dB Reference Levels Chart: dBFS Targets and Real World SPL
The digital tracking target is peaks between -10 and -6 dBFS with an average around -18 dBFS RMS, which is the digital equivalent of 0 VU on analogue gear. SMPTE alignment places the same reference at -20 dBFS RMS. Above that, 0 dBFS is an absolute ceiling, and -6 dBFS peak is the usual delivery target for a master. Below, 16-bit gives about 96 dB of theoretical dynamic range and 24-bit about 144 dB, from the rule that each bit is worth 6.02 dB.
Definition: A decibel is a ratio expressed logarithmically rather than an absolute quantity, so every decibel figure must state its reference: dBFS is measured against digital full scale, dB SPL against the threshold of hearing, and dBu against 0.775 volts.
Almost every level argument in home recording comes from mixing up two scales. dBFS measures against digital full scale and has a hard ceiling at 0. dB SPL measures acoustic pressure in the air and has no ceiling worth speaking of. They are unrelated numbers that happen to share a unit, and this page keeps them in separate tables for exactly that reason.
What are the dBFS levels that matter in digital recording?
Digital level runs downward from a fixed ceiling. There is no headroom above 0 dBFS, no soft saturation, and nothing to recover once a sample clips. Every target below is defined by how far beneath that ceiling it sits.
| Level | What it is | Why it sits there |
|---|---|---|
| 0 dBFS | Absolute digital ceiling | The largest number the word length can express. Anything above it is clipping. |
| -1 dBTP | True peak delivery ceiling | Leaves room for the intersample peaks that lossy encoding creates on playback. |
| -6 dBFS peak | Mixdown delivery target | Gives a mastering engineer, or the next version of you, room to work. |
| -10 to -6 dBFS peak | Tracking peak target | Headroom for a performer who suddenly gets 6 dB louder in the chorus. |
| -18 dBFS RMS | 0 VU alignment, EBU practice | The level analogue-modelled plugins expect. The site's default target. |
| -20 dBFS RMS | 0 VU alignment, SMPTE practice | The American broadcast reference. Two dB more conservative. |
| -23 LUFS | EBU R128 broadcast loudness | A loudness measurement, not a peak one. Used for television delivery. |
| -14 LUFS | Typical streaming normalisation target | Where most platforms land playback. Mastering louder gains nothing. |
| -60 dBFS | Practical noise gate threshold | Below the audible floor of most rooms once the mix is playing. |
| -96 dBFS | 16-bit theoretical noise floor | 6.02 dB per bit, times 16 bits. The CD format limit. |
| -144 dBFS | 24-bit theoretical noise floor | 6.02 dB per bit, times 24 bits. Well below any analogue circuit. |
The 6 dB per bit rule is the arithmetic behind the last two rows and it is worth understanding rather than memorising. Each additional bit doubles the number of quantisation steps available, and doubling amplitude is a 6.02 dB change. Sixteen bits therefore gives 16 x 6.02 = 96.3 dB, and twenty four bits gives 24 x 6.02 = 144.5 dB, conventionally quoted as 96 and 144.
The consequence is the single most useful thing on this page: headroom is free on a 24-bit system. Record 18 dB below full scale and you still have roughly 126 dB of theoretical range beneath you, far more than any microphone preamp or converter can actually deliver. There is no signal to noise argument for tracking hot, and there is a very real argument against it, which is that a clipped transient is unrecoverable.
How do dBFS, dBu and dB SPL relate to each other?
They do not relate by any fixed universal constant, and this is where confusion starts. What exists is a chain of conversions with a manufacturer-specific link in the middle.
| Scale | Reference point | Domain | Fixed or arbitrary? |
|---|---|---|---|
| dBFS | Digital full scale | Inside the computer | Fixed. 0 dBFS is the largest expressible sample. |
| dBu | 0.775 V RMS | Analogue voltage | Fixed. Historically the voltage giving 1 mW into 600 ohms. |
| dBV | 1.0 V RMS | Analogue voltage | Fixed. 2.2 dB above the dBu reference. |
| dB SPL | 20 micropascals | Acoustic pressure | Fixed. The nominal threshold of human hearing at 1 kHz. |
| VU | +4 dBu at 0 VU | Analogue metering | Fixed by convention, with a defined 300 ms ballistic response. |
| dBFS to dBu | Converter dependent | The conversion itself | Arbitrary. Each interface picks its own alignment. |
| dBFS to dB SPL | Monitor gain dependent | Playback | Arbitrary. Set by your monitor controller position. |
The two arbitrary rows are the ones that trip people. There is no universal answer to "how many dB SPL is -18 dBFS", because it depends entirely on where your monitor volume knob is. What professionals do is calibrate that relationship once and leave it alone: play pink noise at -18 dBFS RMS, set the monitor gain so a meter at the listening position reads a chosen SPL, and mark the knob. From then on the meter and the room agree, and level judgements stop drifting session to session.
