How to replicate this raw, detached/boxy vocal texture? (AI isolated stem attached — not a basic phone filter)
Yo guys,
I’ve been trying to reverse-engineer this specific vocal sound from a 2015 underground track (artist: вышел покурить, track: Артхаус). I pulled the vocal stem using an AI separator so you can hear the raw capture without the beat masking it.
Link to audio ~~https://drive.google.com/file/d/1QbGwmph9JdnOOfyAWRTSMWUpW-Tb1y4N/view?usp=drivesdk\~\~
What trips me up is that it’s definitely NOT a generic bandpass / telephone EQ. The full frequency spectrum is basically there (there's low-end and high-end air/sibilance), but the texture feels completely detached:
Zero proximity effect / not in-your-face: It doesn't hit dead-center like a modern treated-room condenser. It feels hollow, slightly wide, and floaty — almost like comb filtering from early reflections or speaking 1.5–2 feet off-axis in an untreated bedroom.
Subtle grit / transistor crunch: There's a weird crustiness on the transients, but it doesn't sound like a typical guitar distortion plugin. More like an onboard laptop soundcard / cheap preamp being pushed into soft-clipping on loud syllables.
The "cardboard/bag" resonance: It sounds inherently boxy without just hacking out mids with an EQ.
My setup is a Fifine K669 condenser running into FL Studio. Whenever I try to record, my vocals sound way too crisp, intimate, and clean. Slapping a low-pass or phone preset doesn't get even close to this mood.
What do you think is going on at the tracking level vs. processing? Is this purely room bleed + distance + bad ADC conversion, or is there a specific chain/saturation trick to nail this exact late-2010s lo-fi aesthetic?
Appreciate any knowledge on this!