Mastering Audio Levels for Clear and Balanced Radio Broadcasts

Understanding the Fundamentals of Audio Levels in Radio

Audio levels govern the perceived loudness and dynamic range of every element in a broadcast. Inconsistent levels cause listener fatigue, while improper headroom invites distortion. Mastering audio levels means controlling the volume of speech, music, and effects so that each element sits naturally in the mix and translates well across any receiver, from car stereos to smartphone speakers.

The human ear is sensitive to sudden jumps in volume. If a host raises their voice or a commercial break exceeds the program level, listeners reflexively reach for the volume control. To prevent this, broadcast engineers rely on standardized measurement units such as dBFS (decibels relative to full scale), dBA (A‑weighted decibels) and, increasingly, LUFS (Loudness Units relative to Full Scale). Understanding these metrics is the first step toward professional audio mastering.

Loudness Standards and Metering

Gone are the days when radio stations simply peaked as high as possible. Modern broadcasting follows strict loudness standards that ensure consistent playback across platforms. The most widely adopted specification is ITU‑R BS.1770, which defines integrated loudness, loudness range, and true‑peak limits. In Europe and many other regions, EBU R128 mandates a target loudness of ‑23 LUFS (±1 LU) for linear programs, with a true‑peak maximum of ‑1 dBTP. Streaming platforms often apply similar standards — for instance, Spotify targets ‑14 LUFS, while YouTube applies ‑14 LUFS integrated. Radio broadcasters must therefore master content that meets both over‑the‑air and digital distribution requirements.

Key metering tools include:

• True‑peak meters – show the actual inter‑sample peak that could cause distortion after D/A conversion.
• Loudness meters – display integrated, short‑term, and momentary LUFS values.
• Loudness range (LRA) – indicates how much the loudness varies over the program; useful for spoken word vs. music.

Using a loudness meter that complies with ITU‑R BS.1770‑4 (or newer) ensures your broadcast meets regulatory limits and sounds consistent across devices. We recommend the free Youlean Loudness Meter for quick checks, or the integrated loudness tools in DAWs like Reaper, Logic Pro, or Pro Tools.

Dynamic Range: Why It Matters for Radio

Dynamic range is the difference between the quietest and loudest parts of an audio signal. In radio, too much dynamic range forces listeners to ride the volume control — uncomfortable during driving or background listening. Too little range (over‑compression) fatigues the ear and removes the natural ebb and flow of a live performance or conversation.

For spoken word, a typical dynamic range target is around 6–10 dB between the average speech loudness and the highest peaks. For music interspersed with voice, aim for a narrower window so that transitions feel seamless. A good rule of thumb: keep the loudest element (usually the anchor’s voice) within 2–3 dB of the music bed’s average level.

Compression: The Workhorse of Level Control

Compression reduces the dynamic range by attenuating signals above a set threshold. It is the single most important tool for taming unpredictable peaks in live broadcasts and for smoothing out voiceovers in pre‑recorded content.

Key Parameters

• Threshold – the level at which compression begins. For radio voice, start around ‑12 to ‑18 dBFS.
• Ratio – how much gain reduction is applied above the threshold. 2:1 to 4:1 works well for voice; higher ratios (6:1, 8:1) are used for aggressive control of screaming or singing.
• Attack – how fast the compressor reacts. Fast attacks (1–10 ms) catch transients like plosives; slower attacks (20–50 ms) let the natural punch of a word through.
• Release – how quickly the gain returns to normal after the signal falls below threshold. Set release between 50 ms and 200 ms for natural‑sounding speech.
• Knee – a soft knee (smooth transition) prevents audible pumping in voice; a hard knee is more suitable for drum bus compression.

Stacking compressors in series — often called serial compression — is common in radio. A first compressor catches large peaks (low ratio, fast attack), while a second adds gentle overall shaping (higher ratio, slower attack). This approach gives more transparent control than a single compressor working too hard.

Limiting: Preventing Digital Clipping

A limiter is essentially a compressor with an infinite ratio (10:1 or higher). Its sole purpose is to prevent signals from exceeding a ceiling, typically ‑1 dBTP for broadcast. A well‑set limiter stops hard clipping without introducing audible distortion. Popular hardware limiters include the Waves L2, FabFilter Pro‑L 2, and the built‑in limiter in many broadcast processors.

Important: limiters should be the last processor in the chain. Do not apply limiting before compression; otherwise, the limiter will fight the compressor and cause unnatural gain‑riding artifacts. Use a true‑peak limiter to catch inter‑sample peaks that standard meters miss.

Equalization and Level Perception

EQ changes how we perceive loudness. Boosting frequencies around 3–5 kHz makes a voice sound more present and “forward,” which can give the illusion of increased volume without raising the peak level. Cutting low frequencies (below 80 Hz) reduces muddiness and allows the compressor to work more efficiently. For radio broadcasts, a gentle high‑pass filter at 60–80 Hz on voice tracks removes rumble and breath pops, freeing headroom for the main signal.

Never over‑EQ in pursuit of loudness. Excessive boosting introduces phase issues and noise. Instead, use subtractive EQ to clean up the mix, then apply gentle compression to bring the apparent loudness up.

