several times because of mispronounced words or abbreviations that made it impossible to follow what the passage was about to the point that just reading the text myself would be more efficient even though I read extremely slowly (hence my investigation into this service). I am currently using the free version, but unless this kind of feature is added, I won't be upgrading to the paid version.