Vocal Compression Guide

Updated May 2026

Vocals live or die by compression. Too little and they disappear in the mix. Too much and they sound strangled, lifeless, and obviously processed. The difference between amateur and professional vocal production almost always comes down to how compression is applied — and in what order.

This guide covers everything: core parameter settings, compressor types and how they behave on voice, serial and parallel compression techniques, genre-specific approaches, plugin recommendations, and the full signal chain used by professional mix engineers. Whether you are compressing a whispered lo-fi vocal or a wall-of-sound pop anthem, the principles are the same — the application changes.

Why Vocal Compression Is Different

The human voice is the most dynamically unpredictable instrument in any mix. A skilled singer can shift 20–30dB between a whispered verse and a belted chorus within a single take. Even within a single phrase, the level difference between consonants and vowels can be 10–15dB. No fader move or static EQ can manage this in real time — only compression can track and respond fast enough to control those rapid shifts.

But the voice is also the most scrutinised element in any production. Listeners are neurologically wired to hear the human voice above everything else. The moment compression becomes audible — pumping, over-squashed consonants, unnatural decay, loss of breath — it breaks the emotional connection with the listener. Vocal compression must be effective and invisible. That is the core challenge, and it is why a single compressor cranked hard almost never gives a professional result.

The professional solution is a chain — automation, then two compressors in series, then potentially a limiter — each stage handling a specific aspect of control so no single stage has to work too hard. When each compressor is only doing 3–5dB of gain reduction, the result sounds natural. When one compressor is doing 12–15dB of gain reduction alone, it sounds like compression.

Understanding this philosophy before touching any parameter is the most important thing you can take from this guide.

The Vocal Compression Signal Chain

Before touching a compressor, understand the order of the full vocal processing chain. Compression does not happen in isolation, and where you place it relative to EQ, saturation, and effects determines what you hear at the end.

Each stage in the chain has a specific and non-overlapping job:

• Gain Staging & Trim: Ensures the signal hits downstream processors at an appropriate level — typically peaking around −18dBFS on the input of a compressor, leaving plenty of headroom.
• Volume Automation (Rides): Manually rides large phrase-to-phrase level differences — the verse being quieter than the chorus, or a particularly shouted ad-lib. By smoothing these out before the compressor sees them, you prevent the compressor from having to work excessively hard, which is the single most important pre-compression step most beginners skip.
• Corrective EQ: A high-pass filter to remove low-end rumble (typically set between 80–120Hz on most vocals, higher on female voices), and any surgical cuts for problem resonances. Cutting problematic frequencies before the compressor prevents those frequencies from triggering excessive gain reduction.
• Compressor 1 — Levelling: The primary levelling compressor. This does the heavy lifting — typically 3–6dB of gain reduction — using a relatively transparent or musically smooth character (optical or clean VCA). Its job is control.
• Compressor 2 — Character: A second compressor for density, glue, and colour — typically 2–4dB of gain reduction. This is where you add harmonic character, the kind of compression you can feel rather than just measure.
• Creative EQ + Effects: Post-compression tonal shaping — adding presence and air, gentle saturation, reverb and delay sends. These come last because you want to shape the sound of the already-controlled signal.
• Limiter / Output: Optional final safety limiter to catch any stray peaks before the signal hits the bus.

For a deeper look at building this kind of layered processing approach, see our guide on how to build a plugin chain — many of the same principles apply across instruments, not just vocals.

Core Compression Parameters — Vocals Specifically

Threshold

Set the threshold so you are achieving 3–6dB of gain reduction on the loudest phrases of the vocal. Watch the gain reduction meter while the vocal plays in context. The needle or bar should be moving — but only significantly on peaks, not constantly pegged.

If the compressor is showing gain reduction on every single syllable at all times, the threshold is too low. The compressor is working on the quiet moments it should be leaving alone. Raise the threshold gradually until quieter phrases trigger 0–2dB GR and only the loudest phrases trigger 4–6dB GR. That is the target operating range.

Ratio

• Sung vocals (transparent levelling): 2:1–4:1
• Pop lead vocals: 3:1–5:1
• R&B and soul: 3:1–4:1 — enough control without killing the expressiveness
• Rap and hip-hop: 4:1–8:1 — the percussive delivery benefits from a firmer clamp

Never start above 6:1 on a vocal without a specific creative reason. High ratios kill the natural breath and expression that make a vocal performance feel human. If you feel the need to push above 6:1 on a single compressor, it is almost always a sign that either the performance needs volume automation first, or you need to split the compression load across two compressors.

