Compass is still being developed and is effectively in beta. Some content and features may be incomplete, change, or not work as expected.
Content is anchored to trusted sources where available, but Compass has not yet been clinically reviewed. It provides general information rather than medical advice.
Feedback helps us identify gaps, errors, usability problems and opportunities to improve Compass over time.
Voice Banking Guide
Part of University of Iowa Speech, Language, and Hearing Clinic
USIndependent publisher
This describes who publishes a source, not how reliable it is.
Open source (opens in a new tab)Link checked August 2026
About this source
A university speech clinic's practical guide to message banking and voice banking, covering recording conditions, how long recording takes, and the services available including options for people whose speech has already changed.
Where this source is used
Used in 3 places across 1 domain.
Questions
Communication
Voice Banking
- How does voice banking work, and what is involved?
Supports this
“Recording does not have to be done in one sitting, and taking breaks is normal advice rather than a compromise.”
Its walkthrough for one of the longer services puts the recording at six to seven hours in total, and tells people to take breaks rather than try to finish it in a single day.
Note: Recording will take around 6-7 hours so make sure you take breaks and do not aim to record in one day.
ModelTalker, Recording the main sentences - How does voice banking work, and what is involved?
Supports this
“Background noise and the microphone you use affect the quality of the recordings, and so the quality of the voice built from them.”
A university speech clinic guide covers both halves of the setup. It lists practical ways to cut out noise, such as closing windows and turning off fans, and it says the quality of the audio affects the quality of the voice, which is why it tells people to use an external microphone rather than the one built into a computer.
Make sure you avoid background noise that can decrease the quality of the recording.
Voice Banking, Before you Begin - What if my voice has already changed?
Qualifies this
“Recordings made before diagnosis can sometimes be used to build a voice, particularly with AI cloning, but only if there is enough good quality material.”
Confirms that services will work from old recordings but treats it as something to be assessed individually, and only where the existing material is good enough.
You can create a voice based on old recordings on a case-by-case basis if there are sufficient high-quality recordings.
CereVoice Me, Features
This is an external source. Compass links to it and describes it but does not hold rights over it. Opening it takes you to the publisher’s own site.