An idea i have is that you could add where whenever you type in the transcript the thing you type is what you hear