Interested in converting text to speech directly on your iPhone? Even better, would you like the text-to-speech feature to mimic your own voice? With iOS 17, you can do just that thanks to a new feature called Personal Voice.
This feature allows you to create a replica, well, sort of, of your voice, which can then be used to read aloud the text you type on your iPhone.
What is Personal Voice?
Before we delve into the setup process, let’s understand what Personal Voice is and what it does.
Personal Voice is a new feature introduced in iOS 17. It leverages the device’s built-in machine learning capabilities to study and analyze your voice, subsequently creating a personalized voice model on your iPhone.
Once this model is generated, you can use it for text-to-speech functions, such as reading out text, phrases, or sentences on your iPhone.
Creating Your Personal Voice on iPhone
To generate your unique personal voice, you’ll need to record 150 sets of random sentences, which should take approximately 15 minutes. Your iPhone will then analyze these recordings and construct a voice model based on your audio input.
Here are the steps to set it up.
- On your iPhone, go to ‘Settings’ > ‘Accessibility’.
- Scroll down and tap on ‘Personal Voice‘ and select ‘Create a Personal Voice‘.
- Tap ‘Continue’, provide a name for your personal voice, then tap ‘Continue’ again.
- You are now in the sound quality check phase. Tap the ‘Record’ button and read out the text. If everything is all good, tap ‘Continue’ again to proceed.
- Now, this is where the real recording of your voice starts. Tap the ‘Record’ button and start reading the text you are prompted. Once you’ve finished reading, you will see ‘Moving to the next phrase‘.
- If you wish to re-record a particular sentence, tap ‘Re-record‘. If you wish to know how a particular line sounds like, tap ‘Preview’.
- Once you have finished reading all 150 out of 150 phrases, tap ‘Continue’, followed by ‘OK’.
You are all set for now. To expedite the generation of the voice model, lock your phone and put it on charge. The entire process may take a few hours, sometimes even days.
Note: In my case, it took around 8 hours with the phone locked and connected to a charging cable.
How to Know if My Personal Voice Is Ready?
When your personal voice is being prepared, you will occasionally receive a notification, especially when your phone is locked and charging.
Once it’s done, you will receive a notification stating that it’s complete and your personal voice is ready for use.
Alternatively, you can check the status by navigating to ‘Settings’ > ‘Accessibility’ > ‘Personal Voice‘. If it’s complete, it will display the date it was created. Otherwise, it will show you the progress as a percentage.
How to Use Personal Voice on iPhone
To hear what your personal voice sounds like during narration, you will first need to turn on ‘Live Speech‘. Here’s how to do it:
- Go to ‘Settings’ > ‘Accessibility’.
- Scroll down and tap ‘Live Speech‘, then tap again to turn on ‘Live Speech‘.
- Locate the ‘VOICES’ label and select the custom voice you previously created.
- To use your voice for narration, quickly press the power button three times to launch Live Speech.
- Enter the phrase or sentence you wish to be spoken and hit the ‘Send’ button.
How long does it take to create a personal voice on an iPhone?
Well, it takes about 15 minutes to record all 150 phrases. Then, it takes a few more hours to days, depending on how fast your iPhone can analyze and generate the voice model. Your best bet to expedite the process is to ensure you lock the phone’s screen and put it on charge.
Can I delete the created personal voice?
Yes, to delete any created personal voice, just click into it, and then tap on the ‘delete’ button.
Can I access my personal voice on other devices?
Yes, as long as the other devices are signed in with the same iCloud ID. Also, in order to share, you must also enable it under ‘Accessibility’ > ‘Personal Voice’ > ‘Share Across Devices’.
Will Apple be able to access my personal voice?
Personal Voice utilizes on-device machine learning, supported by the Neural Engine, to analyze and generate your voice model. So, the answer is no. Apple will not be able to access your personal voice or any of its associated data.