For users who have lost the ability to speak or are at risk of aphasia, iOS 17 has added two text-to-speech functions that can be spoken by the synthesized voice of the mobile phone. They are “Input to read aloud” and “Personal Voice”.
type to read
“Input to read aloud” allows aphasic people to call out the keyboard immediately, and the input text is read aloud by the mobile phone with a designated voice. At the same time, common words and sentences can also be written in advance, which is convenient for one-click reading.
After turning on the “Type to read aloud” switch, just press the side button on the right side of the phone three times to call out the reading aloud keyboard. Here you can also set commonly used words and sentences and select synthetic voices used in different languages.
There are three options for Cantonese voice, but I think Siri’s voice is the best. In addition to Cantonese, Chinese can also choose Mainland Mandarin, Taiwanese Mandarin, and Sichuan, Shaanxi and Liaoning accents. In English, not only the accents of different regions can be selected, but also the voices of ghosts and aliens can be selected. The capacity of each Chinese is about 500MB, while the size of English is only 300MB even if it is of high quality. Most of the voices can be auditioned before deciding to download. For each voice, you can click “i” to adjust the pitch and reading speed. If you don’t like it after downloading, you can swipe left to delete it.
personal voice
As for individual voices, it is for those who have recently been diagnosed with a risk of losing their speech. It uses the machine learning function in the device to learn the user’s voice, and in the future, it can use the user’s voice to speak instead when answering the phone, FaceTime or using other auxiliary communication software.
This feature currently only supports iPhone 14 or above, and only supports English for the time being. To create your own voice on your mobile phone, you need to follow the instructions on the screen to read out about 15 minutes of random content for the machine to learn. Users don’t have to read it all at once, they can read a few sentences when they have time, but it is best to read it in a quiet environment.
