Replica allows you to type text and create AI generated speech. We currently have free early access accounts available. Over the next few months we'll be launching our full product. We'd love your ideas and feedback while in early access.
Just tried this out. VERY cool! Feels like the beginning of a sci-fi movie coming to life! Y'all are going to get acquired so quickly, haha—looking forward to my Google home sounding like Mom 😆
@nicbyron thank you for getting on and testing it out so quickly! You are right in that this is only the begging :) We have a lot more to come in the future, and we are very happy to know you already like the product!!
Hi! Great tool, I was playing around with it last night. I used a number of the 'ready-to-go' voices and surprisingly found that you could get them to say HOW you wanted things just by changing how it would interpret the word. So for example 'acid' wasn't being pronounced properly so A-ecid that this worked better. I can gladly supply this 'cheat sheet' if this would be of use. I was also wondering whether you were also working on expanding the option to use a particular voice for your own voice-command projects? As a separate option. Lastly, I got a few errors and it said consult the FAQ but I couldn't easily find that - perhaps linking to the FAQ would be an option? I'll play around some more and can provide more feedback if you think it's useful!
@charlotte_champion thank you for the awesome feedback!!
Yes, the way to get around pronunciation issues is by spelling the words 'phonetically' : i.e. how they sound. English is a funny language!
Yes we are expanding the voice library over time so you'll have more options. And thanks for pointing out about the FAQ not being easy to find, we had a hunch about this already but you just confirmed it! Thanks again 😊
Really excited to try this, i've already built my first episode with IBM Watson AI voice and the result is outstanding. https://www.youtube.com/channel/...
I'll give it a try in the next episode
I joined the beta 2-months ago and was accepted in and have been working on my voice here and there. I will say that they are ahead of the competition that is out there. They have given me some tips and tricks for helping their service work good like converting numbers to spelled out numbers so the AI voice sounds right. It's a great overall product that has so much potential! Join the beta and see for yourself!
Hello @replica_ai, just wanted to say that the voices you have are incredible. Just awesome, I can see that you are a light year ahead of google tts voices for example.
I noticed that you only have Free and Enterprise plans. I would like to know if this will only be available for big companies and not the for the private user?
I would love to use your product for some projects I have, but need to know if I will be able to use it in the future.
Just incredible, Congratulations to your Team :)
@cameronsmith Ok thanks for the answer, I will follow your project and use it, and when you release the private plan will give that a go too probably.
Congratulations on a great product :)
@harshit_taneja No we don't see it replacing voice acting. We're actually helping voice actors license their Replica voices to studios and enterprise companies and earn a passive income, while they can still continue to work their regular day-to-day job of studio bookings and voice overs. So we're enhancing the ability for voice talent to monetize their skills.
I've interviewed the CEO of Replica and have been following their work. As a writer in the gaming media industry, I expect this tech to make waves once it starts to get adopted by game devs around the world.
Hello, very useful product however I do have some criticism:
1)Many of the accents sound incorrect they are noticeably either not by native English speaking actors or are by people who do not not actually have the accent they are attempting to mimic and I find many of them to be too awkward to use.
2) Lack of any Mexican or South American Accents. As a content developer in North America I find that very limiting.
3) Could use a better way to blend vowel and consonant sounds or a way to trigger alternate pronunciations of different . If it is not pronouncing words correctly I often have to trick it with an incorrect spelling or edit the recordings together afterwards to get what I want, which limits the ability to export directly to iClone.
4) Could use a variety of “resting” and filler sounds ie not speech , light breathing, heavy breathing, gasps, screams, laughs etc.