Just create an account or login with your existing Google/Facebook one, start recording the sentences given to you, and then you will be allowed to make your TTS voice once your recording count reaches 30 (each recording containing a sentence), and the bigger your recording database the better the voice. Once you get your voice ready, you may write any text into the edit box and it will be read aloud by your digital avatar. Note this service is provided as freeware as of now, though you may need to pay for additional features such as a personalized SAPI version of your voice or a more natural, human-like version of it, but not all features are available since the website is still undergoing massive changes as it is still in beta I suppose.
Hmm lyrebird, but their stuff is not so good, it only speaks until a specific ammount of time i don't remember, how ever you can find them at lyrebird.ai.
No, lyrebird is so great for the service it provides IMO.
but it can't read long texts.
At one time my friend was playing with something called Tacotron. It is a library that uses artificial intelligence to synthesize speech. The advantage is that you can use it to generate speech in any language. Sufficient audio samples divided into sentences and the same sentences writing in the text. It requires quite a lot of computer resources, and the thing is rather for programmers, but maybe someone will be able to somehow implement it.
#7 (edited by rory-games 2019-07-04 04:47:55)
I just tried lyrebird, absolutely awesome. now creating, rory version 1.1!
edit: this is truely amazing. here comes version 1.4, lets go!
3.0 of my voice done! uploading a sample when I'm not lazy...
that's lyrebird. and it sounds much better now because he was using an older beta and only had a hundred sentences; I have 300 on mine
#11 (edited by janagirl 2019-07-05 11:29:21)
I myself would really like to create a german sapi voice which I can use with my screen reader. That would be my biggest dream. But sadly, no one could help me with this. Reading sentences and then make a sapi voice with the read sentences.
what about mk voice or serkem reality?
Where to find MK voice or serkem reality? I can't find anything via google.
#14 (edited by akse0435 2019-07-11 18:45:24)
The reason why you can't find anything on Google is, that it's called Circum Reality, not Serkem Reality.
Here is some links to download it:
The program is packed in a 7z-file, so you'll need 7-zip to extract it.
You can download it here:
As the voice says, he only recorded 600 Words, so it isn't as good as it could have been, I guess.
When you've made a voice, you can use it as a normal SAPI voice for your applications.
However, I've no idea how to use this program, so I'll recoment searching on Google for it.
first of all, thank you very very much for the links you posted.
What a pity that you don't know how to use it, I would have so many questions.
I managed to speak a few sentences and I could hear the voice while testing in the program itself. But if I'd like to use it in an aplication it doesn't work.
And, is there any way to make a german voice, too? I really dream of making such one, my language is german and it would be much easier for me to read the sentences.
Perhaps someone can help me.
as he said, he has no idea how to use it.
What I've found out is that you can move through the items with the tab key.
I don't know which screenreader you use but I have to shift+tab out of the application and shift+tab back and then I can see what the cursor is located on. And I use space to select an item.
But it could be much more accessible. The sentence you have to read is shown very short. But there's a file called "englishprompts.txt". So I open it and search for the sentence I have to speak, go back to the program, press space to record and go back to the txt-file to read it.
And for me, the result is that the voice doesn't work! I don't know if I missed something or whatever.
well, you could always use one of the decTalk speak windows and learn how to make it sing, make telephone dial sounds, and different kinds of tones. And with commands in your text you can greatly modify how it speaks. It works best with jaws. But if using jaws, make sure to turn the older accessibility driver to on, under Miscellaneous in your default jaws configuration file. the best speak window is version 4.3 of decTalk. and it works very good in windows 10 even on my $114 RCA cambio tablet with intel atom z8350 processor, well its a netbook tablet basically. I plan on getting a much better computer shortly after the new year in January of 2020.