Provenance of Generated Media
Your platform is powerful, and yet provides no concrete safeguards to prove generated media from real media. Build a system that prevents this platform from being used to generate disinformation and fraud.
Caribbean voices wanted
The Caribbean is a mixture of ethnicities, so a Caribbean voice can have an African or Indian tone. But these are not the voices from those subcontinents
Quality Update for Ultra Realistic Voices
We're working on a major update that will significantly improve the following aspects of our Ultra Realistic voices - Improved audio quality. Fixes the issue of words getting chopped off toward the end of sentences. Fixes the issue of words repeating randomly in the audio generated. Improved prosody, intonation, and flow of sentences.
Voice of "Wilbert" and "Donovan" change into female
As said in title. The ultra realistic voices of "Wilbert" and "Donovan" change with long generation of text. Wilbert changes, after few sentences, into female british voice. ( Even in sample its heard?) And Donovan changes into normal instead of deep sound after few sentences. It would be good if someone could cjeck this out.
Edit multiple voices in one file
Cannot add pauses, change tone, lengthen words etc on any voices other than the main voice in a Premium voice file
Download .SRT file after creating audio
A new feature to download an .SRT file of the audio I created on the Play.ht platform.
Support for SSML tags in the Editor
This functionality would include SSML tags support inside the text editor that would allow users to change voice tones, speed, etc. It would look something like this - <pitch=low> Here is some text that is spoken <volume=low> in a low volume </volume=low> and this is text at the normal volume. </pitch=low>.