Key takeaways
- 4–6 dB-A: Essentially inaudible in any real room. Found on the quietest modern condensers.
- 7–9 dB-A: Inaudible in a room above 30 dBA. Fine for almost everyone.
- 10–14 dB-A: Audible as a faint hiss only if you apply heavy compression and your room is genuinely silent. Still fine for most work.
- Not specified: Common for dynamics. In practice, a dynamic’s electrical noise is so far below any real room that the question is moot.
The best microphone for voice over is the one that matches your room before it matches your budget: in a quiet, treated space — room tone below roughly 30 dBA — a large-diaphragm condenser such as the Neumann TLM 103, Rode NT1 or Lewitt LCT 440 Pure gives you the detail, air and low self-noise that voice-over clients expect, while in an untreated bedroom, apartment or shared flat a large-diaphragm dynamic such as the Shure SM7B, Electro-Voice RE20 or Heil PR-40 will out-record a $3,000 condenser in the same room, because its lower sensitivity and tighter pattern simply hear less of the room. Everything below is organised around that first decision, then around the second one almost nobody talks about honestly: your voice.
Quick picks by situation
If you want the short version, find your row. The “why” column is the part that matters, because two people with the same budget in the same city can need opposite microphones.
| Your situation | Pick the type | Real examples | Why it fits | Typical street price |
|---|---|---|---|---|
| Untreated bedroom, roughly 10 × 12 ft, carpet, HVAC hum | Large-diaphragm dynamic, cardioid | Shure SM7B, Rode PodMic, Electro-Voice RE20 | Low sensitivity rejects room reflections; tight cardioid pattern keeps the fridge and the hallway out | $100–$450 |
| Treated booth or quiet room, room tone under 30 dBA | Large-diaphragm condenser | Neumann TLM 103, Rode NT1 5th gen, Lewitt LCT 440 Pure | Detail and transient accuracy; self-noise low enough to raise gain without hiss | $260–$1,400 |
| Deep, boomy, chesty voice | Dynamic with a low-cut switch | Shure SM7B, Electro-Voice RE20 | Onboard bass rolloff and proximity control keep the low end from turning to mud | $400–$450 |
| Bright or sibilant voice | Warm condenser or smooth dynamic | Warm Audio WA-87 R2, Aston Origin, Rode Procaster | Gentler top-end lift means less de-essing in post | $230–$700 |
| Audiobooks and long-form narration | Low-self-noise condenser, or RE20 | Neumann TLM 103, Rode NT1 5th gen, Electro-Voice RE20 | Consistency over eight-hour sessions matters more than sparkle | $260–$1,400 |
| Podcast plus occasional VO, one person, one desk | Hybrid XLR/USB dynamic | Shure MV7+ | USB today, XLR when you buy an interface; same capsule either way | $250–$300 |
| No interface, no interest in buying one | USB condenser | Rode NT-USB+, Elgato Wave:3, Audio-Technica AT2020USB+ | Plug-and-play, on-mic monitoring, zero gain-staging homework | $130–$200 |
| Commercial VO, agent-level and broadcast work | Multi-pattern studio condenser | Neumann U87 Ai, AKG C414 XLS | Pattern flexibility, client familiarity, resale value | $1,100–$4,000 |
Why voice over is not the same job as podcasting or singing
Three jobs, three different microphones, and the difference is not marketing. Singing rewards a microphone that captures a wide dynamic range, tolerates high SPL and flatters a sustained note. Podcasting rewards a microphone that survives being knocked around, sits close to a moving head and forgives a bad room. Voice over sits in a narrow band between them: you are recording a single voice at a consistent distance, at conversational level, for minutes to hours at a time, and the deliverable is judged on intelligibility, low noise and neutrality.
That combination pushes the requirements in a specific direction. You want a microphone that is quiet enough that the noise floor is your room, not your gear. You want a frequency response that is either flat or gently shaped, not scooped in the midrange — the 1 kHz to 4 kHz band is where consonants live, and a mic that dips there will make you sound like you are talking through a blanket. You want a pattern that stays consistent as you move slightly, because you will move slightly. And you want a body that does not pick up the vibration of your desk every time you hit a key.
