Wątek: How did the real names of the siri voices come about?

Strona: 2 z 5

21 z 84: bomberman29

there is an addon for older nvda, if you have a mac, you can generate apple voices from a local host and play it i don't know where
my github: https://github.com/denizsincar29
26.08.2020 09:10

22 z 84: Enes

As far as I remember, that one was just some add-on that would play a bunch of words whenever NVDA spoke them. So it was far from being a completely functional synthesizer. And I highly doubt it has ever been recompiled to be Python3-compatible.
--Cytat (radio729):
No, but there was an Alex NVDA add-on but that was made as a joke. It didn't work that well anyway.

--Koniec cytatu
Enes Find my sounds at: https://freesound.org/people/sonically_sound/
26.08.2020 21:29

23 z 84: etno

It is possible to use Siri voices with Vocalizer AddOn. I found that on anyaudio.net and I think a friend of me still has atleast german voices storaged on his computer.
Religion ist Opium fürs Volk!
29.08.2020 11:14

24 z 84: Angel

The one I would actually use on NVDA would be Nora, she is the only one I like on my phone and don't have on NVDA.

01.09.2020 13:03

25 z 84: Dark_Raven

While the old siri voices worked with old nvda at least, there are two things to remember if you still have this siri package.

1. These siri voices are only compatible with what is no known as "Vocalizer Expressive V1", meaning your driver needs ve.dll of version 1.3 or newer.
The voices are multiple files per voice and thus won't work with what is now known as "Vocalizer Expressive V2", in which voices are generally one file per voice and ve.dll has version numbers from 2.1.0 up to 3.0 and beyond.

2. The voices are only available in the quality "Premium" or "Compact".
Presumably there are actual Premium High variants of them, but if so they haven't been seen so far.
Other Vocalizer voices for JAWS and NVDA had Premium High variants for most of them.

And it is unknown if the IOS 13 or 14 voices are now all in the format like the current "Vocalizer Expressive 2" voices for JAWS, or if they are mixing old siri voices with new regular ones for IOS.

01.09.2020 18:24

26 z 84: Enes

1. The new iOS Siri voices have no compact variants.
2. The old NVDA Siri voices can definitely be used with the Vocalizer Expressive 2.2 driver. Tried it myself; and they work perfectly fine.
--Cytat (Dark_Raven):
While the old siri voices worked with old nvda at least, there are two things to remember if you still have this siri package.

1. These siri voices are only compatible with what is no known as "Vocalizer Expressive V1", meaning your driver needs ve.dll of version 1.3 or newer.
The voices are multiple files per voice and thus won't work with what is now known as "Vocalizer Expressive V2", in which voices are generally one file per voice and ve.dll has version numbers from 2.1.0 up to 3.0 and beyond.

2. The voices are only available in the quality "Premium" or "Compact".
Presumably there are actual Premium High variants of them, but if so they haven't been seen so far.
Other Vocalizer voices for JAWS and NVDA had Premium High variants for most of them.

And it is unknown if the IOS 13 or 14 voices are now all in the format like the current "Vocalizer Expressive 2" voices for JAWS, or if they are mixing old siri voices with new regular ones for IOS.

--Koniec cytatu
Enes Find my sounds at: https://freesound.org/people/sonically_sound/
04.09.2020 12:53

27 z 84: Enes

So the new Siri voices are completely different from the previous ones, as far as I know.
--Cytat (Enes):
1. The new iOS Siri voices have no compact variants.
2. The old NVDA Siri voices can definitely be used with the Vocalizer Expressive 2.2 driver. Tried it myself; and they work perfectly fine.
--Cytat (Dark_Raven):
While the old siri voices worked with old nvda at least, there are two things to remember if you still have this siri package.

1. These siri voices are only compatible with what is no known as "Vocalizer Expressive V1", meaning your driver needs ve.dll of version 1.3 or newer.
The voices are multiple files per voice and thus won't work with what is now known as "Vocalizer Expressive V2", in which voices are generally one file per voice and ve.dll has version numbers from 2.1.0 up to 3.0 and beyond.

2. The voices are only available in the quality "Premium" or "Compact".
Presumably there are actual Premium High variants of them, but if so they haven't been seen so far.
Other Vocalizer voices for JAWS and NVDA had Premium High variants for most of them.