Attack

Vocal attack time is one of the most critical — and most commonly misset — parameters in vocal compression. The attack determines when the compressor starts compressing after the signal crosses the threshold.

Too fast (under 3ms) and the compressor catches the initial transient of every syllable. The hard consonants — the "T", "K", "P", and "S" sounds that define intelligibility — are the fastest-moving parts of a vocal. Compress them and the vocal loses clarity. Words start sounding muffled and dull. Presence drops.

The sweet spot for most sung vocals is 10–25ms. This lets the consonant pass through before gain reduction kicks in, preserving intelligibility and the natural attack of the voice. The compressor then clamps down on the body of the vowel — which is where the energy lives — creating density without destroying articulation.

Rap vocals can handle slightly faster attack (5–15ms) because the delivery is more rhythmically driven and the attack transients are often part of the aesthetic, not just the articulation. Very slow attack settings (30ms+) work well when you want maximum punch and presence and the compression is being used more for subtle density than peak control.

Release

Release time controls how quickly the compressor returns to unity gain after compressed audio falls back below the threshold. This is the parameter that most affects how natural or unnatural vocal compression sounds in practice.

Too slow and the compressor stays in gain reduction between syllables — this sucks out the space between words and makes the vocal sound dense, airless, and unnaturally sustained, because the compressor is still holding down the signal even during the natural gaps between syllables. Too fast and you get distortion artifacts or rapid pumping on sustained vowels as the compressor tries to chase individual cycles of audio.

The starting point for most vocals is 80–120ms. This allows the compressor to recover between syllables without being so fast it creates audible pumping. For faster-paced rap or spoken word, 50–80ms tracks the rhythm better. For slow ballads or R&B where phrases are longer, 100–160ms gives a smoother, more natural release.

Many compressors offer an auto-release mode — particularly optical designs and modern plugin compressors like the FabFilter Pro-C 2. Auto-release adapts the release time to the audio content, using faster recovery on short transients and slower recovery on sustained signals. On vocals, this is often a very good default starting point before fine-tuning manually.

Makeup Gain

Every time you compress, you reduce the peak level of the signal. Makeup gain (sometimes called output gain) restores the signal to its original perceived loudness so you can make a fair comparison — compressed versus uncompressed — at the same listening volume. This is important because louder always sounds better to our ears, and without matching levels, you will always think the compressed version sounds better simply because it is louder. Always null-test your compression at matched levels. If it still sounds better compressed — it is genuinely better.

Compressor Types for Vocals

Not all compressors behave the same way. The circuit topology — VCA, optical, FET, tube, or digital — determines the character of the compression and how it interacts with the voice. Understanding these differences is essential for choosing the right tool and for knowing why certain combinations work so well together.

VCA Compressors

VCA (Voltage Controlled Amplifier) compressors are the fastest and most controllable type. They respond accurately to program material and offer precise, repeatable settings. The SSL G-Bus compressor, the dbx 160, and the API 2500 are classic hardware VCA designs.

On vocals, VCA compressors deliver a forward, punchy, controlled sound. They work extremely well on pop and rap vocals where presence and density are priorities. The downside is that with aggressive settings, VCA compression can sound clinical or harsh — which is why many engineers pair a VCA with an optical unit rather than using two VCAs in series.

The UREI 1176 — arguably the most famous vocal compressor in recording history — is technically a FET compressor (Field Effect Transistor), which behaves similarly to a VCA but with a distinctive harmonic character and lightning-fast response. The 1176 has been on more hit vocals than almost any other piece of hardware, used at ratios of 4:1, 8:1, or the famous "all buttons in" mode (which engages all ratio buttons simultaneously for a distinctive, driven sound).

Optical Compressors

Optical compressors use a light element and a photocell to control gain reduction. The key characteristic is that the response is inherently non-linear and program-dependent — the optical element takes time to respond and to recover, and this timing changes based on how loud and sustained the signal is. This produces a compression character that feels musical and natural rather than mechanical.