The rest of this article deals with the two variables that decide all of that: pickup type and noise floor. Everything else — the badge, the price, the colour — is downstream.
Condenser versus dynamic: what actually changes in your recording
Every voice-over microphone you will be offered is either a moving-coil dynamic or a condenser. The distinction is not “pro versus amateur.” It is a set of physical trade-offs that show up in your recording as level, detail and room pickup.
A condenser capsule is a thin, charged membrane that moves very little and generates a relatively large voltage for a given sound pressure, which is why condensers are described as sensitive. That sensitivity is the whole story. A typical large-diaphragm voice-over condenser puts out something in the region of 20–30 mV/Pa. A typical large-diaphragm broadcast dynamic puts out 1–3 mV/Pa. That is a difference of roughly 20 dB — a factor of ten in voltage — for the same sound arriving at the grille.
Ten times more output for the same input means the condenser also outputs ten times more of everything else that reaches the capsule: the reflection off your untreated back wall, the hum of the refrigerator, the neighbour’s television through the floor. A dynamic is not more “focused” in some mystical sense. It is simply less sensitive to everything, and the room is part of everything.
The second variable is mass. A condenser diaphragm is measured in microns; a dynamic diaphragm has a coil of wire attached and is orders of magnitude heavier. Heavy things do not start and stop as quickly. This is why condensers capture the fine detail of a breath, the click at the start of a word, the tiny changes in lip and tongue position that make a read sound present, and why dynamics sound slightly rounder and more forgiving. For narration, that extra detail is usually what you are paid for. For a room full of reflections, that extra detail is a liability.
The trade-off, in numbers
| Characteristic | Large-diaphragm condenser | Large-diaphragm dynamic (broadcast type) | What it means in a home studio |
|---|---|---|---|
| Sensitivity | About 20–30 mV/Pa | About 1–3 mV/Pa | Condenser hears the room roughly 20 dB louder |
| Self-noise (published) | 4–14 dB-A | Not usually specified; effectively negligible | Matters only if your room is already quiet |
| Preamplifier gain needed | 20–40 dB | 55–70 dB | Dynamic mics demand a genuinely quiet preamp |
| Transient detail | High | Moderate | Condenser flatters precision; dynamic flatters imperfection |
| Off-axis rejection | Varies; multi-pattern models leak from the rear | Usually strong cardioid or hypercardioid | Dynamic tolerates a desk, a laptop and a wall behind you |
| Proximity effect | Strong, and often used deliberately | Strong but tamed by internal design in models like the RE20 | Dynamic boom is easier to control with a low-cut switch |
| Max SPL | 130–160 dB | 130–150 dB | Irrelevant for speech unless you shout professionally |
| Shock and handling noise | Sensitive; needs a proper shock mount | Robust; tolerates a plain clip | Dynamic is easier on a cheap boom arm |
| Typical price band | $150–$4,000 | $100–$500 | The price gap is real, but the room gap is bigger |
Read that table as a decision, not a ranking. If your room is quiet, the condenser’s extra sensitivity costs you nothing and buys you detail. If your room is not quiet, the dynamic’s insensitivity is not a compromise — it is the feature you are paying for.
Noise floor: the number that quietly decides your microphone
Almost nobody buying their first voice-over mic measures their room. This is the single most common mistake in home voice over, and it is also the cheapest to avoid, because a $70 sound level meter app on a phone and ten minutes of listening will tell you more than any spec sheet.
Do this before you buy anything. Close the windows, turn off the HVAC if you can, sit exactly where the microphone will go, and measure the room in dBA with nothing happening. Then do it again with the HVAC on, because you will not turn it off for a two-hour session in July. Write both numbers down.
What room tone numbers sound like in practice
| Measured room tone | What that environment usually is | Microphone that fits |
|---|---|---|
| 20–25 dBA | Purpose-built booth, isolated, no mechanical noise | Any low-self-noise condenser; the room is not your limit |
| 26–32 dBA | Quiet suburban bedroom at night, HVAC off, hard drive silent | Condenser works; you may still need a noise gate at −55 dB |
| 33–38 dBA | Typical bedroom, laptop fan audible, fridge on the far side of a wall | Condenser usable with treatment; dynamic is safer |
| 39–45 dBA | Apartment with HVAC running, or a window onto a residential street | Dynamic, close-miked, high-pass filtered |
| 46 dBA and up | Home office near a kitchen, daytime traffic, shared flat | Dynamic only, and consider recording at 5 a.m. |
Here is the part that surprises people. A quiet condenser is not the problem in a noisy room — the room is. A microphone with 7 dB-A of self-noise in a room measuring 42 dBA contributes essentially nothing to the noise you hear. The room is 35 dB louder than the microphone’s own noise. Buying a quieter condenser will not help you. Buying a dynamic will, because the dynamic’s lower sensitivity means you turn the gain up less relative to the voice and the room comes up with it in exactly the same proportion — no. That is not right either, and it is worth being precise, because this is where most buying advice goes wrong.
Turning the gain up on a dynamic does raise the room along with the voice, in the same ratio. What changes is the ratio of direct voice to reflected room sound arriving at the capsule in the first place. A dynamic microphone is typically used much closer to the mouth — 2 to 4 inches rather than 8 to 12 — and closeness is what improves the direct-to-reverberant ratio. A cardioid dynamic with a tight pattern also picks up less from the sides and rear. The advantage is acoustic, not electronic. This is why “just turn up the gain on a condenser” does not reproduce what a dynamic does in a bad room.
Self-noise, and when it actually matters
Self-noise, also written as equivalent noise level, is the noise the microphone’s own electronics add. It is measured in dB-A, and lower is better. Published figures for the microphones discussed here range from about 4 dB-A to about 14 dB-A.
- 4–6 dB-A: Essentially inaudible in any real room. Found on the quietest modern condensers.
- 7–9 dB-A: Inaudible in a room above 30 dBA. Fine for almost everyone.
- 10–14 dB-A: Audible as a faint hiss only if you apply heavy compression and your room is genuinely silent. Still fine for most work.
- Not specified: Common for dynamics. In practice, a dynamic’s electrical noise is so far below any real room that the question is moot.
Self-noise only becomes the deciding factor when you are choosing between two condensers for a quiet booth, or when you plan to record quiet passages — meditation narration, ASMR, whispered lines in an audiobook — and then compress them heavily. If you are recording a normal conversational read in a normal room, self-noise below about 10 dB-A is not your bottleneck. Your room is.
Room sensitivity: why a 10 × 12 ft bedroom is the hardest room in the house
A 10 × 12 ft room with an 8 ft ceiling has a volume of about 960 cubic feet. Rooms of that size have a well-known problem: their strongest acoustic modes cluster in the 100–300 Hz range, which is exactly where the fundamental frequencies of a male voice and the lower harmonics of a female voice sit. The result is a low-mid bloom that makes every recording sound slightly boxy and slightly indistinct, no matter how good the microphone is.
Meanwhile, the distance from your mouth to the nearest wall is short — often 3 to 5 feet — which means early reflections come back fast and loud. In a room that size, a condenser placed 12 inches from your mouth is picking up a direct signal and a reflected signal that are only a few milliseconds apart. That is comb filtering, and it is why untreated small rooms sound “phasey” and “boxy” at the same time.
You have three ways out, and they are not mutually exclusive:
- Change the distance. Moving from 12 inches to 3 inches increases the direct sound by roughly 12 dB while the reflection from a wall 4 feet away increases by far less. This is the single highest-value move available to you, and it is free.
- Change the microphone type. A dynamic at 3 inches with a cardioid pattern rejects far more of the rear and sides than a condenser at 12 inches. This is the second-highest-value move, and it costs money.
- Change the room. Treatment, a reflection filter behind the microphone, a closet full of clothes, a blanket fort. Effective, but the amount of absorption needed to tame a 100–300 Hz mode in a 960 cubic foot room is much larger than most people expect. Thin foam panels handle 2 kHz and up and do almost nothing for the low-mid bloom.
