Introduction to AI Vocal Technology and Recommendations

AI vocal currently has several usage requirements, which are:
AI-generated music, AI text-to-speech, AI cover songs, and AI voice changers

Here is a brief introduction and recommendation of related software for the four categories.

1. AI-generated music

sunoAI

If you don't understand music but want an original piece of music or background music (BGM), you can use this tool. It is currently the only recommended tool for generating music from text descriptions. It can generate almost any music based on your accurate description, and you can also customize the lyrics. Although it was not developed by a domestic team, it has a good recognition of traditional Chinese styles or other less common genres. Its usage is very simple; you just need to accurately describe the song you want. It is recommended to use it alongside ChatGPT to generate the prompts you need.

Chord School

Note that it requires a certain level of music expertise, and ordinary people may find it difficult to grasp.

If you have a good understanding of music theory and want an easy way to create accompaniments and chords, you can choose this one. It offers a variety of presets to help you complete your chords, and you can also generate music using AI and then fine-tune it.

II. AI Text-to-Speech

CapCut

The first recommendation is the text-to-speech feature in CapCut. If you just need a soulless text reader, this would be a good choice. CapCut is very easy to use and also offers a rich voice library. Most of the low-quality short videos on the market use CapCut for their audio (mainly because it's convenient).

ChatTTS

The effect of this is extremely realistic. Of course, you'll need to find some tutorials online to help with local deployment and do some learning. However, it's not as complicated as you might think. If you need a high-quality AI voice that's almost indistinguishable from a real one, this is the one to choose. You can either use existing models or choose to train your own voice.

If your only need is text-to-speech, then I wouldn't recommend others. You can check out the following video for other recommendations, as it explains things more clearly.

Ⅲ. AI Cover Singing

VITS

If you want your anime waifu to sing for you, you can choose this tool. Of course, it also supports text-to-speech, but compared to the two tools recommended above, the learning curve is a bit steeper. You'll need to spend some time learning how to configure and use it. But don't worry, there are plenty of tutorials online, just search for them, and you'll find many.

RVC

RVC can also help you create AI covers. I haven't tested the results, so I'm not sure whether it's better or more user-friendly than VITS. For detailed tutorials, you can check this one.

DiffSinger

This tool seems to support functions like lyric modification, pitch adjustment, dry voice extraction, and score conversion. Please test it yourself for more details.

Ⅳ. AI Voice Changer

RVC

The only recommendation for AI voice changing at the moment is RVC. In addition to singing covers, it can also help you change your voice. If you can't find a satisfactory model, you can also train your own. Personally, I highly recommend the integrated package created by Bilibili user "唯有入梦," which allows for one-click usage, making it very beginner-friendly.