How to Mix Strings

Updated May 2026

Strings occupy almost the full audible frequency spectrum. A cello section extends from roughly 65Hz at the bottom of a low C to harmonics that brush 10kHz and beyond. Violins overlap in the midrange with everything that matters most in a mix: vocals, guitars, synths, and snares. Add in the fact that string libraries are often recorded in reverberant halls, and you have an instrument that fights for space with nearly everything else while simultaneously being one of the most emotionally impactful sounds a producer can use.

The mistakes are predictable. Over-bright sampled strings that cut through the mix like a bandsaw. Under-processed real ensemble recordings that sound distant and muddy. Hall reverb used so liberally that strings become more smear than sound. MIDI strings with machine-gun uniformity that no listener consciously identifies as mechanical but every listener feels is somehow wrong.

This guide addresses each of those failure modes directly — covering signal chain setup, EQ decisions, compression, reverb strategy, MIDI humanization, mic blending for real ensembles, and how to integrate strings into electronic productions without losing either the strings or the low-end impact of your drums and bass.

Setting Up Your String Tracks: The Basics First

Before you touch EQ or reverb, make sure your string tracks are organized correctly. Whether you're working with a sampled library or a real recording, the same organizational principles apply.

Separate your string families — violins, violas, cellos, basses — onto their own tracks or bus groups. Do not commit everything to a single "strings" bus until you've done your EQ and balance work at the individual section level. Each section has a different frequency character, a different dynamic range, and a different role in the arrangement. Trying to EQ or compress all of them as a single unit is like trying to EQ a full band mix from a single stereo bus — you end up making compromises that hurt everyone.

Group your sections onto a strings bus and feed that bus into a mix bus. This gives you three levels of control: individual section, strings group, and overall mix. Insert processing happens at the section level (section-specific EQ and compression), send processing happens from the strings bus (shared reverb), and any final glue or bus limiting happens at the mix bus.

This hierarchy matters. Attempting to EQ sampled violin harshness from a master bus is like trying to tune an individual instrument through a wall. The problem exists at the section level and needs to be fixed there.

Set gain staging correctly from the start. String libraries in particular tend to come in hot. Bring all your string tracks down until the strings bus peaks around -12 to -10 dBFS. You want headroom above your strings before you compress them. Compressing a clipping signal produces artifacts that no amount of downstream processing can fix.

EQ for Strings: High-Pass, Harshness, and the Air Shelf

String EQ follows a clear sequence: clean the bottom, tame the harshness zone, then add air if needed. Skipping to the air shelf without addressing the midrange is one of the most common mistakes in string mixing — you end up boosting the top of a signal that still has a problematic, synthetic-sounding midrange.

High-Pass Filtering: Finding the Right Floor

High-pass filtering strings is one of the most commonly misapplied steps in string mixing. Producers either skip it entirely — leaving in low-frequency mud that masks the kick and bass — or overdo it, cutting so high that the warmth that makes strings feel physical and bowed disappears with the mud.

The safe high-pass range for strings is 50–100Hz. Cutting below 80Hz is almost always right. Below this frequency, strings produce very little musical content, but mic bleed, stand rumble, HVAC noise, and low-frequency resonances accumulate. Removing this cleans up the mix without touching anything perceptible as string tone.

Where producers go wrong is cutting above 120Hz in an attempt to "make room for the bass." Frequencies between 80Hz and 200Hz contain the chest resonance of cellos and bass strings — the physical feeling of a bowed instrument producing sound. If you cut these frequencies, sampled strings immediately lose the quality of presence and weight that convinces the ear these are real instruments. They start to sound like a sine wave approximation of strings rather than strings.

Use a 12dB/octave or 18dB/octave slope. A 6dB/octave slope rolls off too gently and lets too much low-frequency energy through at the threshold point. The sharper slope gives you a clean cut without creating phase problems in the region just above the cutoff.

The 2–4kHz Harshness Zone: The Core Problem With Sampled Strings

If you've ever loaded a string library and felt that something was wrong before you'd even added any effects — that harsh, almost electronic quality that makes the strings sound synthetic and aggressive — the 2–4kHz frequency range is usually responsible. This is the harshness zone for strings and it is the number one reason sampled strings sound fake.

Real string ensembles spread across a room. You're hearing dozens of instruments acoustically blending in three-dimensional space before any of that sound reaches the microphones. The 2–4kHz region — which contains the upper partials of bowing noise and string attack — gets naturally diffused and softened in that acoustic process. Individual harshness cancels out across instruments playing slightly differently. The result is smooth, present, but never harsh.