Room-size fit, by room
| Room | Approximate volume | Main acoustic problem | Practical approach |
|---|---|---|---|
| Closet, 3 × 4 ft | About 100 cu ft | Extremely boxy, all modes above 200 Hz | Works with a dynamic close-miked, heavy soft furnishing, and a low-cut filter |
| Small bedroom, 10 × 12 ft | About 960 cu ft | Low-mid buildup at 100–300 Hz; early reflections | Dynamic at 3–4 in, or condenser at 8 in with a reflection filter behind the mic |
| Living room, 14 × 18 ft | About 2,000 cu ft | Longer reverb tail, harder to treat cheaply | Record in a corner-free position, off-centre, with a rug and soft furniture |
| Basement or garage | Variable | Hard parallel surfaces, flutter echo | Broadband absorption first; a dynamic until then |
| Purpose-built booth, 4 × 6 ft | About 200 cu ft | Over-damped, unnaturally dead | Condenser; you may need to add a little liveliness back in post |
Notice that the recommendation changes with the room, not with the budget. A $1,400 condenser in a 10 × 12 ft untreated bedroom will produce a worse voice-over file than a $100 dynamic in the same spot, and it will not be close.
Gain, preamps and the hidden cost of a dynamic microphone
Dynamics solve your room problem and create a different one: they need a lot of clean gain. A Shure SM7B is specified at roughly −59 dBV/Pa, which means a typical audio interface with 55–60 dB of available gain will need to run near its maximum to get a healthy recording level. Running a preamp at maximum gain is where you hear its noise floor, its interference and, on bus-powered USB interfaces, sometimes its limits.
The practical solutions, in order of cost:
- Buy an interface with more gain. Some current interfaces publish 65–75 dB of gain on their microphone inputs. This is the cleanest solution because it adds nothing to the signal path.
- Use an inline preamp. Devices such as the Cloud Microphones Cloudlifter CL-1, Triton Audio FetHead and sE Electronics DM1 Dynamite add roughly 20–27 dB of clean, phantom-powered gain before the interface. They are small, they work, and they cost roughly $50–$200.
- Accept a lower recording level. A 24-bit recording at −18 dBFS peaks has plenty of resolution. If your interface is quiet enough, recording lower and normalising in post is perfectly acceptable and costs nothing.
- Use a channel strip. A dedicated preamp or channel strip gives you gain, EQ and compression in one box. Useful, but it is a bigger investment than the microphone in many home setups.
Condensers do not have this problem. At 20–30 mV/Pa they reach a healthy level on almost any interface, which is one reason they are the default recommendation for beginners — and one reason that recommendation fails in untreated rooms.
Stands, boom arms, shock mounts and the clearance nobody mentions
This is the part of the buying decision that product listings skip entirely, and it is the part that determines whether you actually enjoy recording.
- Weight. A Shure SM7B is roughly 1.7 lb (about 765 g). An Electro-Voice RE20 is roughly 1.6 lb (about 737 g). A Neumann U87 Ai is about 1.1 lb (500 g) plus a shock mount. Cheap desktop boom arms are frequently rated for 2–3 lb total, and that rating includes the shock mount or yoke. Add a 1.7 lb microphone to a 0.5 lb shock mount and a 0.3 lb pop filter and you are at 2.5 lb on a 3 lb arm, which will sag within a month.
- Desk clearance. A microphone 3 inches from your mouth needs to sit roughly 30–36 inches above the floor if you are seated at a standard 29–30 inch desk in a standard chair. A boom arm that clamps to the desk edge and reaches 20 inches horizontally will put the mic in front of your keyboard, not your mouth. Measure your desk depth before you buy.
- Doorway and closet reality. If you plan to record in a closet, a 30-inch doorway and a 24-inch-wide closet interior leave you about 22 inches of usable width once the door frame is accounted for. A boom arm plus a microphone stand plus a stool does not fit in most closets. Plan on a floor stand and a lot of soft furnishing, or a different room.
- Shock mounts. A condenser needs one; a dynamic usually does not. Elastic shock mounts degrade — the bands stretch and eventually snap, typically after two to four years of use. Buy one with replaceable bands.
- Pop filters. A nylon mesh pop filter should sit 2–3 inches in front of the grille, not touching it. A metal mesh pop filter can sit closer. Both add weight to the arm.
- Reflection filters. Products such as the Aston Halo, sE Electronics Reflexion Filter and Primacoustic VoxGuard mount behind the microphone and absorb what comes off the wall in front of you. They help most in the 500 Hz–4 kHz range and least in the 100–300 Hz range where small rooms actually hurt.
