Hi, we're on a pro plan and we've been using Elevenlabs in order to do the following with a quality recorded professional voice:
Generate "Hi [name]!" with EL + Stitch it to a pre-prepared audio file, So for example it'd be "Hi James! I saw you left our webinar in the middle.....". The issue we are seeing is:
- Whenever longer names are generated (i.e "Hi Jonathan!") the quality is fine.
- However when shorter names are involved (Mike, Jack, Joe, Dave..), oftentimes the literal generated voice piece doesn't even resembled the professional voice it is utilizing, completely off. I wonder, are there some special settings / way / method I can use to ensure quality geneeration of short phrases/names in order to be as close to our professional voice as possible? Right now it's just bad.
Thanks for any advice