Don't sleep on AI voiceovers

Hello Colorful Characters.

Thank you all who got my latest guide focused on the Concept stage for 2D artists.

I’m glad to see the information I share is helping you embrace AI in your creative work.

The 50% discount code IMAGINATION is still in effect, so if you’re thinking of learning new techniques for your 2D workflow, you can get the guide here.

Alternatively, I’ve added a referral link at the end of this mail, where you can get the guide as a reward for sharing this newsletter with your friends.

NOW, let’s get to todays topic - AI voiceovers.

We’ve all been in awe of how good AI has become at creating images. So much in fact, that most haven’t been paying attention to Speech Synthesis, or in other words - Voice AI.

Why would you use voice AI in your work, you might ask?

Visuals immerse us in the world, while the music and sounds create the feelings and atmosphere of that world.

It adds another layer to our visual storytelling. Additional layers usually require additional work, which is true for this case as well. BUT Voice AI is SO freaking good, that you don’t need to do much to add this additional layer.

SO, I want to share the workflow I’ve been using to create this recent animation.

With this animation, I started with visuals and moved into the narration afterwards.

To start, I asked ChatGPT to give me a short script for the animation I created.

After a few variations and changing the style ChatGPT wrote in, I landed on this

It was a bit long for the duration of my animation, so I chose the lines that would fit well with it.

I copied them into LOVO to generate my voiceover.

I rendered the voice, put it in my video and that is literally it.

It’s a super quick workflow and even with a simple image sequence it will add another layer to your story.

Platforms to choose from.

I’ve tried a few different platforms for Speech Synthesis. Some free, some paid.

I’ve found myself using LOVO more and more, as it fits my needs very nicely.

The selection of voices is good, most are realistic, but some can extend into more characters. It gives control over the emphasis, speed, pauses and pronunciation of the text.

And best of all, it has ChatGPT inside the interface, which allows me to do the whole process in one place.

Try it here (this is an affiliate link).

ElevenLabs has incredibly realistic voices as well. The word emphasis and speech is spoken with emotions, as if a person was talking.

Try it here.

Uberduck is an open source platform, where everyone can upload their own models. This results in a HUGE library of realistic and character voices. The speech synthesis might be distorted at times though.

Try it here.

Visual storytelling is more than just images. Adding more layers to it will only enrich the experience of the viewer.

If you’re enjoying these mails, why not share them with others?

I would love to hear from you as well!

If you’ve got feedback, comments, suggestions, reply to this mail and let me know.

I’m also planning to highlighting other artist’s workflows, so if you have one you would like to share, let me know!

You can stop by my website or social media and as always,

Keep creating and don’t forget to have fun. ☀