Nowadays, listening to new rap music on the internet can feel like stumbling into a neonatal intensive care unit. Shrieks and squeals, moans and wails emanate from all angles. Fans have dubbed the high-pitched vocal trend “baby voice rap,” “chipmunk rap,” “fetus vocal rap.” But these terms don’t fully capture the mutant craziness reached on recent tunes. Building off the madcap antics of aural acrobats like Young Thug, Future, and Playboi Carti, a new species of surreal SoundCloud rappers are squeezing and stretching their voices beyond where humans have ever gone before, producing music that could easily be mistaken for experimental ASMR or the national anthems of an alien civilization. Say hello to the final frontier of rap-sing insanity.

“When people listen to my music, I want them to have an out-of-body experience, like they’re in another world,” Atlanta-based rapper 645AR tells Highsnobiety. He was one of the first to get attention for squeaking. In fact, critics and fans weren’t really using the term “squeak rap” until the mosquito-voiced artist dropped his single “4 Da Trap” in 2019. At first, skeptics were everywhere. One critic quipped that the song sounded like the debut track from Voldemort, if he were a fetus. On Twitter, rap fans clowned him for sounding like a demented cartoon figure (it didn’t help that the rapper used images of Mickey Mouse, Powerpuff Girls, and Sesame Street folks as cover art). But 645AR insisted he was serious. Quickly, opinions shifted; there was something mysteriously hypnotic about the music which rewarded repeat listens, entrancing you over time. YouTube comments flooded with remarks from haters turned lovers. Then “Yoga,” a tsunami of ear-pleasing squeals and bass jolts, exploded on TikTok during the early months of the pandemic. Thousands of listeners converted to the 645AR cult.

So why did people change their mind about the music?

The appeal of 645AR and others in this rising wave of ultra-weird rappers like lungskull, Yameii Online, and meat computer is multilayered. Initially, people notice and relish the tunes because they’re basically memes – at once inane and infectious, stupid and genius. The ironic allure is similar to that of “Deep Fried Memes” (mega-distorted, hyper-saturated gifs and images) and kitschy remix fads like nightcore and slowed + reverb. You enjoy the music from an equally mocking and admiring distance, recognizing the helium gurgle as a fun, novel gag, an absurd satire of Playboi Carti’s baby voice.

After you listen enough times, though, the squeak seeps into your skull. The detached amusement turns into genuine joy.

This sort of freaky vocal play is the logical endpoint of what Future does on songs like “Oxy,” using vocal effects to replicate the drugged swoon his lyrics describe. For these new artists, though, many of the words are enunciated so weirdly they’re practically indecipherable, language reduced to brain-tickling hiccups. It’s not that the songs have no meaning. Rather, the otherworldly curvature of the voice is the meaning, textures and tones replacing lyrics as the primary vehicle of expression. For instance, 645AR’s pinched falsetto on “4 Da Trap” was conceived as a way to convey the abject misery he experienced growing up, making you feel the melancholy via the tremulous flutter of the voice.

On the most mind bending tunes, there’s no concrete emotion portrayed at all but an amorphous array of squeaks and gasps. The music tingles your scalp; fans often use ASMR-related adjectives like “head-massaging” to describe it. There’s something about the collision of sickly-sweet vocals and dark, dissonant beats that gives you goosebumps. The bass becomes a supercharged trampoline, sending the words flying into the air with a delicious whirl.

Rap has an extensive history of playful messing with vocal effects and timbres, and there’s been a long line of vocally eccentric MCs: T-Pain, Young Thug, Lil Wayne, etc. But the difference is that today’s artists are going so far into off-kilter inflections and astonishing vocal effects that practically none of their original timbre or phrasing comes through. These artists are creating aural avatars; virtually inscrutable vocal costumes they can slip into at whim, and which have essentially no relation to how they sound in real life. There’s a massive gulf between the performer-voice and the real-voice – if you’d only heard 645AR’s music, you wouldn’t recognize his everyday speech. Indeed, his cybernetic squeak doesn’t sound like any human being. It’s the ultimate extension of modifying your image until it’s completely unreal. On TikTok, where most of the tunes explode, the voice is even more spectral, because there’s usually no song title or artist name attached to the videos. Fingerprints erased, the vocals turn into pure alien sound rippling out of the digital ether.

Maybe the freakiest rising vocalist is whisper-warbler meat computer. Whispering in rap isn’t a completely new thing, Ying Yang Twins even put it in the title of their lewd 2005 smash “Wait (The Whisper Song).” Rather than rap, meat computer’s sound works more like unintelligible talking ASMR videos, excising meaning in order to relax listeners with the simple soothing pleasure of hearing a soft-spoken voice. “Eyes wide shut,” a bubbling stream of downcast half-sentences, lulls you into a loopy daze. “No mods” is perhaps the first ever fully ad-lib-based tune, a sort of ultimate scat rap. “I’ll quickly mumble freestyle and then the next day, write to the mumbling,” meat computer told us about their production process. “I think it’s a cool idea to make songs that are [on the surface] incoherent, but if you look deeper and read the lyrics, they’re actually saying a lot more than these other people whose lyrics you can actually understand.”

