What Level Should You Record At? dBFS Targets Explained

What level should you record at?

Aim for an average level around -18 dBFS with peaks landing somewhere between -12 and -6 dBFS. Sound On Sound's technical editor recommends treating -18 dBFS as the digital equivalent of 0 VU, with the average hovering there and peaks typically kicking up to around -10 dBFS^[1].

Focusrite gives a simpler rule for its Scarlett interfaces: aim for a -12 dBFS peak so you keep enough headroom once all your tracks are recorded^[4]. Both targets say the same thing in different ways. Leave space above the loudest moment of the performance, because singers always hit the chorus harder than they did during the soundcheck.

Use the table as a starting point, then adjust for the performer in front of you.

What does dBFS mean on a recording meter?

dBFS means decibels relative to full scale, where 0 dBFS is the loudest level a fixed-point digital system can represent. SSL defines 0 dBFS as the maximum undistorted level, with negative values safely representable and positive values distorted^[2]. Every reading on your input meter is a negative number counting down from that ceiling.

Two kinds of reading matter. Peak is the instantaneous highest sample and tells you how close you are to clipping. Average (often shown as RMS) tracks the body of the sound and matches perceived loudness more closely. Most interface meters and DAW track meters show peaks. That is why the -18 dBFS advice confuses people: it describes the average, and a peak meter on the same signal will read much higher.

If your DAW only shows peaks, use the peak target from the table and ignore the average column. A free metering plugin with an RMS or VU mode shows both at once.

Why is -18 dBFS the usual reference level?

Because -18 dBFS copies the headroom that analog studios always had. The EBU recommends that digital equipment use an alignment level 18 dB below the maximum possible coding level, regardless of how many bits are available^[3].

Sound On Sound explains the logic from the analog side. Analog gear was designed to clip at about +24 dBu, roughly 20 dB above 0 VU, but analog meters never showed that margin^[1]. Recording a digital signal that averages around -18 to -20 dBFS simply recreates the same safety zone. The same answer adds that pushing every peak to the top of the meter makes live performances harder to capture and mixing needlessly difficult^[1].

There is also a practical side for home producers. A track that arrives at a moderate level leaves room for EQ boosts, saturation and compressor makeup gain before anything on the channel turns red. The mixing half of that story is covered in what gain staging is and why it matters.

Why doesn't recording hotter sound better at 24-bit?

Because 24-bit already has far more dynamic range than your mic, preamp and room can use. SSL lists the theoretical dynamic range at 96 dB for 16-bit and 144 dB for 24-bit, and points out that noise in the analog electronics around the converter is the real bottleneck^[2].

In practice, the hiss you hear comes from the preamp, the mic's self-noise and the room, not from the digital format. Sound On Sound notes that recording hotter does not improve the noise floor, since acoustic sources have worse signal-to-noise ratios than the recording system^[1]. Dropping your peaks from -2 dBFS to -12 dBFS only lowers the signal in relation to a digital noise floor that sits far below anything audible. The analog noise drops with it.

Recording hot made more sense with tape and 16-bit converters, where the noise floor sat much closer to the music. Check that your interface is set to 24-bit. In FL Studio, for example, the recording bit depth comes from the audio device's own settings, not from the mixer's save option^[6].

Does 32-bit mean you can ignore gain?

Only for true 32-bit float recorders. SSL is explicit that its interfaces use 32-bit fixed-point converters, and you still need to set an appropriate gain to avoid clipping^[2]. The 32-bit float devices built for field recording use multiple converters per channel to extend range, which is a different design^[2]. The format itself is explained in what 32-bit float audio is.

What do the clip indicators on your interface mean?

A red clip light means the signal hit 0 dBFS, or the preamp itself overloaded, and the peaks were flattened into distortion. On the Scarlett 2i2 4th Gen, the Gain Halo turns green for a good level, amber for pre-clip, and red when the input has clipped^[4]. Focusrite's meter diagram marks 0 dBFS as clipping and tells you to turn down the input gain^[4].

Clipping at the input cannot be fixed later. SSL notes that distortion inside the preamp cannot be undone, and any converter signal that would go above 0 dBFS is clipped at exactly 0 dBFS^[2]. Lowering the fader afterward only makes the distorted take quieter. The DAW's internal floating-point mix engine is a different story, which is why a red meter on a mix bus is not the same emergency as a red light on the interface.

1. Stop and listen back
   Solo the take around the clip. If you hear crackle or a flattened, buzzy edge on the loudest word, re-record it.
2. Lower the input gain
   Turn the interface gain down by 3-6 dB, not the DAW fader. The fader has no effect on what gets recorded.
3. Check for a pad
   Very loud sources such as guitar amps or kick drums may need a pad switch on the mic or interface, if one is available.
4. Punch in the clipped section
   Re-record the affected line instead of the whole take, keeping the new gain.
5. Consider a clip-protection feature
   Some interfaces offer one. Focusrite's Clip Safe reduces preamp gain when the input nears clipping^[4].

Input gain vs fader: which one sets the recording level?

The input gain knob on your interface sets the recording level; the DAW's track fader does not. Apple's Logic Pro guide states that adjustments to the fader do not affect the recording level, they only affect the monitoring level^[5]. FL Studio's recommended External input mode says the same: the recording level is controlled by the audio interface and cannot be changed inside FL Studio^[6].

This trips up a lot of beginners. They see a quiet waveform, push the fader up, record again, and get the same quiet waveform. Every control in the chain has one job:

• Interface input gain Sets how much signal reaches the converter. This is the only control that changes the recorded file.
• DAW track fader Changes playback and monitoring volume after the fact. Use it for the mix, not for tracking.
• Clip or region gain Raises or lowers a recorded file after the fact. Useful for evening out a take before compression; see using clip gain before compression.
• Headphone and monitor knobs Change only what you hear. A performer who wants more of themselves needs more headphone level, not more input gain.
• Instrument or synth output For line sources, the device's own volume knob matters too. Set it near its nominal level, then trim with the interface gain.

How do you set your recording level step by step?

Set the gain while the performer sings or plays the loudest part of the song at full intensity, then leave it alone for the take. Soundchecking with a quiet verse is the most common reason home vocals clip in the chorus.

1. Confirm 24-bit
   Check the session or audio device settings. 44.1 or 48 kHz at 24-bit is fine for almost any home recording.
2. Start with the gain low
   Turn the interface gain down before the mic or instrument is connected, and switch on phantom power if a condenser needs it.
3. Rehearse the loudest section
   Have the singer perform the chorus or the loudest ad-lib at full volume, at the distance they will actually use.
4. Raise gain to the peak target
   Bring the gain up until peaks land around -12 to -6 dBFS on the interface or DAW meter. Stop there.
5. Check the quiet parts
   Make sure verses still show healthy movement on the meter, roughly -30 to -20 dBFS at their quietest phrases.
6. Set a comfortable headphone mix
   Adjust the headphone and fader levels for the performer. None of these changes touches the recording level.
7. Record and re-check after the first take
   Look at the waveform. If peaks got within 1-2 dB of the top, drop the gain by 3 dB and go again.

Once the takes are in, balancing them across the session is a mixing job; gain staging in FL Studio and Ableton Live picks up from there, and what headroom means in mixing covers the space you will need on the mix bus.