Profile Picture

I have a need for Text to Speech software.

Posted By thomas_CWG 4 Years Ago
You don't have permission to rate!
Author
Message
thomas_CWG
thomas_CWG
Posted 4 Years Ago
View Quick Profile
Distinguished Member

Distinguished Member (1.0K reputation)Distinguished Member (1.0K reputation)Distinguished Member (1.0K reputation)Distinguished Member (1.0K reputation)Distinguished Member (1.0K reputation)Distinguished Member (1.0K reputation)Distinguished Member (1.0K reputation)Distinguished Member (1.0K reputation)Distinguished Member (1.0K reputation)

Group: Forum Members
Last Active: Last Year
Posts: 64, Visits: 356
I am sure i am not the only movie maker or games developer on a budget that has experienced this issue but there comes a time when you need to include spoken dialogue in your content. Hiring actors can be costly and unreliable.
Unfortunately for consumers like myself, there is a gap in the market for programs which offer realistic sounding text to speech that includes inflections and convincing human sounding speech. Of the ones that there are they usually have costly periodic subscription charges which i, and am sure many others, don't want to get tied into.

If Reallusion released a TTS program that had realistic human speech suitable for games development or movies i would be willing to pay $500 upfront for a license.
I hope somebody on the Reallusion team reads this and gives it some consideration.
Edited
4 Years Ago by thomas_CWG
animatedbeing
animatedbeing
Posted 4 Years Ago
View Quick Profile
Veteran Member

Veteran Member (852 reputation)Veteran Member (852 reputation)Veteran Member (852 reputation)Veteran Member (852 reputation)Veteran Member (852 reputation)Veteran Member (852 reputation)Veteran Member (852 reputation)Veteran Member (852 reputation)Veteran Member (852 reputation)

Group: Forum Members
Last Active: 3 days ago
Posts: 113, Visits: 1.7K
Google, Amazon, and Microsoft offer text to speech cloud services. They support speech markup for pronunciation, mood, pitch, etc. The cost range is from zero to a few dollars.
https://cloud.google.com/text-to-speech/
https://aws.amazon.com/polly/
https://azure.microsoft.com/en-us/services/cognitive-services/text-to-speech/

You can do text to speech in Windows and OSX but the functionality is limited.
JimmyB7
JimmyB7
Posted 4 Years Ago
View Quick Profile
Distinguished Member

Distinguished Member (2.4K reputation)Distinguished Member (2.4K reputation)Distinguished Member (2.4K reputation)Distinguished Member (2.4K reputation)Distinguished Member (2.4K reputation)Distinguished Member (2.4K reputation)Distinguished Member (2.4K reputation)Distinguished Member (2.4K reputation)Distinguished Member (2.4K reputation)

Group: Forum Members
Last Active: 3 Years Ago
Posts: 198, Visits: 3.2K

Here’s a cheap, easy way I’ve found using Microsoft Edge Browser. Edge has a built in "Read Aloud" Text to Speech reader using whatever voices you have on your computer.  There are some other free voices out there that you can add too. I also use Audacity (free audio software) to record the speech using “Stereo Mix” setting in the Window’s sound settings.  Here is a tutorial by Freedom Arts that shows how to add voices to Windows.

https://www.youtube.com/watch?v=-nbKWJ8ovcM&t=39s



CyberpowerPC, 64 GB Ram, 64bit Windows 10 Pro, Intel Core i9-10980XE 18 Core CPU,  ASUS TUF X299 Mark 1 Motherboard, ROG-STRIX-RTX2080TI-O11G-GAMING graphics card, Toshiba 3TB hard drive, 3 x Seagate 3TB hard drives, RAGNAROK 801 Full Tower, LEPA G1600 1600W power supply, Killer WI-FI 6,  ASUS PA329Q 32" 4K/UHD IPS Monitor.
Edited
4 Years Ago by JimmyB7
animagic
animagic
Posted 4 Years Ago
View Quick Profile
Distinguished Member

Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)

Group: Forum Members
Last Active: 5 hours ago
Posts: 15.6K, Visits: 30.2K
I use Google WaveNet and I'm quite pleased with it. You get like a million characters free each month, so that's plenty.


https://forum.reallusion.com/uploads/images/436b0ffd-1242-44d6-a876-d631.jpg

TimothyMasters
This post has been flagged as an answer
TimothyMasters
Posted 3 Years Ago
View Quick Profile
Distinguished Member

Distinguished Member (2.1K reputation)Distinguished Member (2.1K reputation)Distinguished Member (2.1K reputation)Distinguished Member (2.1K reputation)Distinguished Member (2.1K reputation)Distinguished Member (2.1K reputation)Distinguished Member (2.1K reputation)Distinguished Member (2.1K reputation)Distinguished Member (2.1K reputation)

Group: Forum Members
Last Active: Last Year
Posts: 436, Visits: 801
Thank you for these links.  At the risk of embarrassing myself, I'll share my experience, in case anyone is interested.
Within 15 minutes of going to Amazon Polly I was creating MP3 speech files from text I pasted into a cloud portal.  Very fast and easy to use.