Where meat computer produces whispers authentically, SoundCloud rapper lungskull uses technology: specifically, he speeds his vocals up. All you hear in the 15-second clip of his breakout cut “foreign” circulating on TikTok is an eerie, aggressively high-pitched mewl that swirls into the frame like a haze of fumes rising from an incinerated building. The voice could be coming from anywhere, anyone, anything. The whole song is distorted, scuffed, scratchy – a slick production trick that makes the radioactive voice stand out even more, turning it neon gray.

Lungskull’s tunes are mostly surging on Roblox TikTok, a subgenre of videos where teenagers post montages from the popular sandbox game. Everything about Roblox seems antithetical to “foreign” – its graphics are clunky and dated; its player base is skewed toward kids. Yet for these players, there’s something incredibly alluring about the music, as if the ultra-intensity of it substitutes for the lack of visual vitality in the game, which was made back in 2006. Coupled together, “foreign” and Roblox gameplay attain this cloddy clarity, a harmony of opposites that’s strangely pleasurable, like watching a hydraulic press crush a stuffed animal.

Likewise, Axxturel’s “All Eyez on Me,” an eruption of molten gurgles and convulsing glitches, didn’t cross over on TikTok as the soundtrack to people moshing in their bedrooms, or anything remotely haywire. It’s draped over videos of people standing in front of the camera, sometimes twirling to show off an outfit. The mismatch is striking, but it makes sense: the music captures what it feels like to be locked away in your room for months during a pandemic. “All Eyez on Me” sounds as if it’s perpetually on the brink of implosion, like the vocalist is decomposing in double-time. The vibe of their music, Axxturel says, is “bittersweet chaotic harmony.”

Axxturel was the CEO of the now-defunct Jewelxxet, a low-key rap faction full of innovative vocalists, most of whom don’t squeal but hiss. There’s islurwhenitalk and sellasouls, inventors of what could turn out to be the logical endpoint of bizarro vocals: bona fide breath rap “Xxelene.” “I was just expressing myself… it came from the other side… from my pact with a deity,” says sellasouls about the vocals. “I wanna die at a young age, the mortal body is annoying.”

While 645AR, meat computer, and lungskull only sound like aliens, sellasouls and company claim to be literally ascending, communing with and soliciting advice from spirits. Axxturel’s most recent tune to storm TikTok, “Ave Domina Lilith,” a dystopian slush of chopped refrains and slithering synths, has some religious listeners spooked, believing it’s damned. The videos are chockablock with comments urging people to pray after listening. “Guys I listened to this on mute am I gonna get cursed?” one asks. The fright seems a bit overblown or misguided (is it really that easy to summon Satan in 2021, just rapping with a ghoulish tone?). In addition to these astonishing vocal styles, one of the group’s best innovations is how they’ve mastered mid-track pitch-shifting: songs like islurwhenitalk’s “Champagne” skitter between speeds, scrambling the voice into a gale of syllable shards. The effect is both terrifying and oddly satisfying; it seems to simultaneously itch and scratch every inch of your spine at once.

If you thought embryo- and wraith-voiced humans was strange enough, wait ‘til you hear cyborg squeaker Yameii Online, a Vocaloid designed by rapper Deko and visual artist Oseanworld. There’s no flesh or blood to her bionic falsetto at all – only high-pitched synthetic tones, extracted from a database of stock vocals. You’d expect the timbre to be flat and fixed; how expressive can a robot be, right? But on highlight TikTok crossover “Baby My Phone,” the virtual artist flexes a wondrous palette of flows and feelings, from jumpy robo-libs (“eep, eep, eep”) and tremulous mecha-murmurs to pixelated gibberish in only a few minutes. It’s a cradle song for sentient Siris.

Although Vocaloids have existed for ages, the tech has mostly been used to create avatars with humanlike tones (see Hatsune Miku). Deko and Oseanworld are doing the opposite to Yameii, plunging deeper into artifice, willfully estranging her from our animal race. They’re carving out an entire android hip-hop style with a wacky language and gleaming sound of its own. It’s the next step in terms of virtualization, where it’s not just a voice super-enhanced to the point of disguise but there’s no original in the first place—like a mask with no face behind it.

Where do we go from here? How can rappers (flesh and machine) take this sound even further?

Rather than being the endpoint, this craze for unabashedly bizarre vocals looks likely to be the start of an entire new era of experimentation for hip-hop. There’s a panoply of ideas bubbling in the underground, from meat computer’s babbling ASMR’n’B to Jewelxxet’s sorcerous hiss-rap. Breath rap also seems promising. Who knows – someone might invent an even more exquisitely nonsensical style like spit-rap, groan-rap, wheeze-rap, sea shanty-rap, lullaby-rap (for small children and migraine victims) or a form that combines everything into one: a triple-decker mousse cake crammed with Martian squeals, harmonic whispers, and sinister snarls. Similarly, rappers might discover weirder and wilder ways to distort and reshape their vocals with effects. The possibilities are endless.

Given many of these artists only started producing music after the pandemic began and have never performed in real life, their aural personas are literally trapped inside the digital ether, contained entirely within the pristine sheen of the screen. It’ll be interesting to see how this sound translates to physical spaces—or, maybe, it’ll stay virtual forever.