Monitoring and Room Acoustics

You cannot master audio levels accurately without a truthful monitoring environment. Cheap headphones or untreated rooms mislead you into making bad level decisions. For radio production, use:

• Closed‑back headphones (e.g., Beyerdynamic DT 770 Pro, Sony MDR‑7506) to prevent bleed during live recording.
• Near‑field studio monitors (e.g., Yamaha HS8, KRK Rokit) placed at ear level and away from walls to avoid bass buildup.
• Room treatment – absorption panels to kill first‑reflections and bass traps to flatten low‑end response.

If you cannot treat the room, check your levels through multiple listening systems: car stereo, earbuds, laptop speakers. A mix that sounds good on all three is likely well‑balanced.

Practical Workflows for Radio Broadcasts

Voice Processing Chain

1. High‑pass filter at 60 Hz to remove rumble.
2. Subtractive EQ: cut any resonant frequencies (often 200–300 Hz for boxiness, 3–4 kHz for sibilance).
3. First compressor: moderate ratio (2.5:1), fast attack (10 ms), medium release (100 ms), threshold around ‑16 dBFS.
4. Second compressor: gentle ratio (1.5:1), slower attack (30 ms), release ~150 ms, threshold ‑12 dBFS.
5. Limiter: ceiling at ‑1 dBTP, attack 0.5 ms, release auto.
6. Final peak‑normalize or loudness‑normalize to target LUFS (e.g., ‑16 LUFS for typical FM radio).

Integrating Music Beds

When mixing voice over music, ducking (side‑chain compression) helps the voice cut through without the music disappearing completely. Set a compressor on the music track, keyed by the voice channel. Use a ratio of about 2:1, fast attack (5 ms), and a release that matches the phrase length (100–200 ms). The voice should sit 3–6 dB louder than the music. Adjust the threshold so the music recedes just enough to maintain clarity without sounding chopped.

Common Problems and Solutions

Problem	Cause	Solution
Distortion on peaks	Limiter ceiling too high or too fast	Lower ceiling to ‑1.5 dBTP; increase limiter release time
Breath pops	Microphone technique; lack of high‑pass filter	Engage high‑pass at 80 Hz; move mic slightly off‑axis
Sibilance (harsh esses)	Too much high‑frequency boost; compressor overreacting to sibilants	De‑esser before compression; cut 5–8 kHz by 2–3 dB
Pumping or breathing	Compressor release too fast or threshold too low	Lengthen release; raise threshold; reduce ratio
Level discrepancy between segments	Different source material not normalized	Apply loudness normalization to all clips before assembly

Loudness Normalization: Final Step

After processing, apply loudness normalization to bring the entire program to the target level. For radio, many stations target ‑16 LUFS integrated, with a loudness range under 10 LU for talk formats. Use the integrated loudness measurement over the entire program — not just peaks. Most DAWs and dedicated broadcast processors (like the Wavelength Audio BPM 100) include this function.

If your station stream also goes to apps like TuneIn or iHeartRadio, check the platform’s loudness requirements. For example, Apple Podcasts normalizes to ‑16 LUFS, while Spotify podcasts target ‑14 LUFS. Publishing one master that meets both can be tricky; consider exporting separate versions for different distribution channels.

Advanced Techniques: Parallel Compression and Multiband Dynamics

Parallel compression (New York compression) blends a heavily compressed copy of the signal with the dry signal. This technique preserves natural transients while adding density — useful for radio jingles or voiceovers that need to sound punchy without being squashed.

Multiband compressors split the audio into frequency bands (e.g., low, mid, high) and compress each independently. This prevents low‑frequency thumps from triggering compression on the voice’s midrange. For example, a multiband compressor can tame a boomy announcer’s voice at 150 Hz while leaving the 2 kHz presence untouched. The iZotope Ozone Dynamic EQ and FabFilter Pro‑MB are excellent software options.

Best Practices for Live Broadcasts

Pre‑recorded audio offers time to perfect levels. Live radio demands real‑time decisions. Follow these guidelines to stay consistent:

• Set a reference level before airtime: calibrate the board so that a tone at ‑20 dBFS equals 0 VU on the analog meter.
• Use a hardware compressor between the microphone preamp and the console insert point. Set it conservatively (3:1 ratio, 20 ms attack, 200 ms release, threshold at ‑10 dBFS).
• Double‑check that your limiter (if used live) has a look‑ahead function to avoid clipping on sudden shouts.
• Train DJs to maintain consistent mic distance (2–4 inches from the pop filter) and to avoid shouting directly into the capsule.
• Monitor the broadcast off‑air (delayed) to catch problems early.

Tools and Resources

To master audio levels effectively, invest in reliable tools:

• FabFilter Pro‑L 2 – transparent true‑peak limiting and loudness normalization.
• Voxengo Elephant – free limiter with multiple character modes.
• iZotope Ozone – mastering suite with multiband dynamics, EQ, and loudness matching.
• Hardware processors: dbx 286s for mic channel, Orban Optimod for station processing.
• Free loudness meters: Youlean Loudness Meter (mentioned above) and TBProAudio dpMeter.

Conclusion

Mastering audio levels for radio is a blend of technical knowledge, careful metering, and consistent workflows. By understanding loudness standards, applying compression and limiting intelligently, and monitoring in a truthful environment, you can deliver broadcasts that sound professional, engaging, and comfortable to listen to for hours. Test your processing on multiple playback systems before going live, and keep learning as new standards emerge. Your audience will thank you with loyalty and trust.