How do I download a Speech Studio Voice Gallery voice just for my TTS programs?

Question

I use text to speech programs to help my ADHD brain read long documents, like textbooks from back in college or contracts now that I'm a "real job" adult. The default TTS voices (Mark, David, and Zira) sound robotic and grating, but the "natural" voice option (Guy) is fairly good, but I'd like to get the "Andrew Multilingual" voice from the Voice Gallery to use with Microsoft Narrator, Microsoft Immersive Reader, Thorium eBook Reader, and other TTS programs that can use voices installed on my desktop. I'm having a lot of trouble figuring out how to do that or who even to contact about it. Anyone have any ideas?

Here's where my default TTS voices seem to be installed on my computer:

C:\Windows\Speech_OneCore\Engines\TTS\en-US

Here's the Voice Gallery I see the "Andrew Multilingual" voice in:

https://speech.microsoft.com/portal/5c1daaff7be04db382091243c540b073/voicegallery

Here's my device specificatons:

Device name Florian

Processor 11th Gen Intel(R) Core(TM) i7-11700 @ 2.50GHz 2.50 GHz

Installed RAM 16.0 GB (15.7 GB usable)

Device ID CBCA335A-19AC-47B2-95EA-53EF0F8F8188

Product ID 00342-20789-75709-AAOEM

System type 64-bit operating system, x64-based processor

Pen and touch No pen or touch input is available for this display

End goal:

I used the Kurzweil 3000 TTS program back in college, but that's expensive so today I use NaturalReaders.com instead. That's where I discovered the "Andrew Multilingual" voice. But that website is built primarily for PDFs and overrides any accessibility tagging with its own OCR data instead, and I want to read an EPUB file. EPUBs do work with NaturalReaders.com (it converts them to PDFs first), but it can be a bit tricky sometimes because of the whole "prioritize OCR over tags" thing. I know that the Thorium and Bookshelf eBook reader programs, built for EPUB, can do TTS well and can draw on downloaded TTS voices. I just want to be able to use this particular voice for that, instead of the default ones (David, Mark, Zira) or the "natural" option (Guy). But that requires downloading.

I'm happy to talk with a sales rep about paying for this download, and I'm happy to sign something that says I'll only use it myself. I just have no idea where to go or who to talk to about that.

Not end goal:

I am not trying to build/code this into a custom app, website, service, or plugin, even just for my own use. I know how to program for computation (like Excel or Stata) and graphic design (like Adobe InDesign), but I don't know how to program in any major languages like C-derivatives, Java, or Python. And building a custom plugin would require that skill set - one I don't have.

I am not trying to get an MP3 or other audio version of the text. The EPUB in question is a 600-page English-language textbook on German grammar - an audio-only format would not at all be helpful for my purposes.

I am not trying to use the David, Mark, Zira, or Guy TTS voices in relevant programs.

I am not trying to record voices for distribution (personal, home, education, commercial, or anything else). This is just meant for me, myself, and my desktop.

Answer 1

Andrew Welker oh man, i totally get u,

those default TTS voices can drive anyone nuts after a while )) Multilingual sounds way more natural, good choice, microsoft's speech studio voices are primarily meant for developers building apps with azure ai speech, not for direct desktop use. bummer, i know :/ Any way for microsoft u can't just download and install multilingual like a regular TTS voice pack (wish it was that easy!), but u can use it via azure ai speech SDK in apps that support custom endpoints.

Grab an azure subscription if u don't have one (there's a free tier, worth checking out). create a speech resource in azure portal docs here. in speech studio, u'll find multilingual under "neural voices". copy its voice name (something like "en-US-AndrewMultilingualNeural"). now in apps that allow custom TTS endpoints (some do!), u'd plug in your azure endpoint key + region + voice name.

downside? it needs internet connection since it runs on azure, not locally. and yeah, it's not free after the initial quota. but the quality? chef's kiss microsoft really nailed neural TTS lately %)

cross-platform alternatives )))) yes, bien sur we have it if u want offline voices that sound decent. Check out balabolka :)))) funny name yea it is freeware (https://www.cross-plus-a.com/balabolka.htm) with add-on voices like cereproc or acapela, they have natural sounding options too. this might help in other tools as well. amazon polly and google cloud TTS also have neural voices, but same deal - needs internet and some setup. as a wildcard, look at rvc (voice cloning) tools if u want really custom voices, but that's a rabbit hole :)

epub + TTS can be tricky! thorium should work with sapi5 voices, so any voice u install as sapi5 (like some ivona or nuance ones) could be an option. just saying )))))))))

thanks for the detailed question btw - u made it way easier to help! hope this points u in the right direction, even if it's not the one click solution we'd all want :D

Best regards,

Alex
and "yes" if you would follow me at Q&A - personaly thx.
P.S. If my answer help to you, please Accept my answer
PPS That is my Answer and not a Comment

https://ctrlaltdel.blog/