Dear sir / madam, Nice to meet you! I see that Descript is doing text-to-speech, but I am really looking for speech-to-speech. I am a non-English native speaker, and I want to use speech-to-speech to correct my awful accent and pronunciation. How does this speech-to-speech work? Well, for example, I can upload a voice record of mine to Descript, then Descript changes my voice to another person's voice, and I can download the new voice record. The advantage over text-to-speech is that speech-to-speech does not require me to laboriously type anything word by word, and spend enormous amount of time to manually adjust the tone, pitch, mood of each word, so it not only can save me more time than text-to-speech, but also sounds more like a real human, because the recorded voice not only contains the mood and natural tone, but "ums," "uhs", "you know", and a dozen other filler words, which increase the realism. Your competitor is ahead of you. For example, voice.ai can currently do speech-to-speech. But they can only do this for 5 minutes per record uploaded to their server under the Pro License. But 5 minutes is too short, I am looking for 20 minutes per record. For the ethical concern, the newly converted voice does not have to be from a real person. My purpose is just to correct my awful accent and pronunciation. So the newly converted voice can be from one of voices created entirely by AI, for example, Nancy from one of the stock voices presented on the Overdub page of Descipt: https://www.descript.com/overdub Thank you!
