The mix sounds huge in the studio and thin on your phone. The bass line you worked hard on has vanished, though the meter says it is the loudest thing in the song. Or the reverse: on the monitors the low end is a wobble that eats the kick, and the guitars, which sounded so wide, shrink to a small buzz when you press mono.
These are the standard problems of the bottom and the middle of a rock, pop or band record, and each has a standard fix. This piece is for the desk: three tools, then a checklist, ranges, common mistakes, and a session to practise on. It takes EQ, compression and panning from Mixing & Signal Processing (36/63) as known, and the polarity and delay checks from Mixing Drums (36/85).
- Who takes it
- Part 1 of the Mixing & Mastering specialization, inside the Music Production specialization of Music; also for Sound Engineering students.
- What it is for
- A low end that is solid on big and small speakers, and guitars and keys that are wide and full without covering the voice.
- Comes after
- Session Prep, Gain Staging & Mix Workflow (36/84).
- Leads to
- Finishing the Mix: Space, Automation & Translation (36/88). Beside it: Mixing Drums (36/85) and Mixing Vocals (36/87).
- What this piece gives you
- Not a finished low end; that takes years of listening on many systems. You will see why kick and bass collide, why a bass "disappears" on a phone and how harmonics rescue it, and why a double is wider than a copy. Easier first reads: Music Technology & Production (35/73) and Acoustics, Tuning & Timbre (35/68).
The low end is shared real estate
A kick drum's weight sits around 50 to 100 Hz. A bass guitar's lowest string, E1, is 41 Hz, and its notes up the neck run through the same range. When the two play together, they add: the sum is louder than either, and on the kick's first 100 milliseconds the combined peak may be higher than the mix can hold. Two tracks that occupy the same band mask each other, as Mixing & Signal Processing (36/63) explained for EQ.
There are three ways to split it. By frequency: choose which instrument owns which band, such as a kick that owns the very bottom and a bass that owns the low mids, or the other way round. Not everyone follows this; Sound On Sound has run pieces that set the idea aside, so treat it as one option and listen. By time: a sidechain compressor (or a dynamic EQ on the bass triggered by the kick) turns the bass down for a moment each time the kick hits, then lets it back. By choice of notes: shorter bass notes, or a kick pattern that avoids the bass's accents. Try the second in the tool below. Notice that the spectra do not move: ducking separates the two in time, not in frequency.
Sound isn't available in this reader; the spectra and numbers still work.
Both sounds are synthesised in the page: a kick of two sine sweeps and a click, and a sine-like bass with a few weak harmonics (A1 55 Hz, then E1 41.2 Hz) at 100 bpm. Spectra come from a 32,768-point FFT with a Hann window. The sidechain follows the kick's own level and cuts the bass in 2 ms, releasing over the time you set.
Without JavaScript: with the duck at 6 dB and 120 ms, the combined peak of kick and bass falls from about 0 dBFS to −1.4 dBFS, the bass just after the first kick is turned down by about 6 dB (−8.8 to −14.7 dB) and between kicks it is nearly untouched (−15.8 to −16.6 dB). About 88% of the kick's energy and 98% of the bass note's lie between 40 and 100 Hz.
Two cautions. A duck that is too deep or too slow gives the bass a pumping, seasick motion; match the release to the tempo so the bass is fully back before the next hit. And check the pair in mono: both are centred, so mono shows how they really add up. A related idea is mono bass, covered below.
Bass that survives a small speaker
The fundamental of a bass note, 41 Hz for the lowest string of a four-string bass (about 31 Hz for the low B of a five-string), is below what most phone and laptop speakers can reproduce at all. Yet you hear the note. The reason is that the ear fills in a missing fundamental: if the harmonics at two, three and four times the frequency are present, you perceive a pitch at the fundamental, even if it is not there. J. C. R. Licklider reported in 1954 that listeners still heard it when noise masked any distortion in the ear itself. Products sold as "psychoacoustic bass" exploit this.
So the engineer's task is to make sure the harmonics are there. A clean bass played through a sine-like amplifier is mostly fundamental; a bass with some saturation (soft clipping, tube or tape emulation, an amp's distortion) has strong harmonics, and they survive a small speaker. Try it: pick a string, set the drive, and switch the sound between full range and a phone-style filter. The two bass sounds below have the same level before the filter; look at what is left after.
Sound isn't available in this reader; the spectrum and numbers still work.
The note is eight harmonics with a decaying envelope; saturation is an asymmetric tanh curve, rescaled so the level before the filter is the same. The "speaker" is a 2nd-order high-pass applied twice (4th order); real phone speakers vary, and many can play little below about 150 to 300 Hz (a rough range, not a measurement of any phone).
Without JavaScript: on A1 (55 Hz) with a 180 Hz corner and the drive at 40%, the clean note keeps about 28 dB less level through the filter than before it and the saturated note about 22 dB less, because its new harmonics sit above the corner.
Saturation is a trade: more harmonics also means more grit, and a bass that sounds clean on large speakers may need a parallel saturated copy rather than saturating all of it. Many engineers split the bass, keeping a clean low end and adding a saturated, high-passed copy for the harmonics, which preserves the weight and gives the phone something to play.
DI plus amp, and why the bottom stays in the middle
Bass is often recorded twice: a DI (direct injection; a clean line from the instrument) and a microphone on the amplifier. The DI has the clean low end and the string detail; the amp has the grit and the room. Blending them gives both, with a catch from Mixing Drums (36/85): the mic is 30 cm or so from the speaker cone, about 0.9 ms later than the DI, which cancels part of the low mids. Zoom in, align them (or flip polarity and keep the fuller sound), then balance. A common split is to take the lows from the DI and the mids from the amp.
Keep the bottom mono. Below roughly 100 to 150 Hz, keep the bass, kick and any sub in the centre. The reasons are practical: a 100 Hz wave is about 3.4 m long, so a listener cannot tell where it comes from; a wide low end wastes energy and makes one speaker work harder than the other; two slightly different low signals can cancel when the system is summed to mono; and vinyl cutting is easier with the low end in the middle (see Mastering in Practice, 36/89). It is a practical convention, not a law: electronic music can have deliberate stereo sub effects, and many mixes keep the bass fully in the middle without ever thinking about it. If a wide instrument has low content, high-pass its sides or use a "bass mono" or mid/side tool below 100 to 150 Hz.
Electric guitars: doubles, width and the scooped-mids trap
A single guitar panned hard left leaves the right side empty. The classic fix is double-tracking: the guitarist plays the part twice, and the two takes are panned left and right. Because the takes differ slightly in timing, pitch and tone, the ear hears one very wide guitar. A copy of the same take, delayed by 10 to 30 ms and panned opposite, is often tried as a shortcut (the Haas effect, from Mixing & Signal Processing), and it is not the same. A delayed copy is a comb filter: in mono it has notches, and the notches are also somewhat audible in stereo as a hollow tone. A real second take decorrelates: its mono loss is also about 3 dB on average, but it falls differently on each note instead of at the same frequencies every time, so nothing is systematically hollowed out.
Sound isn't available in this reader; the bars and numbers still work.
The riff is four power chords of sawtooth-like harmonics through a soft clipper and a low-pass, 1.7 s, synthesised in the page. "Two real takes" are generated twice with different random pitch offsets (up to 8 cents), timing offsets (up to 12 ms) and harmonic phases. Correlation is measured on the samples; the bars compare the mono sum (L + R) ÷ 2 with the average of the channels in 20 third-octave bands.
Without JavaScript: with one take in the middle the mono sum matches the channels (0 dB, correlation 1.00). The delayed copy (15 ms) loses about 3.4 dB in mono, with notches at 1 ÷ (2 × delay), such as 33 Hz, 100 Hz and 167 Hz; two real takes lose about 2.9 dB on average.
The scooped-mids trap is the other thing to know. A guitar with its midrange cut and its bass and treble boosted ("a smile on the EQ") sounds enormous on its own. In the mix it has gone: the 400 Hz to 3 kHz region is where the guitar's body and note definition live, and where it competes with the voice, so scooping the guitar makes it vanish behind the vocal and the snare, and a smaller guitar with its mids intact usually sounds bigger in the song. Cut what masks the voice, a few dB at a time, rather than hollowing the instrument out.
Two more points on electric guitars. Reamping means sending the recorded clean DI back out of the interface to a real amplifier (through a reamp box that matches level and impedance) and recording the result again, so the amp sound is chosen at mixing, not at tracking. Amp simulators are good enough that many released records use them; their honest limits are in feel and in how they respond to the player's touch, and in a harsh upper-mid "fizz" that a low-pass often helps. Keep the DI either way, so the choice can change.
Acoustic guitar, keys and pads
An acoustic guitar has a body resonance in the low range that sounds like boom, and a high-pass (see the ranges below) is the usual cure. Pick noise and strum scrape live in the 3 to 5 kHz zone; a de-esser or a light dynamic cut tames them. Keys and pads are wide, sustained and full of low mids, so they are the usual victims and culprits of mud: high-pass them higher than you think, pan them opposite the guitars so the two do not fight, and let each part own its rhythmic gaps. Where a pad and a guitar play the same notes, ask which one the song needs there, and mute or lower the other.
Starting points, not rules
Ranges from books and magazines, to give your sweeps a place to start. Basses, guitars, rooms and songs differ. Let your ears overrule every number, and judge each change at the same loudness.
| What | Typical range | Why people look here |
|---|---|---|
| Bass: high-pass | 30–40 Hz | Just below the lowest note (E1 is 41 Hz) to remove rumble. |
| Bass: weight, growl, definition | 60–120 Hz; 700 Hz–1 kHz; 2–3 kHz | Where the three characters of a bass sit. |
| Bass compressor | 3:1–6:1, 3–6 dB reduction | Evens out notes; slow attack keeps the finger attack. |
| Kick/bass sidechain | 2–6 dB, release 50–150 ms | Short and tempo-matched. |
| Mono below | 100–150 Hz | Bass, kick and sub in the centre. |
| Electric guitar: high-pass | 80–120 Hz | Clears room for bass and kick. |
| Electric guitar: low-pass | 8–12 kHz | Tames amp fizz. |
| Guitar: mud, presence | 200–400 Hz cut; 2–4 kHz | Mud sits low; presence competes with the voice. |
| Doubles | panned 70–100% left and right | Hard enough to be wide, with the vocal in the centre. |
| Acoustic guitar: high-pass | 80–120 Hz | Body boom. |
| Pads and keys: high-pass | 150–300 Hz | Low mids are crowded; use your ears. |
Where these come from: the conventional teaching ranges in Owsinski's The Mixing Engineer's Handbook, Senior's Mixing Secrets for the Small Studio, Izhaki's Mixing Audio and Sound On Sound's mixing articles. Open-string frequencies come from f = 440 · 2(n − 69)/12 (E1 41.2 Hz, A1 55 Hz, D2 73.4 Hz, G2 98 Hz; low B 30.9 Hz).
Common mistakes and their fixes
1. Judging the low end on speakers that can't play it. Small monitors and headphones mislead. Fix: check on a phone, a car and a bigger system, and look at a spectrum meter.
2. Boosting the bass fundamental to make it audible. It eats headroom and the speaker still can't play it. Fix: add harmonics with saturation.
3. Bass and kick both at full weight. Fix: choose who owns the very bottom, or duck the bass off the kick by 2 to 6 dB.
4. A copy delayed by 15 ms as a "double". It collapses in mono. Fix: record a second take, or use a decorrelating effect and check mono.
5. Scooped guitar mids. Fix: cut a narrow band that masks the voice, not the whole middle.
6. Wide low content. A stereo pad or guitar with deep lows muddies mono and the sub. Fix: high-pass the sides.
7. Ignoring the DI and amp timing. Fix: align or flip, then blend.
Genre and world differences
- Rock. Bass and kick split the low end; doubled rhythm guitars are panned hard, with the lead guitar in the middle.
- Metal. Low tunings push guitar fundamentals into the bass's range, so high-pass the guitars firmly and keep the bass tight; four-track doubles (two each side) are common.
- Hip-hop and trap. The 808 is the bass: a long, sine-like note, often with saturation so it speaks on small speakers, tuned to the song's key, and managed against the kick (see 36/69).
- Pop. Synth bass, tightly controlled and mono below 100 to 150 Hz, with guitars as a texture.