The Teletronix LA-2A is the defining optical compressor for vocals. Its only controls are gain (threshold) and output — there are no attack or release controls because the optical circuit sets these automatically based on the program. The result is a smooth, levelling compression that most engineers describe as making the vocal "sit" in the mix rather than sounding compressed. The Waves CLA-2A is a widely used software emulation.

Optical compressors excel as the first compressor in a serial chain because they handle the broad levelling task naturally without introducing obvious compression artifacts. A common professional chain is an LA-2A (or emulation) for levelling followed by a 1176 (or emulation) for density and presence.

Tube / Variable-Mu Compressors

Variable-mu (also called vari-mu) compressors use tubes to control gain reduction, where the tube's gain characteristics change with the signal level. The result is an extremely slow, gentle compression with significant harmonic warmth. The Fairchild 670 and the Manley Variable Mu are the defining hardware examples — both are prohibitively expensive for most project studios (the hardware Fairchild 670 regularly sells for tens of thousands of dollars at auction), but software emulations like the UAD Fairchild 670 and Waves Puigtec EQP-1A bring these sounds within reach.

On vocals, variable-mu compression adds a thickness and warmth that no other compressor type quite replicates. It works best as a subtle second compressor adding character rather than as a primary levelling tool, because its slow response makes it poorly suited to catching fast peaks.

Digital / Algorithmic Compressors

Modern digital compressors — such as the FabFilter Pro-C 2, iZotope Neutron's compressor module, and Sonnox Envolution — offer features unavailable in hardware: precise lookahead, linked sidechain analysis, multiple compression styles in a single plugin, and extremely accurate visual feedback of gain reduction. They can emulate hardware characters or operate in transparent "clean" modes.

For producers who want full control and maximum transparency, a digital compressor as the second stage — after an analog-modelled levelling compressor — gives you the best of both worlds: musical levelling from the first stage and precise, controllable density from the second. See our dedicated FabFilter Pro-C 2 review for a deep dive into one of the most widely used digital vocal compressors.

Serial and Parallel Compression Techniques

Serial Compression (Two Compressors in a Row)

Serial compression is the professional standard for lead vocals and the technique that most clearly separates professional results from amateur ones. Instead of asking one compressor to do all the work, you split the compression task across two compressors, each operating gently.

Stage 1 — Levelling compressor: Optical or transparent VCA, ratio 3:1–4:1, slower attack (15–25ms), auto or medium release, threshold set for 3–5dB GR on loud phrases. This stage handles the large dynamic swings — pulling the loud moments down so the overall performance sits at a more consistent level.

Stage 2 — Character compressor: VCA or FET with more personality (1176, SSL channel), ratio 4:1–6:1, faster attack (8–15ms), faster release (60–100ms), threshold set for 2–4dB GR. This stage adds density, a slight forward push in the midrange, and the harmonic character of the compressor circuit.

The combined result — 5–8dB of total gain reduction split across two compressors — sounds substantially more natural than 5–8dB from a single unit, because each compressor's artifacts are minimal and neither is being pushed to its limits. This is the most important practical technique in this entire guide.

Parallel Compression

Parallel compression — sometimes called New York compression — blends a heavily compressed signal with the dry (or lightly processed) original signal. The compressed signal adds density, sustain, and fullness. The dry signal preserves the natural dynamic shape, the initial consonant transients, and the breath of the performance. The blend of the two sounds fuller and more controlled than either signal alone.

On vocals, parallel compression is particularly effective for adding density to a thin-sounding vocal without over-compressing it. The setup: duplicate the vocal track, compress the duplicate heavily (ratio 8:1 or higher, relatively fast attack and release, threshold set for 10–15dB GR), then blend this compressed copy back under the main vocal at a level where it adds body and sustain without obviously pumping. The ratio of dry to compressed determines how much density you add.

Most modern DAWs handle this with a send/return setup or simply by duplicating tracks. Some compressor plugins — including the FabFilter Pro-C 2 and Waves SSL E-Channel — have a built-in mix knob that achieves parallel compression within a single plugin instance, which is more convenient but gives slightly less control than a true parallel routing.

For a broader look at how parallel processing applies across a full mix, our complete vocal mixing guide covers parallel routing alongside reverb, delay, and saturation techniques.