How loud is everything, in real dB SPL?
This is the acoustic half of the picture, and it decides microphone choice, placement distance and how long you can safely work. All figures are approximate and measured at the stated distance.
| dB SPL | Source | What it means for recording |
|---|---|---|
| 0 | Threshold of hearing at 1 kHz | The reference point for the whole scale. |
| 10 | Anechoic chamber, breathing | Below the self noise of almost every microphone. |
| 20 | Rustling leaves, a very quiet studio | The floor a purpose-built control room reaches. |
| 30 | Whisper at 3 ft | Around the self noise of a quiet large diaphragm condenser. |
| 40 | Quiet suburban room at night | The realistic floor of a home studio with the computer off. |
| 45 | Home studio with a fan-cooled computer | The number that actually limits quiet recordings at home. |
| 60 | Conversation at 3 ft | Normal speaking level. Also typical background in a house. |
| 70 | Vacuum cleaner, busy office | Nothing usable can be recorded over this without isolation. |
| 79 to 85 | Calibrated mixing level | Where the ear response is flattest and fatigue is lowest. |
| 90 | Acoustic guitar strummed hard at 1 ft | Comfortable for a condenser. No pad required. |
| 100 | Loud singer at 1 ft | Where a condenser may need its pad engaged. |
| 110 | Trumpet bell at 1 ft, orchestra peak | Pad territory for most condensers. |
| 120 | Guitar amp cranked at 1 ft | Dynamic microphone territory. Hearing protection needed. |
| 130 | Snare drum at 1 ft, threshold of pain | Only dynamics and high-SPL condensers survive this cleanly. |
| 140 | Jet engine at 100 ft, kick drum inside the shell | Beyond the rated SPL of most microphones. |
Two rows on that table define most microphone decisions. At 120 to 130 dB SPL, a guitar cabinet or a snare drum will drive many condenser capsules into distortion, which is the actual reason a Shure SM57 ended up in front of every guitar amp in the world rather than any tonal magic. At 30 to 45 dB SPL, the room's own noise floor and a condenser's self noise are within a few decibels of each other, which is why quiet sources recorded at home sound noisy no matter how good the preamp is.
How does the inverse square law change your placement?
In a free field, doubling the distance from a point source drops the level by 6 dB. That is not a rule of thumb, it follows directly from the energy spreading over a sphere whose area grows with the square of the radius.
| Distance | Level | Change from 1 ft | What it means at the microphone |
|---|---|---|---|
| 1 ft | 100 dB | 0 dB | Maximum direct sound, maximum proximity effect. |
| 2 ft | 94 dB | -6 dB | Still direct-dominated in most small rooms. |
| 4 ft | 88 dB | -12 dB | Room reflections now a significant part of the signal. |
| 8 ft | 82 dB | -18 dB | In an untreated bedroom, the room is winning. |
| 16 ft | 76 dB | -24 dB | Beyond the critical distance. You are recording the room. |
The number that actually governs a home studio is the critical distance: the point at which reflected energy equals direct energy. Past that point, moving further away stops reducing the level much because the room is filling in, and all you gain is more room and less source. In an untreated bedroom the critical distance is often under 3 ft, which is why home recordings made from six feet away sound like they were made in a bathroom regardless of the microphone.
Treatment pushes the critical distance out, which is the real reason to treat a tracking space. The treatment calculator sizes the job, and the treatment coverage chart shows how much absorption each surface needs.
The single worst habit I had as a beginner was setting gain from a spoken "check one two" rather than from the actual performance. A singer at a conversational level and the same singer in the second chorus are routinely 12 to 15 dB apart, so a level that peaked at -8 dBFS during the check hit the ceiling on the take. The fix costs nothing: have the performer do the loudest part of the song at full commitment, set peaks to about -10 dBFS from that, and leave it. I have never once wished I had recorded something louder. I have thrown away a first take that would have been the keeper more times than I want to admit.
How long can you safely work at each level?
Hearing damage is cumulative and permanent, and the exposure math is not intuitive because decibels are logarithmic. Two standards are in wide use and they disagree in a way worth understanding.
| Level, dBA | OSHA permitted time | NIOSH recommended time | Studio context |
|---|---|---|---|
| 85 | 16 hours | 8 hours | A loud mixing session. Sustainable all day. |
| 90 | 8 hours | 2.5 hours | Loud monitoring. Take breaks. |
| 95 | 4 hours | 47 minutes | Full band rehearsal in a small room. |
| 100 | 2 hours | 15 minutes | Drum tracking without protection. Do not. |
| 105 | 1 hour | 5 minutes | Standing next to a cranked amp. |
| 110 | 30 minutes | 89 seconds | In the room with a drum kit being played hard. |
| 115 | 15 minutes | 28 seconds | The OSHA ceiling for continuous exposure. |
The practical takeaway for anyone mixing at home: calibrate to somewhere between 79 and 85 dB SPL at the listening position and stay there. That range sits under both standards for a full working day, and it is also where the ear's frequency response is flattest, which means the balance decisions you make at that level translate better than ones made loud. Turning up to check a mix is fine. Living there is what damages hearing and produces mixes that sound thin everywhere else.
