Detecting language please wait for.......
We have developed a new app that supports Text to Speech, and also foot pedal control for segment navigation and voice playback. The details are here:
If anyone is interested in this kind of solution, which is extremely helpful for reviewing (listen to source while reading the target, or vice versa), especially when working with NMT or Quality Assessment, please let me know and I will make the current Beta version available to you.
This thread is getting a bit long, but I notice (184.108.40.206) that when altering a segment, say
"Hello world" -> "Hullo world", TTS will only read after the position of the text change (in this case "world").
Unfortunately I did not have the time to test all the variations of this. A speed change will reset this effect. The effect seems to only occur at increased speed (1.3 in my test).
Daniel Hug said:"Hello world" -> "Hullo world", TTS will only read after the position of the text change (in this case "world").
Thank you. Claudia reported this a while ago in the thread, and I noted this too. It happens when working with increased speeds. I don't know whether this has anything to do with the service or if it's something we can address.
Andrea-Melinda Ghisa I wonder if we can enforce small delay before attempting to read the text for the higher speeds? Perhaps that's where the issue comes from.
Paul said:I wonder if we can enforce small delay before attempting to read the text for the higher speeds? Perhaps that's where the issue comes from.
Daniel Hug the same thing happens with the slower speeds we introduced into the production build. We are going to report this to Microsoft and see if we can get it resolved as this seems to be out of our control.