Table of contents

The voices used in Seconds are a part of the iOS SDK provided by Apple.

Recently, in iOS 16 Apple has added new voices. These voices appear to be specifically for blind users.

Currently, Seconds chooses a voice for you by selecting the first voice for your device language. Unfortunately, these new voices have jumped to the top of the list of voices and are being selected instead.

You can still select the nice-sounding voices in Settings > Sound Settings > Language.

Choose either Samantha, Nicky or Aaron.

Choose either Daniel, Martha or Arthur.

Choose either Gordon, Karen or Catherine.

In the Settings app, go to Accessibility > Spoken Content > Voices > German and find the voice you like. Make a note of the name. Then in Seconds, go to Settings > Sound Settings > Language and try and find the voice that matches.

There will be an update soon that sorts out the old voices and the new ones so that this process is easier.

On top of the issues listed above. Apple appears to have added more issues in the latest updates to the voice library. Specifically, there are known memory leaks.

Memory leaks are when the code will use some device memory for an operation, but then fail to release that memory when it is complete. The result of this is that the device runs out of memory and the system will close the app to reclaim the memory.

This will be more noticeable when there are lots of intervals and more separate instances of the speech.

This is something that should be resolved by Apple in an iOS update. In our experience, the voice libraries Apple provide have been plagued with issues for years and only seem to get worse over time. We are investigating ways of moving away from Apple's voice libraries, but this will take time.

You can read about more developers facing the same issue in this thread on the developer forums.