Tutorial

Three Tips for a More Natural Utilization of AI Voice

At VideoStew, we have over 600 AI voices supporting some 130 languages all ready for you.

While most of them perform phenomenally well, there might be times when they don't quite read as you'd like or their pronunciation sounds a bit off.

We've got three sweet tips prepared for our users employing Korean AI voices to make the Text-to-Speech (TTS) process feel a bit more natural.

Make Use of Naver's Clova Engine Voices

VideoStew comes loaded with over 60 kinds of Naver Clova TTS. These voices, based on a neural network, exhibit remarkable quality, and being a product of a Korean tech company makes them exceptionally adept at Korean language.

If you hover your mouse over each voice, you can have a quick listen. You can swap voices at any time while you're working and even set different voices for each slide.

VideoStew Team’s Recommended Voices: Minyoung, Ara(Call Center), and Juan

Slightly Increase the Speed

The current trend for video platforms including YouTube is to deliver information using a fast pace. While the default TTS speed is stable, it might feel a bit slow, so we suggest you kick it up a notch. Differences depend on the voice used, but we recommend a speed increase to about 115%~120%.

We've prepared a comparison video showing the original speed alongside the quickened pace. Would you like to check it out?

<iframe src="https://www.youtube.com/embed/G0rQlAvzbEY" allowfullscreen=""></iframe>

Try Changing the Pronunciation to Match the Sound

While doing narration work, there might be sections where the reading doesn't go as planned, or just sounds unnatural.

In these cases, you can try adjusting with tactics such as eliminating spacing or adding commas. Hell, you can even rewrite the text exactly as you hear it to make the AI voice sound more natural. Want an example?

<iframe src="https://www.youtube.com/embed/RJUD0blp1sM" allowfullscreen=""></iframe>

Disabling auto text recognition allows you to insert messages as you please, independent of what's displayed on screen. In the example above, spacing was removed or content was altogether changed where pronunciation sounded awkward.

We've explored some advanced settings of VideoStew TTS in this post.

With skillful use of VideoStew, you can create the video you want without having to separately record your voice over your edited video!

And if you're finding it difficult to navigate the ropes, don't worry. We offer chat and video consultations to help you get comfortable with the platform.