The roundup: dynamic microphones for voice over
These are the moving-coil dynamics that dominate voice-over work in untreated rooms. All of them are cardioid or hypercardioid, all of them are built to survive being knocked over, and all of them will reward you for getting close.
Shure SM7B
The SM7B is the reference point for home voice over, and it has been for years. It is a large-diaphragm dynamic with a cardioid pattern, a published frequency range of 50 Hz to 20 kHz, a sensitivity of roughly −59 dBV/Pa, and an internal shock isolation system that makes it unusually immune to desk thumps. It has two switches: a bass rolloff and a presence boost, and the bass rolloff is the one you will use most, because at 3 inches the proximity effect is significant.
Why it works in a bad room: it is insensitive, it is cardioid, and it is designed to be used at 2–4 inches. Why people complain about it: it needs 60 dB of clean gain, which many budget interfaces do not have, and its low-end response is generous enough that a boomy voice can sound congested without the rolloff engaged. If your interface cannot get you to a healthy level without audible hiss, budget for an inline preamp alongside the microphone. Typical street price: $350–$450.
Electro-Voice RE20
The RE20 is the other long-standing broadcast standard, and it solves the proximity problem in a way the SM7B does not. Its Variable-D design uses slots along the body of the microphone to reduce the bass buildup that normally occurs as you get closer, which means you can work at 2 inches without the low end ballooning. It is also internally shock-mounted and hum-bucking, which makes it unusually good near computer monitors and power supplies.
For audiobook narration and long-form work, the RE20’s consistency is its selling point: the tonal balance barely changes as you move a couple of inches, which means fewer retakes and less corrective EQ. It is heavy — around 1.6 lb — and it needs a serious boom arm or a floor stand. It also needs plenty of gain, in the same range as the SM7B. Typical street price: $400–$500.
Heil PR-40
The PR-40 is a large-diaphragm dynamic with a neodymium magnet, a cardioid pattern and a published range that extends lower than most dynamics — down to around 28 Hz. The practical effect is a fuller low end and a slightly more open top than the SM7B, which suits voices that sound thin or light on a darker dynamic. It is also more sensitive than the SM7B, so it needs less gain, which makes it a good match for interfaces that struggle with the Shure.
The trade-off is that its fuller bottom end is less forgiving in a small untreated room, where the low-mid buildup is already the problem. If you have a 10 × 12 ft bedroom and a naturally warm voice, the PR-40 will need more high-pass filtering than the SM7B. If you have a lighter voice and a decent interface, it can be the better fit. Typical street price: $300–$400.
Rode PodMic
The PodMic is the budget entry into the same category: a cardioid dynamic with an internal pop filter, a built-in swing mount, and a published response of roughly 20 Hz to 20 kHz. It is noticeably cheaper than the broadcast standards, and it sounds like a slightly brighter, slightly less refined version of them. For a first microphone, a home podcast that occasionally does voice-over work, or a second mic for interviews, it is a reasonable buy.
The internal pop filter works, but it also means the grille is close to the capsule, so plosives still need care. It weighs around 2 lb, which is heavier than it looks, and it will sag a light boom arm. Typical street price: $100–$140.
Rode Procaster
The Procaster is Rode’s step-up broadcast dynamic, with a tighter low-end response than the PodMic, an internal pop filter and a body designed for close work. It sits between the PodMic and the SM7B in both price and performance, and its tonal balance is a little warmer and less bright than the PodMic’s, which suits narration better than podcast chatter.
It is not as insensitive as the SM7B, so it needs less gain, and it is not as forgiving of a very close position — 3–4 inches is the sweet spot. Typical street price: $220–$280.
Audio-Technica BP40
The BP40 is a large-diaphragm dynamic with a hypercardioid pattern, which makes it more directional than the cardioid models above. That extra directionality is genuinely useful in a room with a specific noise source to one side — a computer, a window, a door — because hypercardioid rejects more from the sides than cardioid does. The trade-off is a rear lobe: hypercardioid patterns pick up more from directly behind the microphone than cardioid does, so if there is a wall 6 inches behind your mic, hypercardioid is the wrong choice.
Its published response runs roughly 50 Hz to 16 kHz, with a presence lift that suits speech. Typical street price: $300–$400.
Also worth knowing: Beyerdynamic M 88 and Shure SM58
The Beyerdynamic M 88 is a hypercardioid dynamic with a long history in broadcast and voice work; it has a distinctive, slightly aggressive upper-mid presence that cuts through on commercial reads but can sound hard on long narration. The Shure SM58 is not a voice-over microphone in any serious sense — it is a stage vocal mic with a presence peak and a very forgiving pattern — but it is worth mentioning because thousands of people already own one. If you have an SM58 and no budget, use it at 2 inches with a pop filter and a high-pass filter. It will outperform a cheap condenser in a bad room, and it costs you nothing.
The roundup: condenser microphones for voice over
These are the condensers that make sense once your room is quiet enough to justify them. The order below runs roughly from the most affordable practical option upward, with a note on what each one is actually good at.
Rode NT1 5th generation
The current NT1 is one of the quietest affordable condensers available, with published self-noise around 4 dB-A and a maximum SPL around 142 dB. It has a cardioid pattern, a large gold-sputtered capsule, and a version that offers both XLR and USB output with onboard 32-bit float conversion on the USB side.
Why it matters for voice over: 4 dB-A of self-noise means you can compress heavily without dragging up hiss, which is exactly what audiobook and e-learning work demands. It is also relatively neutral, with a gentle presence lift rather than a bright peak, so it flatters most voices without needing much EQ. The catch is that at roughly 25 mV/Pa it will hear your room in detail. Use it in a treated space or at 6–8 inches with absorption behind the microphone. Typical street price: $250–$320.
Lewitt LCT 440 Pure
The LCT 440 Pure is a fixed-cardioid large-diaphragm condenser with published self-noise around 7 dB-A and a maximum SPL of about 140 dB. It has a slightly brighter, more open top end than the NT1 and a very low handling-noise design, which makes it pleasant for narration and commercial reads where a little air helps.
Its weakness is the same as every condenser’s: it is sensitive. In a room measuring 40 dBA, the LCT 440 Pure will record the room along with you. In a treated corner of a quiet room, it is one of the better values in the category. Typical street price: $280–$350.
Audio-Technica AT4040
The AT4040 is a workhorse large-diaphragm cardioid condenser with published self-noise around 12 dB-A, sensitivity around 25 mV/Pa and a maximum SPL around 145 dB. It has a transformerless output stage and a reputation for a smooth, slightly warm midrange that flatters voices without exaggerating sibilance.
Twelve dB-A of self-noise is fine for normal conversational recording and only becomes visible if you are recording whispered or very quiet material and compressing it hard. The AT4040 is a sensible choice for a home studio doing mixed work — voice over, sung vocals, acoustic instruments — where one microphone has to cover several jobs. Typical street price: $300–$380.
Audio-Technica AT2035
The AT2035 is the budget option in the same family: cardioid, large diaphragm, published self-noise in the mid-teens, and a slightly brighter top end than the AT4040. It is often sold as a starter condenser, and it is a legitimate one — but it is also the microphone most likely to disappoint a beginner, because it is sensitive enough to expose an untreated room immediately.
Buy it if you have a quiet space and a small budget. Do not buy it expecting it to fix a noisy room. Typical street price: $130–$170.
Neumann TLM 102
The TLM 102 is Neumann’s entry-level large-diaphragm condenser, cardioid, with published self-noise around 12 dB-A and a sensitivity around 11 mV/Pa. Its frequency response has a deliberate lift in the upper midrange that makes voices sit forward in a mix, which is why it is popular for spoken word and podcast work.
It is a small-bodied microphone with no pad or low-cut switch, so you rely on your preamp for both. That lift, combined with its sensitivity, means it needs a quiet room and a voice that can take a little brightness. Typical street price: $700–$850.
Neumann TLM 103
The TLM 103 is the most common “serious” condenser in home voice-over studios, and for good reason: published self-noise of about 7 dB-A, sensitivity around 23 mV/Pa, a cardioid pattern, a maximum SPL around 138 dB, and a tonal signature that is detailed without being harsh. It uses a transformerless circuit derived from Neumann’s larger microphones.
