I just learned that someone can record your voice and generate an artificial voice that sounds like you to scam your friends and family. Now I’m concerned that someone might record my voicemail prompts and use it for that purpose.
Instead of recording my voice, how can I create computerized prompts that ask callers to leave a message, etc?
Thanks for any info…
Any AI Text to speech service that will give you an audio file will probably work for this. I have no particular service to suggest but googling “text to speech” gave me the below as the first result.
Any idea the type of file I need for freepbx, and where to place them?
Pretty sure it’s a wave file and you can save them in /var/spool/asterisk/voicemail/default/ and it’s busy.wav, greet.wav and unavail.wav for the different types.
Ideally you want it in multiple formats, especially if you are using computationally expensive codecs.
For “wave” files, the extension must be lower case “wav” and the file must be a very specific type of .wav file: 16 bit, mono, signed, linear, PCM, at 8kHz. The sound files should be placed in /var/lib/asterisk/sounds/XX/, where XX is the language code in use.
You could also hire a professional voice artist.
I’m pretty sure that piecing together sections of a recording is old hat, and deep faking can be done with synthesized speech.
Much better and more precise answer.
Edit: Except for the location of the files. Pretty sure he is looking for the individual voice mailbox greeting location and pretty sure that needs to be inside each individual mailbox folder, no?
Yes, i think I have it, sorry to bother you.
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.