You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The example of this issue I'm going to use here is soundscape:55.9552485,-3.1928911 which is near a road named "North St Andrew Lane". The text to speech is reading this out as "North Street Andrew Lane" instead of "North Saint Andrew Lane". An easy mistake to make and one that we could perhaps solve with a list of Saints?!
However, I suspect that there will be an endless list of this type of issue across all of the various locales and countries and so is there a better way to do it?
The text was updated successfully, but these errors were encountered:
The Vocalizer TTS app provides an alternate text to speech engine. It correctly says "North Saint Andrew Lane" and so is better than the Google default provider. There are numerous other possible engines - see https://accessibleandroid.com/tts-engines-on-android-past-present-and-future/ for a list. I've also tried Acapela Voices and CereProc from that list and both can work with Soundscape (I had to restart my phone to get CereProc to work). Only CereProc fails to say Saint instead of Street in the same way as Google, but its voice is very nice.
Another interesting data point in text to speech. If I use Cerevoices TTS engine then:
"Cafe.Main street goes left"
is read out as
"Cafe dot main street goes left"
If I put a space after the dot, then it's left as a pause. Google TTS doesn't need the space as a hint. We probably need to add a check on the resource strings to try and ensure any issues like this we find are tested against.
davecraig
added a commit
to davecraig/Soundscape-AndroidTest
that referenced
this issue
Oct 25, 2024
See issue Scottish-Tech-Army#201 for discussion on this. Some speech engines read out
full stops as 'dot' if there's not a space following them. This
avoids it in this one situation.
See issue #201 for discussion on this. Some speech engines read out
full stops as 'dot' if there's not a space following them. This
avoids it in this one situation.
The example of this issue I'm going to use here is soundscape:55.9552485,-3.1928911 which is near a road named "North St Andrew Lane". The text to speech is reading this out as "North Street Andrew Lane" instead of "North Saint Andrew Lane". An easy mistake to make and one that we could perhaps solve with a list of Saints?!
However, I suspect that there will be an endless list of this type of issue across all of the various locales and countries and so is there a better way to do it?
The text was updated successfully, but these errors were encountered: