Mixing for social media and smartphones

The context

A huge share of listening happens on smartphones (built-in speakers, cheap earbuds, Bluetooth) and on social platforms (TikTok, Instagram Reels, YouTube Shorts). These systems have precise limits: a mix that ignores them loses low end, vocal, and impact right where the most people discover it.

Limits of small systems

• Low end: phone speakers reproduce little or nothing below about 200–400 Hz (depending on the model); an 808 at 50 Hz simply doesn't exist.
• Mono or nearly mono: a single speaker or two very close together; wide stereo collapses.
• Emphasized mids (1–4 kHz): anything harsh in that zone becomes annoying.
• Noisy environments: listeners are often on the street or the subway: the vocal must be very clear.
• Platform and device processing: lossy encoding, loudness normalization (often adaptive and undocumented on social platforms), limiters and phone “enhancers”.

Mixing strategies

1. Audible low end through harmonics: saturation on the 808 and bass to create the 2nd, 3rd, and 4th harmonics (A1 = 55 Hz → 110, 165, 220 Hz): the brain reconstructs the fundamental (missing fundamental). Tools: saturation, parallel distortion, Waves MaxxBass, a “mid” layer for the 808.
2. Kick with click and punch at 2–5 kHz and 100–200 Hz, not just sub.
3. Intelligible vocal: presence at 2–4 kHz well controlled, careful de-essing (codecs and small speakers make sibilance worse).
4. Strict mono compatibility (see the dedicated lesson).
5. Clean mids: no buildup at 1–3 kHz among synths, guitars, and vocal.
6. Immediate hook: in short-form content, the strongest moment of the track must work on its own (often the 15–30 second fragment that goes viral); make sure that passage is mixed at its best.

Loudness and social media

• Social platforms don't publish an official loudness target and some apply adaptive normalization; as a cautious starting point, use a master around −14/−10 LUFS integrated with true peak −1 dBTP (−2 to be safe with aggressive codecs). Higher values risk distortion in transcoding.
• Keep transients alive: overly pushed limiters stacked with the phone's limiter produce a flat sound.
• Export in high quality (24-bit WAV or high-bitrate AAC when the platform allows it) and check the uploaded result.

Checking

• Listen on a real phone (several models if possible), with cheap earbuds, and with a Bluetooth speaker.
• Simulators and “device preview” plugins: for example Audified MixChecker, Waves Nx/Abbey Road for headphones, iZotope Ozone (codec preview), or simply an EQ with HP at 300 Hz and LP at 10 kHz + mono for a rough simulation.
• Upload privately to the platform and listen from your phone: it's the only reliable test of the real chain.

Example in A minor

Pop-trap track at 140 BPM designed for TikTok: 808 on A1 with a saturated layer filtered HP 120 Hz and LP 2 kHz at −8 dB, creating harmonics at 110–220–330 Hz; lead vocal with presence at 3 kHz and a de-esser at 6.5 kHz; chorus (the 20-second segment) with vocal +1 dB and a mono synth hook. Test: phone simulated with HP 300 Hz in FL Studio (Parametric EQ 2 + Fruity Stereo Shaper in mono): the 808 reads as a “line” thanks to the harmonics.

Common mistakes

• Relying only on the sub: on the phone the track “has no bass”.
• Extremely wide stereo: in mono half the mix disappears.
• Master pushed to −6 LUFS: it distorts in encoding and gains nothing with normalization.

Exercise

1. Build a “phone” simulation chain on the master (to be bypassed on export) and mix for 10 minutes with only that active, then remove it.
2. Compare your mix and a successful reference on the phone at equal volume: note what you hear in the reference and not in yours.

Links

“Psychoacoustics in the Mix” (missing fundamental), “Mono Compatibility”, “Mastering for Streaming”, “Designing 808s and Basses”, “Distortion as Sound Design”.