Designing the balinese script-to-speech synthesis system using noto serif balinese font

Abstract

The existence of regional languages shows an identity of the existence of a tribe. Indonesia has many regional languages, one of which is Balinese with Balinese script as input for the script which comes from Sanskrit. Reading Balinese script requires knowledge of Balinese script and Balinese culture in order to be able to correctly pronounce the written Balinese script. Since learning Balinese script took a long learning process, it certainly causes a slowdown in knowledge about how to read the Balinese script. Therefore, it is necessary to have a system that can translate Balinese script into a speech so that it can be used in learning the Balinese script without any mistakes in pronouncing vowels. This research designed a system synthesis of Balinese script into a speech with the system synthesis method of Noto Serif Balinese Font which consists of three processes namely text pre-processing, prosody generation, and concatenation. Concatenation or combining speech will use the Pitch Synchronous Overlap Add Method (PSOLA). The result to be achieved in this research is to be able to decompose the Unicode from the Noto Serif Balinese font into a sentence which then this sentence will become a message. In the future, this research can use input from an image to directly enter into speech. Making it easier for users to translate the Balinese script.