How the Scary Voice Goanimate Creator Reveals Dark Secrets Behind The Horrifying Phenomena Actually Works
I came across this channel probably two years ago when someone in a design forum linked it as an example of what not to do with animation software. The videos are straightforward — some guy makes these weird horror skits in GoAnimate (now Vyond) using distorted voice changers, then drops "secret" theories in the description like he's exposing something. The whole thing is just a content farm gimmick at this point, but there's a practical side to it that people miss. GoAnimate is a basic drag-and-drop animation tool. You pick a character, type dialogue, hit play, and you have a video. The scary voice trend started around 2018 when people discovered the built-in text-to-speech voices could be pitch-shifted down or run through cheap audio effect plugins to sound unsettling. The "dark secrets" angle is just clickbait formatting — thumbnail shows a distorted face, title promises revelations, video delivers a 90-second cartoon with a narrator saying something like "they don't want you to know that..." and then shows nothing of substance. I spent a weekend reverse-engineering how a few of these were made because I was trying to replicate the audio effect for a legitimate project. The pipeline is embarrassingly simple.
The actual workflow behind these videos
Character selection: most of these creators use the default male characters, often the one called "Rob" or just a plain business guy. They slap on a black hoodie asset if they want it to look slightly creepier. Background is always the same — a dark room, sometimes a hallway, occasionally the green screen option with a stock spooky image pasted in. Audio is where the work actually happens. The text-to-speech voices in GoAnimate are based on older iSpeech and Microsoft Sam-style engines. To get the scary voice, creators export the audio and run it through Audacity with three effects in sequence: pitch shift down by about 300-500 cents, add a touch of reverb (room size around 20%), and sometimes a subtle distortion pass. Some use free plugins like Freeverb or just the built-in effects. The result sounds like a robot talking from inside a tunnel. That's it. Script structure is formulaic: a hook in the first five seconds ("I used to work at GoAnimate and here's what they hide"), a couple of "evidence" screens showing blurry screenshots or cropped images with red circles, and a call to subscribe before the video ends at roughly two minutes. Average production time per video is about twenty to forty minutes once you know the steps.
Common pitfalls I ran into
The biggest issue is timing. GoAnimate's text-to-speech doesn't handle punctuation well for the scary voice effect. A colon or em dash reads as a full stop to the engine, which means your pacing gets wrecked. I ended up removing all punctuation from my script and using comma pauses manually by splitting lines into separate dialogue boxes with tiny delays between them. That gave me control over the rhythm without fighting the engine. Another problem: the exported audio from GoAnimate is low bitrate, usually 64kbps MP3 or worse. Running that through pitch shift amplifies the compression artifacts — you get this chunky, muddy quality that actually hurts the creepy effect more than helps it. My workaround was to record at the highest quality setting GoAnimate offers, then use a noise reduction pass in Audacity before applying the pitch shift. Cut the sample rate to match your target platform too. YouTube compresses heavily, so starting at 44.1kHz instead of 48kHz saves you a generation of degradation.
Get the Full Details
When this approach completely fails
If you're trying to make something genuinely atmospheric or narratively compelling, the GoAnimate + scary voice combo hits a wall pretty fast. The aesthetic is so tied to early 2020s YouTube junk content that any serious short film or educational horror video using it will look like a parody regardless of what you put in the script. The software itself has hard limitations — no custom lip sync, no real acting, no environmental interaction beyond sliding sprites around. You're working with puppets that can nod and blink. For anything beyond a novelty sketch, I'd recommend looking at real-time motion tools like Adobe Character Animator or even Unity with a simple rig. The learning curve is steeper but you get actual performance capture instead of typing text into a box and hoping. If you still want to make these, the GoAnimate account is basically dead — the platform pivoted to Vyond which is enterprise-priced and stripped most of the casual features. You can find archived versions floating on forums, but they're unofficial and carry malware risk. A safer route is to use OpenToonz or Blender's grease pencil for basic animation and process the audio separately. It takes longer but the output is yours and won't break when the service shuts down.
The channel I mentioned doesn't post consistently anymore. The algorithm moved on. But the technique is still useful if you're making weird internet horror content and don't have a budget for real animation. Just don't expect the "dark secrets" to amount to anything beyond watch time.
