Voxabot

Voxabot

Text to speech editor with SSML and API connections

1 follower

Visit website
The Voxabot universal TTS SSML editor allows you use standard code for making speech. We have access to over 468 voices in 127 languages. We use the latest AI technology from the three major platforms: Amazon Polly, Google, and Azure Text-to-Speech.
Voxabot gallery image
Voxabot gallery image
Voxabot gallery image
Launch tags:
Text Editors•Internet of Things•Developer Tools
Launch Team

What do you think? …

Gilberto Segura
Maker
Hi Product Hunters! I'm happy to announce a universal TTS SSML Editor for your enjoyment. We've taken the problem of making Text to Speech and simplified the API's from Google, Microsoft and Amazon--now you can use those three platforms in one place to make great sounding speech using the top Neural TTS engines. We've got an editor that allows you to code SSML for all three platforms--and unlike any other, you can download your SSML code to use on your own. The editor comes with 20k free characters to use across any of the three engines; and if you plug in your own API Keys, that's between you and the TTS provider.
Plamen Petkov
@gilberto_segura For our product we are seeking for a good SSML editor (covering MS Azure text to speech SSML variation, but AWS and Google is also a good plus). It will be used in Angular. Can we contact and discuss this?
Troy Mork
Hi! I have created a virtual coach / guide for onboarding, coaching, guiding new users through our onboarding process and beyond onboarding to be a continuous presence in the user journey. The primary quality of our guide is to connect with people in the most authentic way. The key to opening that door and building rapport in this instance is the voice. I currently use Amazon Polly’s Neural engine ‘English, British Amy as I have found is the most realistic voice. Yet many key features are not available in the editor which makes writing scripts challenging to sound natural and not synthesized. I’m hoping your editor can help bring all the pieces together.
Troy Mork
@gilberto_segura I did an A/B test on your editor vs AWS using the same SSML script and engine: Amazon Polly, Neural Engine, Voice Female, English UK - Amy. The results are really very different. AWS Audio Quality and Voice Realism are significantly more realistic and rich. I'm curious to understand why this would be if you are in fact using the Neural Engine. Another curious component is that in the AWS editor some features are not available for this particular voice using the Neural engine (emphasis, whisper etc.) and yet your editor has these features available and poorly executed. This again brings into question what is causing the discrepancy in quality and tone. Heck, the pitch of the voice is completely different and in theory it should be the same. I do like your editor and would use that over the AWS editor 8 days a week. I look forward to understanding the challenges here.
Gilberto Segura
Maker
@gilberto_segura @troy_mork Hi Troy! I'm glad you liked it. We are using an API call method through the editor; apparently, the engine will need to include a "use Neural" for the few voices that provide that feature. Since that feature is only available in "us East" and for some sets of languages [not all languages, not all voices in a language] it's a little tricky. So to answer your question, we aren't using the Neural Engine in AWS because we aren't explicitly calling it--the features coming back for all other languages are at parity--we will have to address this for sure. Side note, since we are doing an API call, you can do entire sections beyond their character limit [in their editor] and also save/download the TTS/SSML to use later--so the editor could be used to stage the SSML, then you can post it in the Polly editor to allow use of the Neural engine; admittedly, a very clunky workaround.