Generating sounds using a speech synthesis IC
Generating sounds using a speech synthesis IC
To generate sound using a speech synthesis IC, the following components are required.
MCU
Speaker
Speech Synthesis IC
These components are connected as illustrated in the diagram below.
The MCU and speech synthesis IC communicate via serial interfaces such as SPI or I2C. At the same time, the speech synthesis IC features dedicated terminals for speaker connection. Once audio data is stored in the speech synthesis IC, sound can be played by simply sending a command from the MCU.
Below is a sample flowchart for sound playback. Sounds can be generated by sending the following commands to the speech synthesis IC.
Useful Features of Speech Synthesis ICs (Editable ROM Function)
Since ROM capacity is limited, the editable ROM function optimizes storage by maximizing the number of audio data entries within the available space. This feature allows multiple phrases to be linked together to form a single sound, enabling seamless continuous playback. What’s more, silent playback can be inserted between segments for smoother, more natural audio transitions.
For example, instead of storing full phrases like ‘Today's weather is sunny’ and ‘Tomorrow’s weather is rainy,’ separately in ROM, the sentences can be broken into segments, such as ‘Today's,’ ‘Tomorrow's,’ ‘weather is,’ ‘sunny,’ ‘rainy,’ ‘cloudy.’ These segments can then be dynamically combined during playback, significantly reducing the amount of ROM required.
Segmenting the audio allows phrases to be assembled, such as combining ‘Today's,’ + ‘weather is’ + ‘sunny,’ to generate the ‘Today's weather is sunny,’ effectively minimizing ROM usage.
A comparison of the yellow areas in the tables for cases with and without the editable ROM function clearly shows a significant reduction in space when the editable ROM function is used. This demonstrates its effectiveness in minimizing ROM usage.
