Go to Menu

HOW OUR TEXT TO SPEECH VOICES ARE MADE

To create our outstanding text to speech voices, the ReadSpeaker TTS Development Team carefully assesses the most suitable professional voice talents for every TTS persona. A complex selection process is carefully carried out to guarantee that only the highest-quality synthetic speech is rolled out. The initial stage involves a meticulous assessment of the artistic, commercial and technical characteristics of each candidate speaker voice, which means monitoring a comprehensive set of criteria. The voice talents ReadSpeaker collaborates with are usually actors or trained professional speakers.

Individual performance parameters for each voice candidate are measured, compared and a shortlist is drawn up. Once a voice talent has been selected, he or she works alongside ReadSpeaker staff for several weeks. During this period, our voice talent records a demanding script designed to capture all the characteristics needed to cover the sound patterns of the language we’re developing, down to the finest detail. This script is recorded in the speaker’s own, natural speech. The ReadSpeaker voice development team monitor the recording process closely to detect the slightest deviation in pronunciation, accentuation, or style.

The outcome of this process is a massive collection of voice recordings, thoroughly and repeatedly monitored for consistency. If even the slightest glitch is detected, it’s back to the recording studio.

In the second phase of TTS voice creation, the speech recordings obtained are processed and a rich mark-up is created, in which each word, phoneme and stress is annotated along with several other aspects. The recordings and their annotations are combined into a voice database used by the core VoiceText engine.

During the annotation process, the technical team works its magic – using a powerful combination of Artificial Intelligence and machine learning technologies on big amounts of data. Our state-of-the-art methodologies are augmented by means of expert linguistic capability.

This is when a new ReadSpeaker TTS voice persona is born. But it’s also when we scrutinize the new TTS voice and mother-tongue experts rate it in minute detail in a thorough Quality Assurance process. Any errors and imperfections are corrected.

One of ReadSpeaker’s unique-to-the-market characteristics is our ongoing improvement process. Our TTS software keeps learning even after it has been launched, thanks to an optimal feedback and updating system.

At ReadSpeaker, we have a passion for developing the most advanced, high-quality TTS voices. In fact, expert third party industry observers rate ReadSpeaker TTS voices as being the most accurate on the market. The enthusiastic feedback we receive from our customers confirms that we deliver the very best TTS solutions for successful online, offline, embedded and server-based applications around the world.

Our commitment to providing outstanding TTS solutions is made possible by our uncompromising production process, designed to guarantee the quality levels that have earned ReadSpeaker TTS the trust of customers from across countries and markets.


CUSTOM VOICES

If your strategy is to offer an exclusive customer experience and you want to take your brand appeal to a whole new level, one of the most powerful ways to differentiate yourself is by using a custom voice to represent you. A custom voice sets your brand apart, and creates a powerful bond with your customers across your various communication touch points. If a preferred celebrity or other voice talent reflects your brand best and you want to be able to use their voice any time you need it, ReadSpeaker can create a Custom TTS voice for you.

ReadSpeaker’s Voice Text engine leverages a unique combination of text normalization, lexical lookup, grapheme to phoneme modelling, prosody modelling, manuscript creation, recording of voice talent and acoustic database creation to ensure that the outstanding, lifelike quality of VoiceText TTS voices enhances the user experience across all markets. As long as we can get your voice talent in the studio, we can create a custom voice for your exclusive use.

Fields of application include cars, customer contact centers, digital assistants and chatbots, e- and m-commerce, electronic devices and appliances.