Hello, I am building a project and I want to use an AI text-to-speech software to say few words in the voice of famous person. I am pretty sure I can build such a thing, but it needs to start with a wav file, or similar, with the voice. Does anyone know of a free software or website that can this? Its for non commercial purpose and only very simple. I would rather not pay for expensive software access for this simple job. Thanks for any help.
Steven Lightfoot
Really? You can't find free text to speech?
I should have been more clear - Is there any free software that regular users like and recommend? As you imply there is ton of stuff out, lots of pay-to-play, I am looking for recommendations for a free service/software/app that people use and trust and can recommend. Thanks, Steven Lightfoot
I am also specifically looking for the free software to probably use AI, in the sense it will take a famous person's voice and use to create the speech. I will select the famous person for whom I wish to do this project, but just for example, lets use Arnold Schwarzenegger.
Google "Deep Fake" and start learning how they're made. It's definitely not done with an Arduino.
Sure of course. I want a free AI to text to speech generator, that uses a famous person's voice to generate a sound bite, that I can use as a WAV type file to use in a Arduino project. Should not be too hard.
I found this, and it works OK: AI Celebrity Voice Generator | Parrot AI (tryparrotai.com)
But you can only do maybe one or two free examples, I want something like this that is more open with no limits. I don't mind paying a but if I have to, I just think there should be something free out there that some experienced Arduino types might use, or be aware of. Thanks.
Why do you expect something like that to be free? Nothing is free my friend. If you aren't paying for the product then you ARE the product.
Being as this has nothing to do with Arduino I don't know that I'd expect that. Not everything "computer nerd" is Arduino. This is a project for a PC. I bet the YouTube community would know more about something like this.
Well there a variety of free open type AI projects, so its possible something like this exists, at no cost, possibly from an academic type place.
Also, I would fully expect that some Arduino projects have sound effects, including playing WAV type files, so it is not unreasonable to assume some Arduino users may have experience with generating WAV files from various places, including synthetic created. I cant believe I am the only Arduino person interested in this, I am not THAT cutting edge.
Thanks for your input.
The crux of your problem is the "famous person voice" part. There are many sites with sound effects and free phrases in .WAV format, but in generic "ordinary person" voices. I just google them when I need to playback a voice like "stop" or "start."
Thanks, that is a useful comment. You may be quite right. The website I posted above is actually pretty good, you can type some text and it will have the person you want say the words, it's quite amazing. I am building kind of a novelty/joke product, and so I need a specific person's voice. That website have a fairly low fee, so I may end up just paying it and having the access I need. Cheers.
In the early 2000s, I made "Great Gig in the Sky" using Steven Hawking's chosen TTS. I've heard nothing since that can rival it.
That sounds cool!
https://lingojam.com/StephenHawkingVoiceGenerator
It's just that easy. (sorry Clair Torry)
Thanks! That might be quite useful.
This has some potential, appears to be free, although the voice accuracy I am looking for is a little crappy. But its better than nothing.
You must upload this.
Hah. The text was a string of "oo", "oh" and "ah"... with days of experimenting on time length to approximate the "ooh oh, oo ooh oh, oo ooh, oh aah ah, oo ooh oh..." My sigo thought I was seeing someone else by the end of the week. I'm quite sure it is still on my Gateway 33Mhz laptop with Redhat (replaced Windows 3.11), in a dump, somewhere, near my box of 3.5"fdd ... long gone.
I can only ask that you imagine (or try your own) Stephen Hawking's TTS, reasonably timed, "oho, oho, ooo, ooh ooh..."). It was a magical song. I had fun with it. Mind you, the TTS of 2003 (?) was one speed, so I had to adjust to it... and I did that because I recalled working with a guy who used his speech synth in the early '80s to make rude sounds on ?Macintosh? ?Franklin?... "Mp Mp Mp" was the sound of a two-stroke golf cart (back before LPG or electric) or two people getting friendly. Computers always digressed to game toys... even front-line electronic warfare test systems had Marple Manor, Zork, Trek, and some gravity simulator... the computer world was bootleged.