- Posts
Since I first tried the text to speech tools a couple years ago they appear to have much improved.
Found this – https://ttsmp3.com/
Good sound quality and the pronunciations are not terrible, so they’re easy to understand. I stuck a foreign language translation into it and now I have a foreign lanuage station ID.
Also allows you to download the audio file for easy use.
Very good. Additionally, for any text-2-speech engine you can fix pronunciation mistakes by changing your text to a phonetic equivalent. Additionally, if you are trying to programmatically/dynamically generate speech for use on-air, you can run your text through a script which converts words and abbreviations into their phonetic equivalents.
So, if you need an address to be read as “State Road 25” when it sees “St. Rd. 25” this can fix that, rather than the text-2-speech engine saying “Street Road 25”.
The sed command line tool from GNU/Linux is easy to use for this and you should be able to find it for Windows as well. Back when my station was still operating I kept a number of “dictionary scripts” updated which could fix the most common mistakes I heard my text-2-speech engine making.
Example:
echo “St. Rd. 25” | sed ‘s/[Ss]t./state/g’
state Rd. 25
- You must be logged in to reply to this topic.