Sampled libraries, even excellent ones, are typically built from close-mic recordings or heavily processed spot mics. The acoustic diffusion that smooths the 2–4kHz region in a real room doesn't happen. Instead, every bowing attack, every scratchy upper partial, is captured at full intensity. When you then layer multiple string tracks from the same or similar samples, the harshness compounds rather than cancels.

The fix is a narrow bell cut in the 2–4kHz range, typically -2 to -5dB with a Q of 2–4. Don't use a broad cut — you'll start removing frequencies on either side that carry useful musical information. Use a spectrum analyzer to find the specific resonant peak within that range (it moves depending on the library and the note being played) and cut there with surgical precision.

Then listen. The strings should immediately sound less harsh, less electronic, less like they're competing for the same midrange real estate as your lead vocal. If they now sound slightly dull, that's correct — you may have over-cut. Back off the cut by 1dB until you find the balance point where harshness is controlled but the strings still have presence.

For real recorded ensembles, the 2–4kHz problem is less severe but can still appear if the recording used mostly close mics or if the session happened in a very dry room. The same cut applies — just narrower and shallower, typically -1 to -2dB at most.

The Air Shelf: Adding Shimmer Without Harshness

Once you've cleaned the bottom and tamed the midrange, a gentle high-shelf boost can transform strings from merely competent to genuinely cinematic. The target is 8–12kHz, boosted 2–3dB as a shelf. This range adds the shimmer and breathiness of real string tone — the overtone sparkle of rosin on horsehair, the complex high-frequency texture that makes strings sound alive.

A boost at 200–300Hz can add warmth to strings that feel thin, particularly useful for solo violin parts or libraries that prioritize articulation accuracy over body and tone. Keep this gentle — no more than 2–3dB — and check it in the context of the full mix, where it can quickly create a bloated, woofy quality if overdone.

Avoid boosting in the 1–3kHz range. This is the zone where string harshness and nasality live, and adding energy here makes strings sound like they're being amplified through a telephone. The only reason to touch this range is to cut it.

Compression on Strings: Less Is Almost Always More

Strings are one of the most dynamically expressive instruments in an ensemble. The swell of a section moving from pianissimo to fortissimo is a large part of what makes strings emotionally powerful. Over-compression removes exactly this quality, turning strings into a flat, monotonous wall of texture that technically occupies frequency space in your mix but does nothing for the listener emotionally.

For sampled strings, light compression with a 2:1 ratio, slow attack (30–80ms), and moderate release (150–300ms) is a good starting point. The slow attack lets the initial transient of each note through — this preserves the articulation of the passage and keeps the strings feeling live and responsive. The compression kicks in slightly after the attack to gently manage the sustain and tail of each phrase.

For real ensemble recordings, you may not need compression at all on individual section tracks if the performance was well-played and the dynamics were appropriate to the arrangement. Where compression helps most with real strings is on the strings group bus, as a very light glue compressor — a 1.5:1 or 2:1 ratio with a very slow attack and a moderate release, gaining maybe 1–2dB of gain reduction at peaks. This ties the sections together subtly without killing their dynamics.

Avoid fast attack times on strings. A fast attack (under 10ms) catches the bow attack of each note and turns strings into a punchy, staccato instrument — which is the opposite of what you typically want unless you're specifically working with pizzicato passages. Even then, the natural attack of plucked strings is part of their character.

If you're using a plugin like the FabFilter Pro-C 2, the "Natural" or "Mastering" style modes with a slow attack give you transparent compression that doesn't color the string tone. Avoid aggressive "Punch" or "Vocal" modes designed for transient-heavy sources. See our in-depth guide to compression for beginners for more on ratio and attack choices.

Reverb Strategy for Strings: Hall Sends, Pre-Delay, and Genre

Reverb is where string mixing goes wrong more consistently than anywhere else. The failure mode is almost always the same: too much reverb, applied as an insert rather than a send, with no pre-delay, resulting in strings that sound like they were recorded inside a swimming pool. You can feel the space more than you can hear the strings.

Send vs. Insert: Why It Matters

Always use reverb on a send bus for strings, not inserted directly on the channel. There are two reasons for this.

First, a send bus allows you to feed multiple string sections to the same reverb, which means the reverb treats the entire ensemble as a single acoustic source — the way a real room would. If you insert reverb on each individual track, each section gets its own reverb tail, and they don't blend the way they would in a real hall.

