We have been working hard behind the scenes for the last 3 months, through extensive user feedback and power user adoption the Veritone Voice team has built an entirely new Voice experience. Today we are happy to release our biggest update to date, detailed below.

Lexicon Support

As we know not only synthetic voices, but also organizations, have their own unique personalities. Because of this, edge cases arises when specific words are not pronounced correctly. Typically this will happen with names and domain specific terminology. The good news is we have a way to solve for this: enter Lexicon support. Lexicons allow you to save phoneme and aliases against specific words. For example on a clip by clip basis, if you want all instances of D-M-H to be replaced with “Digital Media Hub” you can create an alias in your lexicon. When you want the engine to pronounce "tomatoes" as [TUH] + [MAA] + [TOHS] you can use the phonetic definition of təˈmɑːtoʊz.

Intonation Editor

The ability to have fine grained control over pitch changes has been a long requested feature. In order to assist users with this we have developed the Intonation Editor (IE). The IE allows you to select text and specify the amount of pitch change at specific locations within the text.

Preview In Place

We've also added the ability to preview a selected block of text, rather than previewing the entire document.

Autosave

As TTS becomes second nature for our users, the ability to pick up where you left off has become important. Veritone Voice now allows users to edit text post-clip creation with built-in autosave.

Reading Rules

Along with using phonemes, users frequently would like to instruct the TTS engine to pronounce words as more accurately as they would. Examples include spelling out a call to action such as, "Go to “Veritone.com!" This can instead be read as “ v-e-r-i-t-o-n-e dot c-o-m” to ensure listeners hear that exact address.

Speech to Speech

Veritone’s STS offering is now fully integrated with ReSpeecher, the Emmy Award winning engine provider and creator of the synthesized young Luke Skywalker’s voice for Disney+’s The Mandalorian. STS clip creation is now available within the Veritone Voice application with applicable licenses. Once the voice model is generated, the authorized user can create STS clips without dependency upon managed services.

Did this answer your question?