Tags:
podcast
This is a list of limitations of the types of automated audio translation offered by such services as Talkr.com. Since we do not see a list in their help center, we thought we would compile our own list and offer it as a wiki page for any customers to keep a list of limitations. Limitations
- The pronunciation of certain words, which seem familiar to us but are not real English words, so they are hard for the automated translation. For example, the word Notepad. It was pronounced in three syllables as "note-uh-pud".
- They read code and URLs, which you may want to skip over. It's hard to read code. Often what the automated translation does it read it letter by letter.
- While it reads the bullets of numbered (or ordered) bullets correctly "One ...", it doesn't say anything for an unordered or unsequenced bullet list. There is no audible cue for unordered bullets.
- For em-dashes and other characters that are not handled by translation come out as A's, so it sounds like a Canadian reading, eh? For example, the title of one blog entry that has an em-dash between two parts of the title comes out with the first part, then "Eh" then the second part.
Sample
Here is a Talkr-powered "podcast" of this blog entry.{SUBMIT()}{SUBMIT}
