|
|
Use a high-quality microphone
|
|
A podcast microphone is the best way to capture your voice sample. If you don’t have access to one, you can use the Voice Memo app on your iPhone. Be sure to hold the phone at least 8 inches away from your mouth and speak clearly.
|
|
|
|
Speak as you want the Delphi to speak
|
|
We provide a script within the Voice portal to read. If you prefer a more conversational style, it’s encouraged to record yourself giving a mock presentation or having a conversation with yourself. Test out different delivery formats to see what works best.
|
|
|
|
No background noise
|
|
Ensure that your environment is as silent as possible. Background noise will impact the quality of the sample.
|
|
|
|
90 second sample minimum
|
|
It’s recommended to have at least 90 seconds of continuous audio. The longer, the better. Feel free to record yourself for 2-4 minutes if you want a robust sample.
|
|
|
|
One sample is best
|
|
It’s not recommended to upload multiple voice samples, especially if they are recorded on different devices and in different environments. One high-quality audio sample will produce the best replication.
|
|
|
|
Harness our settings
|
|
Utilize the Stability (fluctuation in pitch), Clarity (annunciation of vowels and consonants), and Style Exaggeration (cadence) settings to finely tune your Delphi voice. Voice Description allows you to give the Delphi voice further personalization.
|
|