- Reggaetón and dembow. A programmed sub with long notes and glides; tuning and keeping it out of the kick's way, with the vocal clearly in front.
- Cumbia and Latin traditions. The bass often carries the melody of the rhythm, with guitars and keys playing short, interlocking patterns; panning them apart keeps the parts clear.
- Afrobeats, highlife and amapiano. Interlocking clean guitar lines work best panned apart; in amapiano the log drum and the sub share the bottom and need the same care as kick and bass.
- Jazz, folk and acoustic. An upright bass is often a microphone and a pickup blended; compression is light, and the natural resonance is part of the sound.
Practise on a real session
Mike Senior's free multitrack library on his Cambridge Music Technology site (cambridge-mt.com) is offered for mixing practice and education. Its search lets you require instruments: ask for a song with bass guitar, electric guitars and a kick, so you get the DI-versus-amp and kick-versus-bass problems with real recordings. Read the terms beside each download before sharing a mix.
Start with the kick and bass alone, solved in mono, then add the guitars, then the rest. Compare to a commercial track in the same style at the same loudness, on a phone as well.
The bass and guitars checklist
In the order you would do it at the desk. Ticks aren't saved.
DAW notes. A sidechain compressor takes a key input from another track: in REAPER from a send to channels 3/4, in Pro Tools a key input on the plug-in, in Logic, Ableton Live and FL Studio from the side-chain menu or routing of the plug-in. Polarity and delay: see 36/85.
What a graduate just sees
- Kick and bass add in the same band; separate them by frequency, by time (a sidechain duck) or by note choice.
- A duck changes when the bass plays, not what it contains.
- A bass that vanishes on a phone has no harmonics; saturation puts them there, and the ear supplies the missing fundamental.
- DI and amp are two copies a fraction of a millisecond apart; align or flip, then blend.
- Keep everything below about 100 to 150 Hz in the middle.
- A delayed copy is a comb filter; a real second take is a different signal, so it is wide and survives mono.
- A scooped guitar sounds big alone and gone in the mix.
- Every range here is a place to start looking; the song decides.
Where it shows up next
- Mixing Vocals (36/87): the guitars' 2 to 4 kHz presence and the voice share a band; the cuts you make here make room for it.
- Finishing the Mix (36/88): mono checks, width and translation tests are where bass and guitar choices pay off.
- Mastering in Practice (36/89): low end in mono matters for vinyl and for a limiter that reacts to the loudest low notes.
- Beatmaking, Sampling & Arrangement (36/69): 808 bass is tuned and shaped in the same way.
If you want to go deeper
- 📘 Textbook. Roey Izhaki, Mixing Audio (Routledge, 3rd edition, 2017); Mike Senior, Mixing Secrets for the Small Studio (Routledge, 2nd edition 2018).
- 📘 Short and practical. Bobby Owsinski, The Mixing Engineer's Handbook (Cengage).
- 🌐 Article. Sound On Sound, "Mix Tips For Kick & Bass".
- 🌐 Free multitracks. The Cambridge Music Technology "Mixing Secrets" free multitrack library (cambridge-mt.com).
- 🌐 Free course. MIT OpenCourseWare 21M.380 Music and Technology: Recording Techniques and Audio Production (Fall 2016).
The bottom of a mix is one shared space: give it to one thing at a time, in frequency or in time, keep it in the middle, and put harmonics in the bass so that small speakers still play it. In the middle of the mix, width comes from different takes, not copies, and the guitar's mids are the part to protect, not remove.
Try this: play a favourite song on your phone, then through earbuds or good headphones. Which instrument changes most? Then hum the bass line, and notice you can follow it even on the phone.
Sources. J. C. R. Licklider, 1954, on the missing fundamental (summarised on Wikipedia, "Missing fundamental"); Sound On Sound, "Mix Tips For Kick & Bass" and "Phase Demystified"; Wallach, Newman and Rosenzweig, 1949, and Haas, 1951, for the precedence effect, as in Mixing & Signal Processing (36/63); Owsinski, Senior and Izhaki for the conventional ranges.
- 📘 Senior, Mixing Secrets for the Small Studio
- 🌐 Sound On Sound: Mix Tips For Kick & Bass