And it is unknown if the IOS 13 or 14 voices are now all in the format like the current "Vocalizer Expressive 2" voices for JAWS, or if they are mixing old siri voices with new regular ones for IOS.

--Koniec cytatu

--Koniec cytatu
Enes Find my sounds at: https://freesound.org/people/sonically_sound/
04.09.2020 12:55

28 z 84: Dark_Raven

They might work with the new drivers.
But I am not sure if you can for example install old siri voices and current new Vocalizer voices for the same language and let them run together with one Vocalizer engine.

On the voices page for JAWS, they state that you can't use Vocalizer Expressive 1 and 2 voices together on the same computer.

They claimed that you would not be able to install version 1 Daniel and version 2 Serena in the same installaton and switch between at will within JAWS.

You would be required to use both voices in either version, but not accross versions within the same instance of the engine.

08.09.2020 17:40

29 z 84: Seva

Which language is this voice for?
-- (Angel):
The one I would actually use on NVDA would be Nora, she is the only one I like on my phone and don't have on NVDA.

--
-- Seva https://github.com/sevapopov2
09.09.2020 11:33

30 z 84: hozosch

English US, new since IOS 11.

09.09.2020 11:34

31 z 84: Enes

What I said also applies to ordinary Vocalizer voices.
I just added the two Vocalizer 1 Japanese voices to my list of Vocalizer 2 voices; and now they work perfectly fine.
I'm using all of them with one single driver.
--Cytat (Dark_Raven):
They might work with the new drivers.
But I am not sure if you can for example install old siri voices and current new Vocalizer voices for the same language and let them run together with one Vocalizer engine.

On the voices page for JAWS, they state that you can't use Vocalizer Expressive 1 and 2 voices together on the same computer.

They claimed that you would not be able to install version 1 Daniel and version 2 Serena in the same installaton and switch between at will within JAWS.

You would be required to use both voices in either version, but not accross versions within the same instance of the engine.

--Koniec cytatu
Enes Find my sounds at: https://freesound.org/people/sonically_sound/
09.09.2020 14:46

32 z 84: Dark_Raven

What is the exact version of the NVDA driver you use and what is the exact version number of ve.dll, if you can actually combine all voice versions and variants?

Back to "normal" vocalizer, do you know where you can get current version premium high variants of voices like Noelle and Joelle (US English)?

Theese should be found in the variant Embedded-High on https://harposoftware.com/en/2-main/s-1/index/category-all_voices/language-english_usa.

I think latest CodeFactory voices for NVDA are also only Embedded-High and no up to date Premium High are there, except if you use older versions of the CodeFactory addons.

And finally, do you know where you can get voices like Ava ML, Zoe ML, Rishi ML,Anna ML or Petra ML?

Theese apparently can be found here: https://voxin.oralux.net/voice.php

However the sold product is not for Windows and if you look at the list, again nothing better than "High".

But you probably know that most voices are available in Premium High also in older CodeFactory versions and as part of the JAWS voices.

10.09.2020 00:34

33 z 84: Enes

Driver version: 2020.02.27
ve.dll version: 2.2.1.16181
As for Noelle and Joelle, I think I made some add-on using the files contained in the installer that can be downloaded from the page you provided. Sorry, I don't really remember how I did it... I can attach my add-on here if you wish.
I tried to install the Code Factory voices and use them with the driver I had as well, but they didn't work.
And the "ML" thing I have never heard of before. I will examine the link you posted, and inform you of whether I come up with anything useful.
To download and use premium-high variants of the old voices, you can just download the premium-high add-on of any voice intended for Vocalizer 1, change the file extention to .zip, edit 'manifest.ini' to make the add-on compatible with newer versions of NVDA, and then repack and install the add-on to be used with the Vocalizer 2 driver. You can also just copy the folder directly to the NVDA add-on directory.
--Quoted (Dark_Raven):
What is the exact version of the NVDA driver you use and what is the exact version number of ve.dll, if you can actually combine all voice versions and variants?

Back to "normal" vocalizer, do you know where you can get current version premium high variants of voices like Noelle and Joelle (US English)?