Mid-Side Compression on Vocals

Mid-side compression processes the mid channel (centre of the stereo field) and the side channel (stereo width information) independently. On a lead vocal that sits in the centre of the mix, mid-side compression is less commonly used than on bus processing — but it becomes relevant if you are dealing with a stereo doubled vocal or a vocal recorded with two microphones. By compressing the mid channel more aggressively than the sides, you can tighten the centre image while preserving the natural width of the doubled parts. Plugins like the FabFilter Pro-C 2 and Brainworx bx_digital offer mid-side compression modes.

Genre-Specific Vocal Compression Approaches

The technical principles are universal, but the application varies significantly by genre. What works on a jazz vocal is the wrong approach for a trap vocal. Here is how to adapt the fundamentals to specific genres.

Pop Vocals

Pop vocals need to be loud, present, consistent, and emotionally engaging throughout. The compression approach is typically the most aggressive of any genre, because pop mixes are dense and competitive, and the vocal must cut through without dynamics causing it to disappear in any phrase.

A typical pop vocal chain: volume automation first, reducing the loudest phrases by 3–4dB. Then an optical compressor (CLA-2A or LA-2A emulation) at 3:1–4:1 for smooth levelling — 4–5dB GR. Then a VCA (SSL channel or 1176 at 4:1) for density and forwardness — 3–4dB GR. Total compression: 7–9dB across two stages. Creative EQ adding presence (2–4dB around 8–12kHz) and air after the compressors.

Rap and Hip-Hop Vocals

Rap vocals are percussive, rhythmically precise, and often designed to sit on top of or alongside heavy 808 bass and kick drum. Compression on rap vocals is both a technical and aesthetic decision — heavily compressed rap vocals have their own sonic identity that is part of the genre's sound.

Higher ratios (4:1–8:1) are appropriate because the delivery is more dynamic and rhythmically varied than sung vocals. Faster attack times (5–15ms) help catch the transient peaks of aggressive consonant delivery. Release times that track the rhythm of the rap — typically 80–150ms — let the compressor breathe in time with the performance.

De-essing is often more critical on rap vocals than sung vocals because of the density of sibilant sounds in speech. A de-esser placed after the compressors — or in a sidechain — is typically essential. For more on compressing rap-specific elements, see our guide on advanced vocal mixing techniques.

R&B and Soul

R&B and soul vocals are emotionally expressive — melismatic runs, dynamic whispers, full-power belt notes. The challenge is controlling the wild dynamic range without killing the expressiveness that makes these performances powerful. Over-compression kills soul vocals; the nuance of a melismatic run disappearing into a wall of gain reduction destroys the very thing that makes the performance special.

Optical compression is the natural first choice — the program-dependent response of an LA-2A or similar moves with the vocal rather than mechanically clamping down on it. Ratios of 3:1–4:1 with slower attack times (15–25ms) preserve expression. Heavy volume automation before the compressor is arguably more important here than in any other genre — riding individual phrases to a consistent range before the compressor sees them allows much gentler compression settings.

Country and Folk

Country and folk vocals prioritise naturalness, warmth, and intelligibility. Listeners in these genres are highly attuned to the human qualities of the voice — breath, room sound, the slight roughness of a genuinely felt performance. Over-processing is audible and unwanted.

Gentle ratios (2:1–4:1) with slow attack times (15–30ms) preserve the natural attack of the voice. Optical or tube compression adds warmth without the clinical precision of a hard VCA. Total gain reduction of 3–5dB is usually sufficient when combined with careful volume automation. Many country engineers use a single optical compressor rather than a serial chain, prioritising naturalness over density.

Rock and Alternative

Rock vocals often need to compete with loud guitar amps, heavy drums, and dense arrangements. Compression here is often more aggressive — higher ratios, faster attacks, a more obviously compressed sound that has become part of the genre's aesthetic identity. The 1176 FET compressor is the archetypal rock vocal compressor: fast, punchy, and capable of adding an aggressive forward energy when pushed hard.

At higher ratios (4:1–8:1) with the 1176's faster attack settings, rock vocals get a compressed, in-your-face quality that cuts through loud arrangements. The harmonic saturation of the FET circuit also adds a slight edge that complements distorted guitar tones without clashing.

Plugin Recommendations for Vocal Compression

The plugin market for compressors is enormous. These are the most widely used, most trusted, and most sonically appropriate options for vocal compression across different use cases and budgets.

Waves CLA-2A

The CLA-2A is a software emulation of the Teletronix LA-2A optical compressor, developed by Chris Lord-Alge, one of the most commercially successful mix engineers working today. It captures the smooth, program-dependent optical compression character of the hardware original. On vocals, it excels as a first-stage levelling compressor — set the peak reduction to achieve a moderate amount of compression and let the optical circuit do its natural work. Very few controls means very little to get wrong. The Waves CLA-2A is part of the Waves CLA Classic Compressors bundle — current pricing available on the Waves website.

Universal Audio 1176 Classic Limiter Collection

The UAD 1176 emulations (available as UAD DSP plugins and as UAD LUNA plugins) are some of the most accurate software recreations of the hardware 1176 FET compressor. The collection includes multiple revisions of the hardware — the Rev A through Rev E — each with slightly different tonal character. On vocals, the 1176 adds presence, density, and a slight harmonic edge that translates well in dense mixes. It works extremely well as the second compressor in a serial chain — the levelling compressor first, then the 1176 for character.

FabFilter Pro-C 2

The FabFilter Pro-C 2 is arguably the most versatile compressor plugin available. It offers eight different compression styles (Clean, Classic, Opto, Vocal, Mastering, Bus, Punch, and Safe), a built-in sidechain EQ, a mix knob for instant parallel compression, and a detailed gain reduction display. The "Vocal" mode is specifically designed for the program-dependent, smooth compression that works on voice — it is an excellent choice for producers who want a single high-quality plugin that covers all use cases. The Pro-C 2 is particularly strong as a second-stage character compressor after an optical levelling unit. See our full FabFilter Pro-C 2 review for detailed analysis of each mode.

Waves SSL E-Channel

The SSL E-Channel plugin emulates the channel strip of the SSL 4000E console — including its fast VCA compressor/gate section. The E-Channel compressor is clean, punchy, and fast — very similar in character to the hardware SSL channel strip that has shaped the sound of countless pop and rock records since the 1980s. On vocals, it adds a forward, present compression quality with the classic SSL tonal signature. It also includes an EQ section, making it a convenient single-plugin option for producers who want to use a single console strip for both compression and tonal shaping on a vocal channel.

UAD Neve 33609

The UAD Neve 33609 emulates the hardware Neve 33609 stereo bus compressor — one of the warmest, most musical compressors ever built. On a lead vocal, it adds a thick, glue-like quality that is slightly different from the precision of a VCA or the smoothness of an optical. It is best used as a subtle second stage adding density and Neve warmth — ratio 4:1, moderate attack, 2–3dB of GR. At higher gain reduction amounts it becomes obviously coloured, which is either a problem or a creative tool depending on the context.

iZotope Neutron — Compressor Module

The iZotope Neutron plugin suite includes a compressor module with vintage and modern modes, transient shaping, and the iZotope machine-learning Assist features that can suggest starting settings based on the audio content. For producers who are still developing their compression instincts, the Neutron Assist feature is a useful starting point — it analyses the vocal and proposes threshold, ratio, and attack/release values that are at minimum a musically intelligent starting point. The iZotope Neutron guide on this site covers the full plugin workflow in detail.

Budget and Free Options

• Digitalfishphones Dominion: A free Windows VST compressor with a warm, musical character that punches above its zero price point. Useful for home studio producers on tight budgets.
• Variety of Sound FerricTDS: Free saturation and compression tool with a vintage tape-machine character. Adding mild compression and harmonic saturation simultaneously — excellent for lo-fi and vintage vocal sounds.
• Native DAW compressors: Ableton Live's Glue Compressor (an SSL G-Bus emulation), Logic Pro's Vintage VCA and Vintage Optical, and FL Studio's Fruity Peak Controller and Parametric EQ 2 compression mode all offer surprisingly capable vocal compression without additional cost. For producers starting out, the built-in DAW compressor is a better learning tool than a plugin purchased before the fundamentals are understood.

For a broader look at the compressor plugin landscape across all use cases, our best compressor plugins roundup covers hardware emulations and digital designs for every budget.

Advanced Techniques and Common Mistakes

Sidechain EQ for Vocal Compression

Most modern compressors — hardware and plugin — allow you to insert an EQ into the sidechain, meaning the compressor reacts to a frequency-filtered version of the signal rather than the full-band signal. On vocals, this is an extremely useful technique because certain frequencies — particularly the upper bass and low-midrange buildup around 100–300Hz — can cause the compressor to over-react and trigger excessive gain reduction even when the vocal does not perceptually need it.

