Understanding Audio Compression: A Broadcaster's Guide to Dynamic Processing

Nothing separates amateur from professional broadcast audio as clearly as dynamic range control. A raw microphone signal has enormous dynamic variation—a whisper and a shout from the same person can differ by 40–50 dB. Without any dynamic processing, your listeners will constantly be adjusting their volume: turning it up to hear your quiet moments, then frantically turning it down when you raise your voice. Professional radio has used compression as its primary tool to manage this problem for over 60 years. This guide explains how compression works and how to configure it effectively in BroadcastMySelf for broadcast-quality results.

What Is Dynamic Range and Why Does It Matter for Radio?

Dynamic range is the ratio between the quietest and loudest sounds in an audio signal. Human speech has a natural dynamic range of approximately 30–40 dB in conversational contexts—your voice is significantly louder when you emphasize a word or laugh than when you speak softly. Music has an even wider dynamic range: the quiet intro of a track might be 20–30 dB quieter than the chorus.

For radio broadcasting, a wide dynamic range is a problem because listeners consume radio in noisy environments—cars, kitchens, gyms—where the ambient noise competes with quiet passages in your broadcast. A listener whose car has 65 dB of road noise will simply not hear content at -15 dB relative to your peaks. Additionally, wide dynamic range content sounds inconsistent and unprofessional to trained ears.

Broadcast compression solves this by automatically reducing the level of loud passages to bring them closer to the level of quiet passages. The result is a "tighter" dynamic range—the loud parts are not as much louder than the quiet parts as they were in the raw signal. This makes the entire broadcast more audible and consistent at any playback volume.

How a Compressor Works

A compressor monitors the level of your audio signal continuously. When the level exceeds a set threshold, the compressor reduces the gain by a ratio you specify. Below the threshold, the signal passes through unaffected.

The five core parameters of a compressor are:

• Threshold: The level in dB at which the compressor begins working. Signals below this level are not processed. Example: -12 dB means the compressor activates whenever your voice exceeds -12 dB on the meter.
• Ratio: The degree of gain reduction applied to signals above the threshold. A ratio of 4:1 means that for every 4 dB a signal exceeds the threshold, the output only increases by 1 dB. A ratio of 2:1 is gentle compression; 8:1 or higher approaches limiting behavior.
• Attack: How quickly the compressor responds after the signal exceeds the threshold. Measured in milliseconds. A fast attack (1–5 ms) catches transients immediately. A slow attack (50–100 ms) lets the initial transient through before compression engages, preserving punch and clarity.
• Release: How quickly the compressor stops reducing gain after the signal falls back below the threshold. A fast release can cause the compressor to "pump" audibly with the rhythm of speech. A slow release sounds smoother but can hold the gain down longer than needed.
• Make-up Gain: Since compression reduces the level of loud peaks, the overall signal is often quieter after compression. Make-up gain (also called Output Gain) is added after the compressor to restore the overall level to approximately where it was before processing, or higher.

Broadcast Compression vs Music Production Compression

The way compression is used in broadcast differs significantly from how it is used in music production. Music producers often use compression creatively—to change the feel of a drum kit, add punch to a bass line, or create a particular tonal character. In broadcast, the goal is almost always transparent, functional dynamic control. Listeners should not hear the compressor working; they should simply experience a consistently leveled, easy-to-listen-to voice.

This distinction matters for your settings. Aggressive compression that adds "color" to a drum sound would make a broadcast voice sound unnatural and processed. Broadcast compressors are set conservatively to control the dynamic range without audible side effects.

Recommended Compressor Settings for Voice Broadcasting

This preset provides meaningful dynamic control without heavy-handed processing. The 4:1 ratio tightens your dynamic range without making your voice sound compressed. The 10 ms attack preserves the initial consonants and plosives that provide clarity and intelligibility. The 100–150 ms release prevents pumping between words.

After applying this preset, check your output meter. The goal is to have your average speaking level consistently around -6 dB on the output meter, with occasional peaks reaching up to -3 dB. If the average level is lower, increase the make-up gain. If peaks are regularly hitting 0 dB, reduce the input gain or lower the threshold slightly.

The Compressor's Relationship with Other DSP

Compression does not work in isolation—it interacts with every other processor in your signal chain. The most important interactions to understand are with the noise gate and the equalizer.

Noise Gate before Compressor: This ordering is critical. If you place the noise gate after the compressor, the compressor will amplify your background noise (by applying make-up gain to quiet passages) and then the gate must work harder to suppress it. Placing the gate before the compressor means only clean, intentional voice signal reaches the compressor's input.

EQ before Compressor: For broadcast voice, applying EQ before the compressor (cutting the low-frequency rumble below 80 Hz, for example) prevents the compressor from reacting to low-frequency energy that you have already removed. This is called "feed-forward" EQ and produces a more natural-sounding result than placing EQ after compression.

The optimal signal chain order for broadcast voice in BroadcastMySelf is: Input Gain → High-Pass Filter (80 Hz) → Noise Gate → EQ → Compressor → Output/Make-up Gain.

Identifying and Fixing Common Compression Problems

Once you understand what compression sounds like when applied correctly, recognizing and fixing problems becomes straightforward:

• Pumping or breathing: The most recognizable compression artifact—the background level audibly rises and falls in rhythm with your speech. This is caused by too fast a release time. Increase the release to 200–400 ms to smooth this out. Alternatively, the threshold may be set too low, causing the compressor to engage and disengage on nearly every word.
• Loss of clarity and punch: If your consonants (the "t," "k," "p" sounds that define word clarity) sound soft or smeared, your attack time is too fast. Increase it to 15–30 ms to allow these transients to pass through before the compressor engages.
• Voice sounds thin or over-processed: The compression ratio is too high, or the threshold is set too low. Reduce the ratio to 3:1 and raise the threshold until the gain reduction meter is only showing 4–8 dB of reduction on loud peaks.
• Overall level is too low after compression: You need more make-up gain. This is not a compression problem—it is simply a gain staging issue. Increase make-up gain until your average speaking level hits -6 dB on the output meter.

Loudness Standards for Internet Radio

Unlike broadcast television, internet radio does not have a mandatory loudness standard. However, the streaming platforms that carry your station (TuneIn, Radionomy) and many smart speakers apply their own loudness normalization to streams. Streaming platforms typically normalize to approximately -14 to -16 LUFS (Loudness Units Full Scale).

If your stream is significantly louder than the normalization target, it will be turned down on these platforms, and your deliberate mastering choices will be undone. If your stream is significantly quieter, it will sound diminished alongside other stations. The practical recommendation is to aim for an average loudness of -14 to -16 LUFS for your broadcast, measured over a representative 10-minute segment. This places your station at the normalization target before any platform processing is applied.

Achieving this consistently requires appropriate compression settings combined with correct gain staging at the input. A well-configured compressor with 4–6 dB of make-up gain, applied after correct input gain calibration, will typically produce a broadcast that lands in this range naturally.

Conclusion

Dynamic compression is the single most important audio processing tool for broadcast quality. It transforms an inconsistent, variable microphone signal into the polished, consistent level that defines professional radio. The settings are not mysterious—a threshold at -12 dB, a 4:1 ratio, a 10 ms attack, and a 100 ms release will serve most voices in most environments. Configure once, verify with a recording, and trust the tool to do its work. Your listeners will hear the difference, even if they don't know why it sounds better.