I spent over an hour navigating the Microsoft site, going through countless intermediate web pages, often wildly guessing at what links to click to progress forward on TTS.  Along the way, I learned that after 30 days of free usage, I would have to explicitly REMOVE ALL SPENDING LIMITS ON MY CREDIT CARD in order to proceed with more usage, even in the free tier, just trusting them to not charge me anything.  When I finally found my way to the no-programming console page, I learned that I would have to install on my computer multiple pieces of Microsoft software in order to proceed.  Their instructions as to what to install, based on my computer's configuration, were cryptic.  No way.

I spent about three hours navigating the Google site, trying to find text-to-speech.  You'd think it would be easy, but none of the pages were clear.  I never did find text-to-speech.

Bottom line: I'm using Amazon Polly and finding it very easy to use and reasonably decent quality.

PS: I wish, as another member suggested, that Reallusion would form a partnership with a quality TTS provider.  Reallusion's visual capabilities are excellent, but those of us who want to create videos in which characters speak are left behind.





Data Juggler
Data Juggler
Posted 3 Years Ago
View Quick Profile
Distinguished Member

Distinguished Member (12.3K reputation)Distinguished Member (12.3K reputation)Distinguished Member (12.3K reputation)Distinguished Member (12.3K reputation)Distinguished Member (12.3K reputation)Distinguished Member (12.3K reputation)Distinguished Member (12.3K reputation)Distinguished Member (12.3K reputation)Distinguished Member (12.3K reputation)

Group: Forum Members
Last Active: Yesterday
Posts: 1.7K, Visits: 5.3K
I use Cereproc.com sometimes.

Mostly I learned to train my voice because I got tired of paying narrators. After a little practice I found I have a pretty wide range of characters as long it is from a squeaky mouse to a female or a man that was never a lumberjack on a battle field.

James Earl Jones or Jeremy Irons I am not. There are some good YouTube videos on narration.

That is my reason for thinking self driving cars are not ready, if voices can't be generated.

Some new software has been developed for deep fakes, so soon you should be able to do this in something commercial:

(Sorry if it shows you an add, but this channel is a good channel. 2 Minute Papers on AI, I don't mind supporting him as much as Google).





Creator of the free website:
PixelDatabase.Net
A Free Online Text Based Image Editor
https://forum.reallusion.com/uploads/images/6e6663e2-0ecf-447a-ab3d-d49d.png

animagic
animagic
Posted 3 Years Ago
View Quick Profile
Distinguished Member

Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)

Group: Forum Members
Last Active: 5 hours ago
Posts: 15.6K, Visits: 30.2K
Google WaveNet uses an API which you need to subscribe too. I started with Polly, but I've found WaveNet better. Once set up it's easy to use. 

Traditional TTS providers are extremely restrictive in what you can do with the speech. RL used to have relationships with those, but it was costly and not worth it.

There also was an online TTS service but RL stopped offering that. Sometimes it better to stick with your core offerings.


https://forum.reallusion.com/uploads/images/436b0ffd-1242-44d6-a876-d631.jpg

gordryd
gordryd
Posted 3 Years Ago
View Quick Profile
Distinguished Member

Distinguished Member (4.5K reputation)Distinguished Member (4.5K reputation)Distinguished Member (4.5K reputation)Distinguished Member (4.5K reputation)Distinguished Member (4.5K reputation)Distinguished Member (4.5K reputation)Distinguished Member (4.5K reputation)Distinguished Member (4.5K reputation)Distinguished Member (4.5K reputation)

Group: Forum Members
Last Active: Yesterday
Posts: 696, Visits: 7.4K
@ animagic - a few questions about Google Wavenet (edit - I found the Google page and got the answers - below for anyone who cares):
Can you type English into an international voice (say, German) and will it speak English with a German "accent"?  Or does it only accept German input for German voices?
 Answer:  If you select an international voice but type in English, it will speak English with an "accent" - this gives you a LOT of variability.  Some of the pronunciations can be a little wonky, but that adds realism, IMHO...
Can you vary things such as pitch, speed, emphasis, pause, etc.?\
 Answer:  I saw that it accepts normal SSML commands
Thanks.



Reallusion Certified Director / Reallusion Best Visual Award / Reallusion Certified Content Developer
See all the G-Tools Plug-Ins available for iClone here: G-Tools Plug-Ins


Edited
3 Years Ago by gordryd
animagic
animagic
Posted 3 Years Ago
View Quick Profile
Distinguished Member

Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)Distinguished Member (32.3K reputation)

Group: Forum Members
Last Active: 5 hours ago
Posts: 15.6K, Visits: 30.2K
In general, a TTS doesn't really care what language the text is in, except that you have to rewrite the original text so that the words produce the desired phonemes in the target language.

This can be a challenge but also a lot of fun. I've done it for French in one project.


https://forum.reallusion.com/uploads/images/436b0ffd-1242-44d6-a876-d631.jpg

Edited
3 Years Ago by animagic



Reading This Topic