Hello, I'm looking to use an AI voiceover app for my courses, as I'm working remotely full time now and don't have access to great voiceover volunteers at my office any longer. Anyone have any good recommendations or reviews? I am looking at LOVO and Well Said Labs. Thank you!
Have you tried Articulate 360 ? The built in AI voiceover text to voice works excellent and allows a good deal of customization. I use it on my courses and have been very happy so far. Its built into the system too. What functions do you need it for that would require outside voiceover/narration?
Hello, Allison! My name is Dean T Moody, and I am a professional Voiceover artist. I specialize in ELearning and Online Training projects. I'd love to help you out! I have demos at www.deantmoodyvoice.com, including eLearning demos. Let's chat!
Thank you for the info!
I've been actually trying a free trial with them, but the "conversational" functionality just isn't there- for example, I can't get my characters to ask questions the way I want them to, or ask questions, period. The WellSaid representative admitted the technology is just not there yet. But for narration, it is nice!
I've been using Balabolka - free text-to-speech program which can use some of the MS voices, as well as third-party purchased voices. You could also check out MS Word's Narrator feature and capture it with Audacity or some other s/w.
Still a bit robotic, so if you're looking for a truly human-sounding voice these won't get you there...
I've been using AWS POLLY on the Amazon Web Service, its not perfect but its better than the built in one. (free for up to 1 million characters a month.)
Down side is that it looses the CC option, but you can put that in as well and silence the generated audio file so you still get CC.
Is Amazon Polly a high learning curve? (There were a number of 1 hour videos on the site; and it looked like it required scripts to be inserted for pauses.)
Also, for the interface, does it allow a grouped import and export? (Such as importing an entire module at once, then the clips are separated based on paragraphs?
Re- opening this comment thread because we are having issues with Rise appending AI generated files. Rise is reading the first sentence of the "script" and using that in the file title and we are ending up with titles that are over 30 characters long! Our LMS is kicking them back. How can we stop Rise from reading that content and saving the files in this way?
I'm using SL360 and the only choices for text to speech are now showing as neural. When I add SSML codes I get an error saying the voice may not support the codes or the codes are wrong. I've verified my codes but wonder if I'm doing something wrong as the list of SSML codes apply to Neural voices. Anyone have insight to using the SSML codes?
My company used PlayHT in the past. Their prices were reasonable, and had a wide selection of voices (including dialects). That was a couple of years ago. Im sure technology has improved since. We were using Lectora at the time which did not have a text to speech. Something I realized after looking at other applications, many are using the same voices. Articulate's stock voices have improved with the last update. However, there is a noticeable different.
Regardless of what software is being used, I still think its easy to hear when its AI talking. For shorter classes I use Voice.AI. It allows the user to speak in famous voices. Morgan Freeman, James Earl Jones, and William Shatner sound good.
We use Well Said Labs and have not had any issues. They have introduced different languages this year. I have not used that as the contractor we use for all of our translations takes care of that for us. I will explore Voice.AI I'd love to have James Earl Jones facilitate training. :-)
Which are you referring to? The stock voices Articulate has? The last update brought a couple of additional options. The higher tier companies sound like they are all using the same sources for their AI.
Specifically, how Rise takes these AI generated audio files and makes them so long that our LMS kick them back when we upload. We use Play.ht and that is how I discovered this issue. I plan to test Storyline's AI audio voices and see if Rise does the same thing. Have you noticed a difference in the file title length Rise generates between Play.ht and Storyline?
A really nice free one is Microsoft Clipchamp. Many good voices and great settings to adjust the voice. I have used it on several projects. you will need to google how to open the text to speech. after that, very intuitive.
I appreciate all the recommendations and they are less expensive than play.ht- so I will definitely be checking them out. However, I would really appreciate a response from the Articulate crew about what they can do about Rise and how it adds all these extra characters. It seems to be a thing that has impacted people and their LMSs for a while now. They know this creates an issue no matter what tool is used. Articulate.....do you have a response?
Thanks for sharing your experience regarding the audio filename length in Rise 360; I'm sorry you ran into this!
Our engineers are currently looking into shortening the filenames in the published output, so I'll go ahead and include your voice in the report. We'll let you know once this change is available. In the meantime, we're happy to troubleshoot your course for any tailored options. Feel free to connect with us in a case using this link.
25 Replies
Have you tried Articulate 360 ? The built in AI voiceover text to voice works excellent and allows a good deal of customization. I use it on my courses and have been very happy so far. Its built into the system too. What functions do you need it for that would require outside voiceover/narration?