Second, a send bus allows you to blend the wet reverb signal independently of the dry string signal. You can automate the reverb send amount separately from the string fader, pulling back the reverb in dense passages and pushing it forward in exposed moments. This is impossible when the reverb is inserted on the channel.

For more on building effective send routing, see our guide to using send effects in your DAW.

Hall Reverb for Orchestral Strings

For orchestral or cinematic strings, use a long hall reverb with a 2.5–4 second decay. Set the pre-delay to 10–20ms. Pre-delay is the gap between the dry signal and the first reflection — it creates perceptual separation between the strings and the reverb, keeping the strings sounding present and defined rather than buried in the tail.

Keep the reverb's low-frequency rolloff active. Most hall reverb plugins allow you to cut the low end of the reverb itself — set a high-pass on the reverb return around 150–200Hz. This prevents the hall from adding a boomy, washy low end that muddies your kick and bass without affecting how the strings themselves sound.

The Lexicon 480L and the Bricasti M7 are the gold standards for real hardware hall reverb on strings. For plugins, Altiverb, Lexicon PCM Native, and Valhalla Room are all widely used. Valhalla Room's "Concert Hall" and "Dark Hall" modes are particularly well-regarded for cinematic string work and represent excellent value. Spaces II (Logic Pro's built-in convolution reverb) contains high-quality real hall impulse responses that rival many dedicated paid plugins.

Shorter Reverb for Pop and Trap Strings

For pop or trap strings, a shorter room or plate reverb with a 0.8–1.5 second decay keeps strings punchy without washing out the mix. Trap in particular relies on a kick-dominated low end and a clear stereo field — a 4-second hall reverb on strings will compete with and obscure the kick every time the reverb tail coincides with a hit.

A plate reverb gives strings a lush, slightly vintage quality without the depth of a full hall. A medium room keeps strings feeling physical and present. Both can work — the choice depends on whether you want the strings to feel close and immediate (room) or to have a sense of studio-era sweep (plate).

Sidechaining Reverb Tails to the Kick

One of the most effective techniques for blending strings with trap or electronic drums is sidechaining the reverb return (not the dry string channel) to the kick drum. When the kick hits, the reverb tail ducks briefly, then swells back up after the kick's transient passes. This creates a natural breathing quality that feels less like engineering and more like acoustic physics — the same way a loud sound in a room momentarily dominates before ambience fills back in.

Sidechain the reverb return with a fast attack (1–5ms) and a medium release (80–150ms). Keep the gain reduction modest — 4–6dB is usually enough to create the effect without making it obvious. If the pumping is clearly audible, either reduce the gain reduction or slow the attack slightly.

MIDI Humanization: Making Sampled Strings Sound Real

No amount of EQ or reverb will make mechanically sequenced MIDI strings sound human. The ear is extraordinarily sensitive to the patterns of real musical performance — the microscopic timing variations, the way velocity changes note by note, the way a real player breathes before a phrase, the way bow pressure changes mid-note. MIDI strings that ignore all of these qualities sound wrong in a way that's hard to articulate but impossible to miss.

Velocity Variation

The most immediately effective humanization technique is velocity variation of ±15–25 across notes. In a real string passage, no two notes are played at identical dynamic levels. Even in a section where thirty players are attempting to match each other, micro-differences in bow speed, bow pressure, and contact point create a naturally varied velocity profile across the ensemble.

Apply velocity variation in your MIDI editor. If your DAW supports it, use a randomize or humanize function with a range of ±15–25 velocity points. Alternatively, draw variation in manually — it should feel slightly irregular but still follow the broad dynamic shape of the phrase. A rising phrase might have an overall upward velocity trend with individual notes varying around that trend by ±15 points.

Be careful not to let velocities drop so low that the library switches to a quieter dynamic layer mid-phrase at an inappropriate moment. Most libraries have velocity-layer switching points — learn where your library switches and keep your variation within the intended dynamic layer unless you're specifically trying to cross a layer boundary for a dynamic swell.

Timing Offsets

In a real string section, players do not all start notes at exactly the same millisecond. Even a well-rehearsed ensemble has micro-timing differences of 5–20ms between individual players, and larger differences — 20–50ms — between different sections. This timing spread is part of what creates the characteristic ensemble sound: the slight smear of multiple attacks layered together, which is perceptually very different from a single perfectly synchronized attack.

If your library provides multiple mic positions for different string sections, offset them by 5–15ms in your DAW. If you're layering a sampled violin 1 and violin 2 track, delay one by 8ms. If you have separate section tracks (first violins, second violins, violas, cellos), apply different offsets to each — perhaps 0ms, 7ms, 12ms, and 5ms across the four sections.

These offsets will not be perceptible as timing errors at these small values — the human ear requires approximately 20–30ms to perceive two events as separate rather than as a single event with a particular quality. What the brain hears instead is ensemble blend and depth.

Expression and CC11 Automation

Most professional string libraries respond to MIDI CC11 (Expression) or CC1 (Modulation) for dynamic control within a performance. Using a static velocity with no CC automation is the equivalent of recording a guitar player who never varies their pick attack or pressure across an entire song — technically possible, but clearly wrong.

Draw CC11 automation curves for each phrase. A long sustained note might start at CC11 value 80, swell to 110 at the phrase peak, and release back to 75 as it decays. A staccato passage might have small CC11 bumps on each note — up on the attack, down at the release. Even small variations (10–15 CC values up and down across a phrase) make a dramatic difference to perceived realism.

Vibrato Depth Modulation

Real string players do not apply uniform vibrato from the first millisecond of every note. Vibrato typically starts slightly after the note attack, grows to full depth over 100–200ms, and may change depth on longer notes. Many libraries allow you to control vibrato depth with a MIDI CC (often CC1 or a dedicated CC).

Program vibrato to start at a lower depth (CC value 30–40) and grow to full depth (80–100) over the first 200ms of each sustained note. On shorter notes, vibrato may not have time to fully develop — and leaving it at a low depth for these notes makes them sound more like a real performance. For a detailed look at how expression control translates to other instruments, see our guide to mixing vocals, where similar automation principles apply to pitch and dynamics.

Mixing Real String Ensembles: Mic Blending and Session Workflow

Recording a real string ensemble gives you something no sampled library can fully replicate: acoustic ensemble blend, natural room sound, and the performance energy of living musicians responding to each other in real time. But real ensemble recordings come with their own set of mixing challenges — particularly around mic blending and managing the relationship between what the room mics captured and what the close mics reveal.

The Room Mic / Close Mic Relationship

Professional string sessions typically capture several mic positions: a stereo room pair at some distance from the ensemble (often a Decca tree or an ORTF pair), spot mics over individual sections or desks, and sometimes a mono ambient mic for depth. The instinct for producers who haven't worked with ensemble recordings is to start with the close or spot mics because they sound clear, defined, and controlled. This is the wrong starting point.

Start with the room mics as your primary sound. The room mics carry the acoustic blend of the ensemble — the way the sound of thirty instruments has merged into a single coherent wash before it reaches the capsule. They carry the hall reverb, the natural frequency balance of the space, and the ensemble depth that spot mics cannot provide. Room mics are the soul of a real string recording.

Blend the spot or close mics underneath the room mics to add definition. The close mics provide attack clarity — the detail of bow articulation, the crispness of staccato passages, the precision of entrance timing. A ratio of roughly 70% room / 30% close is a good starting point, but this will vary depending on the recording environment and how the session was miked.

If the close mics dominate, strings sound clinical and over-defined — you hear too much individual instrument detail and lose the ensemble blend. This is a common problem with live-to-multitrack sessions recorded in pop or film scoring contexts where engineers default to closer mic positions for safety. In these cases, the room mics (if available) become even more essential as your primary source.

Phase Alignment Between Mic Positions

Whenever you're blending multiple mic positions from the same ensemble, phase alignment is essential. Room mics and close mics capture the same sound at different distances — the room mics hear the sound slightly later than the close mics, creating a time offset that causes comb filtering when both signals are mixed together.

Use a plugin like iZotope RX or a dedicated phase alignment tool (such as SoundRadix Auto-Align) to align your mic positions. The process involves delaying the closer mic to match the timing of the room mic, or using a sample-accurate nudge in your DAW. Even small phase misalignments of a few milliseconds can significantly affect the frequency balance of the combined signal.

Check phase alignment in mono. Collapse both mic positions to mono and listen for thin, hollow, or frequency-cancelled qualities. If the combined mono signal sounds thinner or more hollow than either individual signal, you have a phase problem to correct.

EQ on Real Ensemble Recordings

The same EQ principles apply to real recordings as to sampled libraries, but with less aggressive settings. The 2–4kHz harshness is usually milder (especially if room mics dominate), but the high-pass filter is equally important — room mics especially capture significant low-frequency room rumble that needs to be removed.

For spot mics, you may find a problematic boxiness between 300Hz and 600Hz from close proximity to the instruments. A narrow bell cut in this region (-2 to -3dB) opens up the sound and makes the spot mics blend more naturally with the room mics. Avoid boosting anything in the spot mics before listening to how they blend — what sounds right in solo may sound completely different in context.

Blending Strings With Electronic Elements: Pop, Trap, and Hybrid Scoring

The collision of live or sampled strings with electronic elements — 808s, trap drums, synthesizers, programmed bass — is one of the most interesting and most challenging mixing scenarios a modern producer encounters. Done well, it sounds like the most natural combination in the world. Done poorly, it sounds like two completely different productions playing simultaneously.

Frequency Territory and the 808

The 808 bass dominates modern trap and hip-hop production from roughly 40Hz up to 200Hz. This is also where the warmth and chest resonance of a cello section lives. Trying to preserve both at full level in the same frequency range creates a competition that neither wins — the 808 loses its sub punch and the cellos lose their physical weight.

The solution is not to EQ the cellos to remove all frequencies below 200Hz (that destroys their character) but to give the 808 clear ownership of the sub range below 80–100Hz while allowing the cello warmth to occupy the 100–200Hz range above it. High-pass the strings at 80Hz (as recommended throughout this guide) and let the 808 own everything below. Use a high-shelf boost on the 808 if needed to reinforce its presence in the 100–200Hz range so it still feels powerful above its sub zone. For more on building this low-end relationship, see our dedicated guide on how to mix bass.

Stereo Width and the Kick/Snare Relationship

In electronic productions, the kick and snare are typically centered — they hit at the exact center of the stereo field for maximum impact. Wide stereo strings that extend into the center of the stereo field compete with the kick and snare for that central space, making the rhythm section feel less impactful.

Keep strings narrow or mid-heavy in stereo when mixing with electronic drums. You can achieve this by reducing the stereo width of the strings bus using a mid-side plugin, or by keeping string parts that occupy the low and low-mid ranges (cello, viola) more centered while allowing only high strings (violins, upper harmonics) to extend wide. The wide upper frequencies add atmosphere and presence without competing with the punch of the rhythm section.

Alternatively, for a cinematic trap aesthetic where the strings are an intentional stylistic feature rather than a background texture, you can spread strings wide as a deliberate statement. The key is intentionality — a wide string spread that competes with the kick is a mistake, but a wide string spread that is clearly the dominant atmospheric element in a mix can be a deliberate production choice. For reference techniques on this kind of arrangement, see our guide to making trap beats.

Panning: Orchestral Seating vs. Pop Flexibility

For cinematic or orchestral projects, the traditional orchestral seating arrangement provides a natural panning scheme: violins hard left, violas center-left, cellos center-right, basses right. This replicates the physical seating of a real orchestra as heard from the audience, and it creates a natural frequency gradient across the stereo field — lighter, brighter tones on the left, warmer, heavier tones on the right.

For pop and trap, you have more freedom. Many producers keep strings relatively centered for punch, especially if the strings are providing harmonic or rhythmic support rather than a sweeping melody. For atmospheric string pads or textural elements, spreading wide in stereo works well — these parts don't need to be punch-compatible because they're not competing with the rhythm section for impact.

In hybrid scoring (film music that combines orchestra with electronics), a mixed approach is common: use the orchestral panning scheme for the acoustic sections, but allow electronic elements to occupy the full stereo field independently. The electronic bass and drums center, the strings follow orchestral seating, and synth pads or electronic textures fill whatever space remains. This preserves the realism of the orchestral elements while fully exploiting the spatial possibilities of electronic production.

Layering Strings With Synth Pads

String pads and synth pads occupy similar frequency territory and serve similar emotional functions. When you layer them, they can either reinforce each other beautifully or fight each other to the point where neither is recognizable.

The key to successful layering is frequency differentiation. Use the synth pad to occupy the upper harmonic range (above 2kHz) where strings are thinned out, and use the strings to provide the mid-frequency warmth and body that synth pads often lack. Low-pass filter the synth pad to remove its upper frequencies, and high-pass the strings (more aggressively than usual, perhaps 200Hz) to keep them in the mid-forward range. This creates a composite sound where the strings and pad are heard as a single, rich texture rather than two competing elements.

Alternatively, reserve strings for moments when synth pads are absent and vice versa. Many film scores and cinematic pop tracks use strings and synth pads as a call-and-response dynamic — the synth builds tension, the strings release it, or the strings establish a theme and the synth develops it electronically. If you're struggling to make strings and synths coexist, separating their appearance in the timeline may be more powerful than trying to layer them simultaneously. This kind of arrangement thinking is covered in depth in our guide to making cinematic music.

Choosing and Working With String Libraries

The mixing techniques in this guide apply regardless of which string library you use. However, the quality and flexibility of your library significantly affects how much work is required to achieve a convincing result — and which problems you're likely to encounter most often.

What Makes a Good String Library

The most important features in a professional string library are multiple mic positions, multiple dynamic layers, and a comprehensive articulation set. These three features directly enable the mixing and humanization techniques described throughout this guide.

Multiple mic positions (typically Close, Tree, Mid, Far, and Outrigger) allow you to blend your own acoustic perspective — close-dominated for pop, room-dominated for cinematic work — rather than being locked into whatever perspective the library designers chose as their primary sound.

Multiple dynamic layers (typically pp, mp, mf, ff, and sometimes ppp and fff) allow you to use velocity not just as a volume control but as a true dynamic character change — the difference between a pianissimo passage where the bow is barely touching the string and a fortissimo passage where the section is playing at full force. Without multiple dynamic layers, velocity variation (as described in the humanization section) only changes volume, not character.

Comprehensive articulations — sustain, short bow, spiccato, pizzicato, col legno, harmonics, sul ponticello, sul tasto — give you the palette to write and arrange for the full expressive range of a string ensemble rather than relying on a single all-purpose sustain patch.

Notable Libraries

The BBC Symphony Orchestra (BBCSO) Discover/Core/Professional from Spitfire Audio is one of the most widely used orchestral libraries for cinematic and film work. It was recorded at the Maida Vale Studios in London with a genuinely large ensemble, multiple mic positions, and a comprehensive articulation set. The Discover edition is free and provides a full-range orchestral palette with a single close mic position. Core and Professional add additional mic positions and ensemble sizes.

Cinematic Studio Strings is particularly well-regarded for its legato performance and its realistic string sound in the 200–2kHz range — the area that most libraries struggle with. Its slower attack times and more diffuse initial transient make it sound more like a real ensemble and less like a sample trigger. It requires more careful MIDI programming than libraries with a more immediate attack but rewards the investment with a more authentic sound.

Spitfire LABS Strings is a free library that provides a surprisingly usable string sound with a single articulation and mic position. The mixing techniques in this guide — particularly the 2–4kHz cut and the hall reverb send — are especially effective on LABS Strings, which tends to be bright and harsh without EQ treatment.

Kontakt Factory Library strings and the strings included with stock DAW instruments (Logic Pro's Studio Strings, Ableton's built-in instruments) are workable starting points when dedicated library budget is unavailable. Stock samples can work with careful mixing, but the limited dynamic layers and single mic position mean more work is required at the humanization and processing stage to achieve a convincing result.

When evaluating any library, load a sustained note in the 2–4kHz harshness zone (approximately A4 to D5 on violin) and listen critically before applying any processing. If you hear harshness immediately, you know the 2–4kHz cut will be essential. If the library already sounds smooth and warm at this pitch, your EQ work will be lighter and you can focus more energy on humanization and arrangement.

Sample Rate and Bit Depth Considerations

Most professional string libraries are available in 44.1kHz/24-bit or 48kHz/24-bit versions, with some offering 96kHz/24-bit options. For mixing purposes, working at 44.1kHz/24-bit or 48kHz/24-bit is entirely sufficient. The frequencies above 22kHz captured in 96kHz recordings of acoustic instruments are largely inaudible and do not meaningfully affect the mixing decisions described in this guide.

More important than sample rate is the bit depth of your project session. Work at 32-bit float internally (which most modern DAWs do by default) to ensure that gain staging operations, EQ calculations, and effect processing happen with maximum precision before any conversion to 24-bit for export. String libraries with high dynamic range particularly benefit from the headroom that 32-bit float processing provides.

For an overview of how monitoring affects your string mix decisions, see our guide to mixing on headphones vs. studio monitors — strings in particular reveal significant differences between monitoring environments because of their wide frequency range and their sensitivity to room acoustics.

Practical Exercises

Frequently Asked Questions