A high-pass filter in the sidechain at 100–150Hz means the compressor ignores the low-frequency content of the vocal and responds primarily to the presence and midrange energy — where the perceptual loudness of the voice actually lives. This produces a more natural-sounding compression with fewer unwanted pumping artifacts.

Similarly, a gentle de-emphasis of the 6–10kHz region in the sidechain prevents sibilant sounds from triggering excessive compression. This is a more subtle alternative to a dedicated de-esser, and many engineers use sidechain EQ and a de-esser together in the chain.

Compression and EQ Order

The question of whether EQ should come before or after compression is one of the most debated topics in mixing, and the professional answer is: both, at different stages, for different purposes.

EQ before compression (corrective EQ): Cutting problematic frequencies before the compressor changes what the compressor responds to. If a vocal has a harsh resonance at 3kHz, cutting it before the compressor means the compressor will not over-react every time that frequency peaks. This produces more natural, even compression. The high-pass filter always goes before the compressor for this reason.

EQ after compression (creative EQ): Shaping the tonal balance of the already-compressed vocal. Adding presence at 3–5kHz, adding air at 10–16kHz, cutting boxiness at 300–500Hz. Because you are working on a signal whose level has already been controlled, the EQ changes are more consistent and predictable. For a full breakdown of vocal EQ technique, see our dedicated guide on how to EQ vocals.

The standard professional approach is corrective EQ before compression, creative EQ after compression — using both for their respective strengths.

Automation Before Compression — The Most Underrated Step

Volume automation before the compressor is not just a nice-to-have — it is arguably more important than any compressor setting for achieving a polished, natural vocal. The principle is simple: manually ride the fader of the raw vocal to smooth out the largest phrase-to-phrase level differences before the signal hits the first compressor.

If the verse is consistently 8dB quieter than the chorus, and you let the compressor try to handle that difference, the compressor will be working very differently on the verse versus the chorus — which means the compression character changes between sections. By riding that 8dB difference with automation first, the compressor sees a much more consistent signal in both sections and can work uniformly throughout the song.

This is the technique that makes the compression sound "invisible" — the compressor is only managing micro-dynamics within phrases, not desperately trying to bridge large structural level gaps. Most professional mix engineers spend significant time on vocal automation before touching a compressor setting, and it is the step that most beginners skip entirely.

For a practical guide to DAW automation, including how to ride vocals efficiently, our guide on how to use automation in your DAW covers the full workflow in detail.

Common Mistakes to Avoid

Attack too fast: Setting the attack under 5ms on a vocal compressor is the most common mistake in vocal compression. It kills intelligibility by catching consonants before the compressor should be engaging. Start at 15ms and move from there.

Release too slow: A release time of 300ms or longer on vocals almost always causes inter-syllable pumping — the compressor is still in gain reduction during the space between words, creating an unnatural tightening of the performance. If you hear the vocal "breathing" or the track getting quieter during pauses, the release is too slow.

Threshold too low: If the gain reduction meter is constantly pinned, the compressor is always working — meaning it is effectively just turning the signal down by a fixed amount, which is what a fader does, not what a compressor should do. Raise the threshold until the compressor is only engaging on the loud moments.

Not using makeup gain for comparison: Without matching levels between compressed and uncompressed signals, you will always prefer the compressed version simply because it is louder. This leads to over-compression because you keep thinking more compression sounds better — when actually it just sounds louder.

Ignoring volume automation: Using compression to handle all level management — including large structural differences between song sections — forces the compressor to work in very different conditions through the song. Automation before compression is the fix.

Soloing the vocal when setting compression: Vocal compression should be set in context — with the full mix playing. A compression setting that sounds perfect on a soloed vocal may be completely wrong in the mix, and vice versa. Always check compression decisions in context.

Over-de-essing: De-essing is a form of frequency-selective compression that targets sibilant sounds ("S", "SH", "T" sounds). Over-de-essing produces a vocal where sibilants are completely removed — creating a lispy, hollow quality where the "S" sounds are missing. De-ess to reduce harshness, not to eliminate sibilance entirely.

Practical Exercises

Frequently Asked Questions