My SMEs don't like the text to voice in Articulate. It's too robotic sounding. I wish they did because it'd be so easy!
Thanks, Allison
Allison Znachko
Instructional Designer
Hello, Allison! My name is Dean T Moody, and I am a professional Voiceover artist. I specialize in ELearning and Online Training projects. I'd love to help you out! I have demos at www.deantmoodyvoice.com, including eLearning demos. Let's chat!
This is the very best AI voice site out and very cost effective.
https://wellsaidlabs.com/
Thank you for the info!
I've been actually trying a free trial with them, but the "conversational" functionality just isn't there- for example, I can't get my characters to ask questions the way I want them to, or ask questions, period. The WellSaid representative admitted the technology is just not there yet. But for narration, it is nice!
Allison Znachko
Instructional Designer
Good point as I did use it for a narration.
I've been using Balabolka - free text-to-speech program which can use some of the MS voices, as well as third-party purchased voices. You could also check out MS Word's Narrator feature and capture it with Audacity or some other s/w.
Still a bit robotic, so if you're looking for a truly human-sounding voice these won't get you there...
This post was removed by the author
I've been using AWS POLLY on the Amazon Web Service, its not perfect but its better than the built in one. (free for up to 1 million characters a month.)
Down side is that it looses the CC option, but you can put that in as well and silence the generated audio file so you still get CC.
We use well said labs. Its been very good and they make updates to help improve performance
Wellsaid Labs.
We use Speechelo.
Is Amazon Polly a high learning curve? (There were a number of 1 hour videos on the site; and it looked like it required scripts to be inserted for pauses.)
Also, for the interface, does it allow a grouped import and export? (Such as importing an entire module at once, then the clips are separated based on paragraphs?
Thanks for your help.
Re- opening this comment thread because we are having issues with Rise appending AI generated files. Rise is reading the first sentence of the "script" and using that in the file title and we are ending up with titles that are over 30 characters long! Our LMS is kicking them back. How can we stop Rise from reading that content and saving the files in this way?
I'm using SL360 and the only choices for text to speech are now showing as neural. When I add SSML codes I get an error saying the voice may not support the codes or the codes are wrong. I've verified my codes but wonder if I'm doing something wrong as the list of SSML codes apply to Neural voices. Anyone have insight to using the SSML codes?
My company used PlayHT in the past. Their prices were reasonable, and had a wide selection of voices (including dialects). That was a couple of years ago. Im sure technology has improved since. We were using Lectora at the time which did not have a text to speech. Something I realized after looking at other applications, many are using the same voices. Articulate's stock voices have improved with the last update. However, there is a noticeable different.
Regardless of what software is being used, I still think its easy to hear when its AI talking. For shorter classes I use Voice.AI. It allows the user to speak in famous voices. Morgan Freeman, James Earl Jones, and William Shatner sound good.
We use Well Said Labs and have not had any issues. They have introduced different languages this year. I have not used that as the contractor we use for all of our translations takes care of that for us. I will explore Voice.AI I'd love to have James Earl Jones facilitate training. :-)
This is the tool my company is using that Rise places all these characters in after publication. UGH!
Which are you referring to? The stock voices Articulate has? The last update brought a couple of additional options. The higher tier companies sound like they are all using the same sources for their AI.
Specifically, how Rise takes these AI generated audio files and makes them so long that our LMS kick them back when we upload. We use Play.ht and that is how I discovered this issue. I plan to test Storyline's AI audio voices and see if Rise does the same thing. Have you noticed a difference in the file title length Rise generates between Play.ht and Storyline?
Try Techsmith Audiate
A really nice free one is Microsoft Clipchamp. Many good voices and great settings to adjust the voice. I have used it on several projects. you will need to google how to open the text to speech. after that, very intuitive.
If you have an Azure subscription, then Microsoft's text-to-speech is one of the best I've tried. Their neural voices are amazing.
I appreciate all the recommendations and they are less expensive than play.ht- so I will definitely be checking them out. However, I would really appreciate a response from the Articulate crew about what they can do about Rise and how it adds all these extra characters. It seems to be a thing that has impacted people and their LMSs for a while now. They know this creates an issue no matter what tool is used. Articulate.....do you have a response?
Hello Jennifer,
Thanks for sharing your experience regarding the audio filename length in Rise 360; I'm sorry you ran into this!
Our engineers are currently looking into shortening the filenames in the published output, so I'll go ahead and include your voice in the report. We'll let you know once this change is available. In the meantime, we're happy to troubleshoot your course for any tailored options. Feel free to connect with us in a case using this link.