For audiobook narration, e-learning and commercial work in a properly treated space, it is close to a default answer. In an untreated 10 × 12 ft bedroom it will sound worse than a $100 dynamic, and no amount of EQ will fully fix the room it captures. Typical street price: $1,200–$1,500.
AKG C414 XLS
The C414 XLS is a multi-pattern large-diaphragm condenser with nine selectable polar patterns, published self-noise around 6 dB-A and a maximum SPL around 158 dB. The pattern flexibility is the reason to buy it: cardioid for solo narration, figure-8 for two-person work or for a mid-side recording, omni when you want the room, hypercardioid when you want less of it.
It also has switchable pads and low-cut filters, which makes it unusually adaptable. It is a heavier investment, and its multiple patterns mean you can accidentally record in the wrong one — check the switch before every session. Typical street price: $1,100–$1,400.
Warm Audio WA-87 R2
The WA-87 R2 is a large-diaphragm condenser with three patterns and a circuit inspired by the classic FET studio microphone. Its tonal character is warmer and rounder than the modern Neumanns, with a softer top end, which suits bright or sibilant voices that make most condensers sound edgy.
Published self-noise is in the low teens, which is fine for normal speech but not the quietest option if you are recording very quiet material. If your voice tends bright and you want a condenser that tames it, this is one of the more cost-effective routes. Typical street price: $600–$750.
Aston Origin
The Origin is a cardioid large-diaphragm condenser with a built-in shock mount and a distinctive slotted body. It has a slightly warmer, more forgiving character than the bright “budget condenser” stereotype, and its integrated mount reduces the amount of handling noise that reaches the capsule — useful on a desk-mounted boom arm.
It is a solid single-mic choice for a home studio that does both voice and instruments. Like every condenser here, it assumes a room that is at least reasonably quiet. Typical street price: $280–$350.
Neumann U87 Ai
The U87 Ai is the microphone most often specified by commercial voice-over clients, and it is the one most likely to be found in a professional studio you are booked into. It is a multi-pattern large-diaphragm condenser with published self-noise around 12 dB-A in cardioid, a sensitivity around 28 mV/Pa, and a maximum SPL of roughly 127 dB.
Should a home voice-over artist buy one? Only if the work justifies it. The U87 Ai’s advantages are pattern flexibility, a familiar tonal signature that clients recognise, and outstanding resale value. Its disadvantage is that it is brutally honest about an untreated room — arguably more so than the TLM 103, because of its low-mid character. If you are buying one microphone to build a career on and your room is treated, it is the safe answer. Typical street price: $3,400–$4,000.
USB microphones for voice over
USB microphones solve a real problem: no interface, no cables, no gain staging. They solve it by moving the preamp and converter inside the microphone body, which is fine for the level of work most people start with. Three current options are worth knowing.
Rode NT-USB+
A cardioid condenser with an onboard headphone output for zero-latency monitoring, a built-in pop filter and a desktop stand. It records at up to 24-bit/48 kHz over USB and includes onboard processing in Rode’s companion software. Its main limitation is that it is still a condenser, so it hears the room — if your room is bad, a USB dynamic is the better choice even though it costs more.
Typical street price: $150–$200.
Elgato Wave:3
A cardioid condenser with a capacitive mute sensor, a headphone output, and 24-bit/96 kHz conversion. It is designed around streaming and podcasting, and its software mixer is genuinely useful if you also record game audio or calls. For pure voice over it is adequate rather than exceptional, and it shares the condenser’s room sensitivity. Typical street price: $130–$170.
Audio-Technica AT2020USB+
A cardioid condenser with a headphone jack and a mix control that blends microphone and playback. It has been a beginner standard for years and remains a reasonable entry point if your room is quiet and your ambitions are modest. Typical street price: $130–$170.
Shure MV7+
Listed here as well as in the dynamic section because it is the most useful hybrid: a cardioid dynamic capsule with both USB and XLR outputs, onboard DSP, a headphone output, and a published frequency range of roughly 50 Hz to 16 kHz. It gives you a dynamic’s room rejection with a USB microphone’s convenience, and it upgrades to a proper interface later without replacing the microphone.
