human voice vs. text to speech preference?

[slutinmyhead]

Just wanted to weigh in... No matter how insecure you are about your voice, your files will sound better by you than using text to speech. Anyone else want to put in their two cents?

[zapnosis]

I think that the sound has to match the file content. If you want something personal, seductive and flexible, then of course use your own voice. If you are using more mechanical programming - unwavering, uncompromising and relentless - then TTS has the edge. Always provided it's done properly, of course.

And if you have a speech impediment... where does that leave you?? Potentially a lot of editing!
Z

[Jeshi]

I think the effort it takes to get a TTS to sound actually good is about on par with editing your files so you don't stutter (and unless your stutter is really very bad then the subject probably wouldn't even care if you stuttered.) I definitely agree that even the worst human voice, so long as the audio quality is fine, is still better than the best TTS voices. And I've heard some amazing TTS voices before. Usually british women sound the most realistic, but no TTS can ever get the prosody to sound natural, which is a big hurdle. The right prosody can really make a file amazing or bad.

[gdrileyx]

I think it depends on what type of listener the subject is. I think, for an active listener, a human voice can be too expressive, and keep the person from going into trance, because the active listener cannot help listening to nuance in the voice, trying to find hidden meanings or true intent.

[joecomp2000]

definitely voice ... unless you have the worse stutter in the world and have access to the best text to speech ..then I doubt anything but a real voice would be better ..

just in cantor alone, the proper pace the proper pauses .. this is all needed .. sure you can trance with a text to speech ..but it not as deep.. and no matter what you are going to flow better with a human voice IMHO

plus with many of the t-to-s you need to spell things funny to get it to pronounce right..

if you are making a mp3 use audacity and just keep reading .. any mistakes just keep reading... you can edit them out ..play back while reading ...and see and do the edits..

finally reading allows you to see if the.file actually flows.. I don't know how many time I was recording and realize...that sentence just doesn't make sense.. where as t to s , you covert ,play back muck around with spelling , covert and hope you got every thing..usually takes longer , and
have worse results ..


my $.02