How to Monitor Audio Levels While Recording and Never Clip Again

Why Audio Levels Are a One-Way Door

I've worked with enough creators to know that nothing gets a video instantly rejected — whether by an audience or a client — as fast as bad audio. And of all audio problems, the two most common are entirely preventable with a 2-minute pre-roll check: clipping (audio that's too loud and distorts) and too-quiet audio (audio that needs amplification in post, dragging up the noise floor with it).

Both problems are invisible until you watch the playback. The solution is to make them visible before you record, using level meters and headphone monitoring. Here's how to do that in any setup.

The Target Zone: Where Your Levels Should Land

Audio levels in digital recording are measured in dBFS — decibels relative to full scale. Full scale (0 dBFS) is the ceiling. Cross it and you get hard digital clipping: a harsh, crackling distortion that cannot be fixed in post.

Your target peak level for talking-head video is -12 dBFS to -6 dBFS. This gives you:

• Sufficient loudness so you don't need extreme amplification later
• 6 dB of headroom above -6 for occasional louder moments (laughter, emphasis, a word you stress harder)
• A final output level that will be loud enough after normalization in your editing app

If your loudest peaks are below -18 dBFS, your gain is too low. If they're hitting -3 dBFS or above, your gain is too high. Adjust accordingly before rolling.

How to Monitor Levels on Different Setups

Recording on iPhone

Apple's built-in Camera app doesn't show audio meters. To monitor levels on iPhone, use:

• Filmic Pro: Shows real-time stereo meters with color-coded zones (green safe, yellow caution, red clip). Lock your manual audio level by tapping the level bar.
• USB audio interface + headphone jack: Connect a Focusrite Scarlett or similar interface to your iPhone via Lightning-to-USB adapter. Monitor through headphones plugged into the interface.

Recording on a DSLR or Mirrorless Camera

Most Sony, Canon, and Fujifilm cameras display audio meters in the viewfinder or on the LCD during video recording. Go to your audio menu, set levels to Manual, and adjust mic gain while watching the bars. The red zone on a camera meter typically starts around -12 dB, so aim to stay out of the red during normal speech.

Recording on a Mac (OBS, ScreenFlow, Riverside, etc.)

Every major recording application shows audio meters in the interface. In OBS, watch the audio mixer. In ScreenFlow, watch the audio track meter during the system check. Set the input gain in your System Preferences > Sound > Input or in the audio interface software, and do a loud test read before starting the actual recording.

Headphone Monitoring: The Right Way

Watching a meter on a screen is useful; hearing the audio in your ears in real time is better. Direct monitoring — where the audio from your microphone reaches your headphones with near-zero latency — lets you immediately catch problems:

• The mic is too far from your mouth (audio suddenly thin)
• Background noise spiked (air conditioning kicked on, traffic outside)
• You turned your head and your voice went off-axis (level dropped 6 dB)

For direct monitoring on Mac, an audio interface like a Focusrite Scarlett 2i2 or Audient ID4 has a direct monitor knob that routes the input signal to your headphones before it hits the computer — true zero-latency monitoring. Plug your headphones into the interface, not the Mac, for this to work.

If you don't have an interface, plug headphones directly into your Mac and monitor with a slight delay using GarageBand or QuickTime audio input — it's a 20-40ms delay but still useful for catching obvious problems.

Gain Staging: Setting Levels From Source to Destination

Gain staging means controlling the signal level at each point in your chain, from the microphone to the recorder. For a simple home studio setup, your chain is:

1. Microphone sensitivity: Fixed (built into the mic design)
2. Interface or preamp gain knob: Adjust this first. Set it so normal speech peaks at -12 dBFS on the interface's meter.
3. Recording app input level: Leave this at 100% (unity gain) once the interface is set. Reducing it here just degrades your signal without the benefit of a cleaner preamp.
4. Editing app normalization: In post, use Loudness Normalization to -14 LUFS for YouTube/web or -16 LUFS for podcast. This is an even, clean boost across the whole file — not the same as adding raw gain in post.

The Test-Read Ritual: Do This Every Single Session

Before every recording session, do a 30-second test read at your actual recording volume and energy level. Not a whisper, not a projection — the exact voice you'll use. Watch the meters throughout:

1. Normal speech: should peak around -12 dBFS
2. Intentional emphasis (louder words): should peak no higher than -6 dBFS
3. Quiet moments: floor should be at least -40 dBFS or lower (indicating low noise floor)

If you're also using a Mac to run Telepront during this test — so your script scrolls voice-activated while you do the read — this is a natural way to run a full system check: you're checking levels, testing your script pacing, and confirming the voice-scroll is tracking your cadence all at once.

The Most Common Mistakes

• Setting gain for the test and then moving the mic: If the mic moves, the level changes. After setting gain, tape the mic in position or use a mic stand with a locking mount.
• Monitoring with speakers instead of headphones: Speakers bleed back into the mic, creating feedback or affecting your reading. Always monitor with closed-back headphones during recording.
• Only checking one extreme: Don't just make sure you're not clipping — also check that your floor levels are quiet. Open the waveform in your editing app after the test clip and make sure you can see clear silence between words.