AI Vocals for Hardstyle

I use ttsmp3.com. It's free and has a bunch of different options. You can also search for AI voice generator or text to speech and you'll find a bunch of these websites. First you need lyrics, so write something really ominous.

Put in just two words so you can test the voices without having to wait. Look for something a little bit monotone that doesn't sound too characteristic. I want a female voice, because I don't really hear male voices in the scene. I went with Kimberly. Download it as MP3 and load it into your DAW.

Before any processing, put the vocal on different stretch modes and change the time multiplication. Some work better than others. On e3 mono it gets this weird robot stretch, which works really well.

Then a neat trick. Add Auto-Tune with the retune speed right down. It really stretches out the notes and makes it sound more artificial. Render that and load it into NewTone (you can use Melodyne as well). Change everything to go to the center, change the variation, and turn on transition. That's where you get this really robot kind of voice.

I always start with EQ. Get rid of the sub, we really don't need that, and take the highs down a little bit. The sh in the vocal is really prominent, so control it. FabFilter Pro-DS works but it's a paid plugin. A Patcher preset like Intelligent Deesser does the job too. Then Pro-Q 3 as a dynamic EQ to compress the highs a little bit. We're just clearing up the vocal and making it as flat as possible.

Then it's time to excite the vocal. A little bit of saturation with GSatPlus, not too much. Then OTT, that's a big difference already. Then a compressor. I like the CLA-2A with a vox preset.

Step three is special effects. The first one is dBlue Crusher. It's a free plugin, you just add some bitcrush and play around with the reduction. That's a huge improvement already.

Then I add a Patcher and create different variations of the vocal. One totally dry channel where nothing really happens. One with Little AlterBoy for a nice top layer, and another Little AlterBoy with a different intonation to build it up. Then VocalSynth for a really weird layer. That adds a lot of bass and other stuff we don't need, so excite it one more time with OTT and compress it again.