Theese should be found in the variant Embedded-High on https://harposoftware.com/en/2-main/s-1/index/category-all_voices/language-english_usa.

I think latest CodeFactory voices for NVDA are also only Embedded-High and no up to date Premium High are there, except if you use older versions of the CodeFactory addons.

And finally, do you know where you can get voices like Ava ML, Zoe ML, Rishi ML,Anna ML or Petra ML?

Theese apparently can be found here: https://voxin.oralux.net/voice.php

However the sold product is not for Windows and if you look at the list, again nothing better than "High".

But you probably know that most voices are available in Premium High also in older CodeFactory versions and as part of the JAWS voices.

--End of Quote
Enes Find my sounds at: https://freesound.org/people/sonically_sound/
11.09.2020 14:22

34 z 84: Dark_Raven

Here is the only ml voice I know of, but it was released for the official Vocalizer for NVDA driver in the 3.0 range, so actually using multilanguage features would never have worked with the original ve.dll which is version 1.1X:
http://www.tiflotecnia.net/ftp/vocalizer-voice-petra-PremiumHigh.nvda-addon

But this is obviously the old version and not a version 2 voice.

About CodeFactory:
redownload the NVDA addon, install it an redownload any voice, in the current version you should be able to use the files out of the product as well.
But beware that for some reason the new CodeFactory voices use a ve.dll version > 3.0 and all voices exist only in the Variants Compact, Embedded-Pro and Embedded-high.


About voices like Joelle: the Versions I have seen are in Embedded-High only, but no Premium High variant.

And most current voices from Harposoft were recently updated, so that even voices like Daniel were downgraded to Embedded-High, where we know that Premium High exists going by the JAWS voices.

Did you make the driver yourself, or where did you get a current version from?

I am asking because Tiflotecnia hasn't updated their download page with a version of 3.X for current NVDA and neither have they released any Version 4 driver on their own (official) page at https://vocalizer-nvda.com/downloads.

So if you can link to the driver if possibly for both NVDA branches, I could experiment with some voices I can get my hands on.

11.09.2020 16:56

35 z 84: Enes

I will unfortunately be unable to share a link to the driver as I am unsure of its legal status.
Enes Find my sounds at: https://freesound.org/people/sonically_sound/
21.09.2020 09:32

36 z 84: Dark_Raven

Forget the driver for a moment.

How can any driver accept two voices of different versions and file structures within one instance meaning for example one installation with V1 Serena and v2 Daniel?

Surely that's against the software's rules for handling the engine core and data files.

And moreover, how do theese things work with the multilanguage voices?

And what about conversational and deep learning variants of voices apparently offered for sale on nextup.com, allthough they are apparently under a different brand name.

And it does not explain why Codefactory voices for NVDA have been downsiced from Premium-High to Embedded High in subsequent updates.

If you can't get your hands on CodeFactory's files, compare the Harpo voices with the ones from the downloads page for the JAWS voices if you do need hard evidence.

22.09.2020 20:18

37 z 84: TheBlindTechGuy

hi.
If there is any way to get the nvda addon for the old siri voices, please let me know.
I love the nicky voice and I would love to have it on nvda.


18.03.2023 21:46

38 z 84: marchoffmann

People tryed them it's out of date and not working anymore.
Ist euch denn nicht klar , dass für Menschen , die Unrecht tun, in Gottes Reich kein Platz sein wird? Täuscht euch nicht: Wer sexuell unmoralisch lebt, Götzen anbetet, die Ehe bricht, wer sich von seinen Begierden treiben lässt und homosexuell verkehrt, wird nicht in Gottes Reich kommen. 1. Korinther 6, 9
18.03.2023 21:52

39 z 84: TheBlindTechGuy

is there anyway I could get the files and see what I could do with it?

18.03.2023 21:54

40 z 84: Enes

Who is people? One should still be able to use those voices with the appropriate Vocalizer drivers compatible with later versions of NVDA, as I can. Well, not sure if it's legal though.
-- (marchoffmann):
People tryed them it's out of date and not working anymore.

--
Enes Find my sounds at: https://freesound.org/people/sonically_sound/
19.03.2023 15:36