Headphones deserve their own warning, because there is no distance to fall back on and no room to absorb anything. Open back designs like the Beyerdynamic DT 990 PRO, 250 ohm leak enough that you naturally run them quieter. Closed back designs like the Beyerdynamic DT 770 PRO, 80 ohm isolate well, which is exactly what you want for tracking and exactly what lets you run them dangerously loud without noticing.
How do you actually set a level to hit -18 dBFS RMS?
Meters make this easy once you know which meter you are looking at. Most DAW channel meters are peak meters by default, and a peak meter tells you nothing about RMS. Switch the meter to RMS or add a loudness meter on the channel while you set gain.
- Have the performer play the loudest section, at performance intent. Not a spoken check, not a warm up.
- Set preamp gain so peaks land around -10 dBFS. On a typical vocal or instrument, an average of about -18 dBFS RMS follows automatically, because the crest factor of most performed material is 8 to 12 dB.
- Watch the RMS reading for a full verse and chorus. If it is drifting well above -14 or well below -24, adjust and re-check.
- Do not add gain after the fact to "get it hotter". The file is fine. Louder is a mix decision, not a tracking one.
- Leave the interface gain alone between takes. Consistent gain across takes is what makes comping painless later.
Interfaces with real metering make this materially easier. A unit with a proper level display like the MOTU M4 lets you set gain without staring at a screen, and units with an assisted gain feature like the Focusrite Scarlett 2i2 (4th Gen) will get a nervous first-timer into the right region automatically. The gain staging calculator works out the relationship between your peak reading and your RMS target, and the recording levels guide covers the full workflow.
Related tools and charts
- Gain staging calculator: peak to RMS, and where to set the preamp
- How to set recording levels: the full tracking workflow
- Cable and connector chart: the +4 dBu and -10 dBV side of the same subject
- Microphone polar patterns chart: choosing a capsule for the SPL you are pointing it at
- Best audio interface for a home studio: metering and gain features compared
Frequently asked questions
What level should I record at in dBFS?
Aim for peaks between -10 and -6 dBFS and an average around -18 dBFS RMS. That -18 dBFS RMS figure is the digital equivalent of 0 VU on analogue gear, and it leaves 18 dB of headroom above the average for transients you did not expect. Recording hotter than this buys you nothing on a 24-bit system, because 24-bit already gives about 144 dB of theoretical dynamic range and the noise floor is never the limiting factor.
Why is -18 dBFS the standard tracking level?
Because it aligns digital with analogue. European broadcast practice sets 0 VU, which is +4 dBu, equal to -18 dBFS. The American SMPTE alignment uses -20 dBFS for the same reference. Both exist so that a signal reading 0 on an analogue meter reads the same place on a digital meter. Most analogue-modelled plugins are calibrated to expect a signal around -18 dBFS RMS at their input, and they behave as designed at that level.
How much dynamic range does 24-bit give you?
About 144 dB in theory, from the 6.02 dB per bit rule multiplied by 24 bits. In practice the analogue electronics in any converter limit you to somewhere between 110 and 125 dB, so the extra bits are not delivering more range at the bottom. What they deliver is permission to leave headroom at the top: you can record 18 dB below full scale and still have far more usable range than a 16-bit system recorded to the ceiling.
How loud is a snare drum?
Roughly 130 dB SPL measured a foot from the head, which is at or above the threshold of pain and comparable to a jet engine at a hundred feet. This is why the microphone in front of a snare has to tolerate high sound pressure and why drummers need hearing protection at every session. A guitar amp cranked in a small room sits near 120 dB SPL at one foot, which is quieter than the snare by a factor of ten in perceived loudness terms.
What is the 6 dB per doubling of distance rule?
In a free field, sound pressure level falls by 6 dB every time you double your distance from a point source. A source measuring 100 dB SPL at one foot measures 94 dB at two feet, 88 dB at four feet and 82 dB at eight feet. Real rooms deviate from this because reflections add energy back, so the rule holds well close to a source and progressively less well as reflected sound starts to dominate.
How long can I safely mix at a given volume?
The OSHA permissible exposure is 90 dBA for eight hours, halving the allowed time for every 5 dB increase, so 95 dBA gives four hours and 100 dBA gives two. NIOSH recommends a stricter 85 dBA for eight hours with a 3 dB exchange rate, which allows only fifteen minutes at 100 dBA. Mixing at 79 to 85 dB SPL sits below both limits and happens to be the level at which the ear response is flattest.
Working out your own room and signal chain? The Home Studio Build Planner is the paid version of these pages: 8 printable worksheets you fill in with your own numbers, plus the full PDF, $29.