Using AI Voiceover

By clicking on "AI Voiceover" on the left navigation bar, you access the voiceover creation page, or "Text-To-Speech" (TTS). The voiceover function allows an artificial voice to read the text provided by the user. To use this service, you need to have purchased a top-up or have subscribed to a plan that includes available "characters".

Of course, the user can customize all desired aspects: language, gender and type of voice, and other parameters.

First, select one of the dozens of available languages. Then, choose one of the voices associated with the selected language. You can listen to a free preview of the chosen voice by clicking on the speaker icon next to the selected voice. In the box right after, you can enter the name of the file with which you want to save your work and finally the reference workbook.

In the next line, you can select the format in which you want to synthesize the voice: Mp3, Wav, Ogg, and WebM.

Then there are the parameters you want to assign to the voice. "Speaking Style" allows you to choose the tone of voice; "Say As" specifies how a certain string should be read; "Volume" allows you to increase or decrease the prosodic tone of the reading; “Speed” increases or decreases the reading speed; "Pitch" performs tuning on the voice, and "Pauses" allows you to insert reading pauses.

On each of these parameters, it should be known that they act as a "tag" on the text. Therefore, it may be necessary to specify to the AI which portion of the text we intend to apply the parameter to. For example, to read a portion of text in a formal tone, the program adds the following tags: <mstts:express-as style='newscast-formal'> and </mstts:express-as>

However, it is up to us to insert between one </> tag and another </> the text to which we intend to apply the parameter. So, if we want a certain sentence to have the "formal reading" parameter applied, we will need to insert it between the tags. Here is an example:

<mstts:express-as style='newscast-formal'> this is the portion of text that will be read in a formal tone </mstts:express-as>

In the window below, an image of the chosen voice appears along with the box in which we can insert our text to synthesize. In the current version of the software, it is possible to insert up to 5 voices each comprising up to a maximum of 5000 characters.

Once the text is entered, we can:

- Click on the icon on the right, containing a musical note, to create a temporary synthesis of the written text. Please note that even for just listening, the relative characters will be subtracted from the number of characters available for the account. This temporary speech synthesis is not automatically saved and can be downloaded immediately by clicking on the download icon next to the player. If you request to listen to another text or if you change pages, the temporary file generated is lost.

- Click on the "listen" button at the bottom of the page. This operation is identical to the previous one and generates a temporary file that can be downloaded immediately. If you do not download it while it is on the page, it is lost.

- Click on "synthesize". In this case, all voices and all texts present in the project are synthesized within a single file (in the initially selected format) that is stored in the "AI Voiceover Results" section and is immediately visible on the same page. All voiceover files produced and stored by a user can also be seen by accessing the user archive by clicking on Documents>All Voiceovers on the left navigation bar.