For a single-person home setup in a normal room, it is one of the most practical answers in this entire article. Typical street price: $250–$300.
Shotgun and specialty microphones: when they make sense
Shotgun microphones appear in voice-over conversations because the Sennheiser MKH 416 has been used in film and television ADR and trailer work for decades. A shotgun’s interference tube makes it more directional than a cardioid, but it also makes it more sensitive to off-axis colouration, and its directionality only holds at higher frequencies. For a home voice-over booth, a shotgun is rarely the right answer unless you are matching a specific sound or working in a very controlled space.
Other specialty categories worth knowing about:
- Lavalier and headset microphones. Good for movement and for e-learning video, poor for the neutral, close, full-bodied sound that commercial voice-over clients expect.
- Small-diaphragm condensers. Accurate and neutral, often used for measurement and instrument work. Some narration engineers prefer them for their lack of colouration, but they are typically more sensitive to room than a large-diaphragm dynamic.
- Ribbon microphones. Figure-8 pattern, warm top end, but extremely sensitive to room from both front and rear and fragile in the wrong hands. Not a beginner’s voice-over microphone.
- USB/XLR hybrids. The most useful compromise for anyone who wants to start today and upgrade later without rebuying.
Decision matrix: match the microphone to your situation
| Your situation | Best fit | Why | Budget band |
|---|---|---|---|
| Small apartment, neighbours, HVAC you cannot turn off | Cardioid dynamic, close-miked, high-pass at 80–100 Hz | Room tone will dominate; insensitivity is the only fix short of construction | $100–$450 |
| Kids and pets in the house | Dynamic, floor stand, no desk-mounted arm | Desk thumps and door slams are the main interruptions; dynamics reject them better | $100–$450 |
| Renter who cannot drill or glue panels | Dynamic plus a portable vocal booth or a reflection filter | Reversible treatment only; a dynamic reduces how much treatment you need | $150–$500 |
| Quiet treated booth, under 30 dBA | Low-self-noise condenser, cardioid | The room is no longer your limit; detail and self-noise become the deciding factors | $250–$1,500 |
| Deep, resonant voice | SM7B or RE20 with bass rolloff engaged | Proximity effect plus a naturally warm voice equals mud without a low-cut | $350–$500 |
| Light, bright, sibilant voice | Warm condenser (WA-87 R2) or smooth dynamic (Procaster) | Avoids adding a presence peak on top of a voice that already has one | $220–$750 |
| Audiobook narration, hours at a time | RE20 or a quiet condenser with consistent off-axis response | Consistency across takes matters more than a flattering tone | $260–$1,400 |
| Commercial and corporate work, client-facing | U87 Ai, C414 XLS or TLM 103 | Client familiarity and pattern flexibility reduce friction | $1,100–$4,000 |
| No interface, no plans to buy one | USB dynamic (MV7+) over USB condenser | Same room rejection as an XLR dynamic, without the interface cost | $130–$300 |
| Recording in a closet | Dynamic at 3 inches, floor stand, heavy soft furnishing | Closets are extremely boxy; a dynamic plus absorption is the only workable combination | $100–$450 |
Placement: the free upgrade that beats any microphone
If you change nothing else, change where the microphone sits. These positions are practical starting points, not rules.
- Distance. Dynamic: 2–4 inches. Condenser in a treated room: 6–10 inches. Condenser in an untreated room: 4–6 inches with absorption behind the microphone. Every inch you move away from a cardioid costs you roughly 6 dB of direct sound and increases the relative level of the room.
- Angle. Speak across the capsule rather than straight into it — roughly 15–30 degrees off-axis. This reduces plosives substantially and slightly softens the top end, which is usually welcome.
- Height. The microphone should sit level with your mouth, not your chin. Aiming up at the mouth from below produces a hollow, nasal tone; aiming down produces a dull one.
- Position in the room. Avoid the exact centre of the room and avoid sitting with your back 12 inches from a wall. A position roughly one third of the way along the room’s length, facing the longest dimension, usually gives the most even low end.
- What is behind the microphone. Cardioid microphones reject most of what is behind them, so the wall behind the mic matters less
Affiliate disclosure. As an Amazon Associate we earn from qualifying purchases at no extra cost to you. Prices accurate as of the date shown.