fbpx

Voice Banking with TheVoiceKeeper in three simple steps

Voice Banking with TheVoiceKeeper in three simple steps

TheVoiceKeeper was built with simplicity in mind. We wanted to create the fastest and easiest voice banking service.

The process of using our product is a testimony to this goal. In this post we’ll walk you through the three simple steps to create your personalized voice.

Registration process? 5 seconds.

Complete the registration process in 5 second using Google or Facebook login. We’ll ask you three optional questions, and you can start recording.

Too slow? We can even do it faster! If you don’t want your data to be saved, or agree to any terms no problem!

Take our technology for a test drive by starting anonymously. Create a voice and test it as much as you want and, after you close the tab, we’ll delete all your information!

Recording time? Not long!

You can record from any Windows or Mac desktop computer, or using our iPhone application (coming soon).

Note that recording from an iPhone is recommended only as a test drive of the system. If you are working on a voice that you’d like to use, we recommend investing $40 into a good microphone, and doing your recordings on the desktop.

How many sentences? As few as 30 for the start!

To get the first sample of your personal voice, you only need to record 30 sentences. The process typically takes around 3 minutes. After that, our server will get to work to generate your synthetic voice. This should take another 8-10 minutes.



After recording 30 sentences, you can either wait for the system to create your voice or—better still!—use the time to complete your first official milestone—recording 100 sentences.

The quality of the first, 30-sentence, version of your voice can surprise you. It can be very good, or you may hardly recognize yourself! We’ve seen both results multiple times, it’s perfectly normal.

30 sentences is not really enough material to create a top-quality synthetic voice. In fact, the greatest leap in quality comes with completing the next two milestones, that is recording 100 or 200 sentences more.

Recording 200-300 sentences usually allows you to reach 80% of the possible quality of your personalized voice. This of course varies from speaker to speaker.

With each milestone we will create a new, improved version of your voice. So, in theory you can keep recording until you don’t hear an improvement. Use our “Is that enough?” wizard (coming soon!) to help you decide when to stop.

Rule of thumb? When in doubt about how many sentences to record, aim at 100 sentences. This will allow you to gain a better picture of what our technology can achieve.

Recording instructions? Not many!

One of the main advantages of TheVoiceKeeper is that you can record anytime. We’ll save al the sentences you’ve recorded, and next time you log in you can resume the process where you left off.

Recording with TheVoiceKeeper is simple—you don’t have to try hard. For most natural results: Record as if you wanted to get it over with it!

We even advise people not to replay the sentences they recorded. Of course, if you slipped your tongue, it’s better to re-record the sentence. Re-recording and “perfecting” your sentences takes time, and results in a less natural voice.

Remember, a small mistake is OK—our technology is very forgiving!


Creating a voice with TheVoiceKeeper takes 10 minutes and costs you nothing.

Ready to try?


Cover photo by Z S on Unsplash