Understanding the Root Causes of Crackles and Pops

Before you can effectively clean up your podcast audio, it’s critical to distinguish between the different types of crackling noises. A “crackle” is not a single sound; it can be a short, sharp pop, a series of rapid clicks, a static buzz, or a low-frequency rumble. Identifying the specific type helps you choose the most precise removal tool without harming the vocal clarity.

Common Sources of Crackles

  • Microphone and cable issues: Loose XLR connections, damaged cables, or dirty contacts create intermittent crackles. Even a slightly corroded jack can produce pops that are hard to isolate in post-production. The cable itself acts as an antenna when its shielding is compromised, picking up environmental noise that manifests as crackling.
  • Electrical interference: Fluorescent lights, dimmer switches, nearby power supplies, and unshielded cables can introduce high-frequency static or rhythmic crackling into the signal chain. This is especially common with USB microphones that run alongside power cables.
  • Digital clipping: When audio levels exceed 0 dBFS during recording, the waveform gets “clipped,” creating harsh, square-wave crackles that sound like digital distortion. Once clipping occurs, it cannot be fully removed—it can only be mitigated with a declipper tool that reconstructs the waveform.
  • Plosives and sibilance: While not exactly crackles, explosive “p” and “b” sounds can cause low-frequency rumbles that may be mistaken for crackling. They often require separate de-essing or high-pass filtering rather than a standard de-clicker.
  • Room noise and background artifacts: HVAC systems, computer fans, or external traffic can produce low-level crackling when compressed or normalized. These sounds are often continuous but become transient-like when gated or heavily processed.
  • Buffer underruns in the DAW: A less obvious source: when your computer’s CPU can’t keep up with real-time processing, it may introduce short dropouts that sound like clicks. Increasing the buffer size in your audio interface settings can eliminate these.

Why Simple Filters Often Fail

Many podcasters reach for a broad noise-reduction filter, but these can dull the voice or introduce a “wet” reverb effect. Crackles are usually short, sharp transients that exist in the same frequency range as consonants like “t,” “k,” and “s.” A blanket filter that removes those transients also removes crispness from speech. That’s why you need targeted solutions that treat crackles as isolated events, not continuous noise. Broad noise gates also fail because they only affect the noise floor between words, not the transients that land on speech.

Essential Tools for Removing Crackles

1. Dedicated Noise‑Reduction Software

The most reliable way to remove crackles without distortion is to use software built for audio restoration. Each tool has a slightly different approach, but the principle is the same: you sample the noise floor, then apply a reduction algorithm while preserving the rest of the signal.

  • Audacity (free): Open‑source and widely used. Its Noise Reduction effect works best for steady background hums, but for crackles you’ll need the “Repair” tools or a separate de‑click plugin. The Audacity manual provides detailed steps for capturing a noise profile. For transient crackles, use Effect > Repair on a small selection (around 10–20 ms) around the pop.
  • Adobe Audition: Offers both a “DeNoise” effect (for continuous noise) and “DeClicker” / “DeClipper” effects specifically for transient crackles. The spectral frequency display allows you to see each pop as a vertical line and remove it manually. Adobe’s noise reduction guide covers these tools in depth. The DeClicker lets you adjust sensitivity across three frequency bands—low, mid, high—so you can target only the band where crackles occur.
  • iZotope RX: The industry standard for audio repair. Its “De‑click” module uses machine learning to distinguish between clicks, crackles, and voice. The “Spectral Repair” tool lets you paint over a crackle in the spectrogram, and the algorithm fills in the gap with predicted audio. iZotope offers a free trial and extensive tutorials. For podcasters, the RX Elements edition includes De-click, De-clip, and Voice De-noise at a lower price point.

2. Specialized De‑Click Plugins

If you already use a DAW (like Reaper, Logic Pro, or Pro Tools), consider a dedicated de‑click plugin rather than a general noise reducer. These plugins are optimized for high‑speed transient removal and often offer finer control than the built-in tools in Audacity or Audition.

  • Waves X‑Click: Part of the Waves Restoration bundle. It gives you control over click sensitivity, frequency range, and release time, so you can treat light crackles without affecting vocals. It also includes a “monitor” mode that lets you hear only the removed clicks, helping you dial in the exact threshold.
  • Accusonus ERA De‑Click: Uses a one‑knob interface that’s beginner‑friendly but still powerful. It analyzes the audio in real time and removes clicks while preserving transients in speech. The accompanying ERA Noise Remover can be used in series for a complete cleaning chain.
  • Clarity Vx by Waves: A newer plugin that uses neural networks to separate voice from noise, including crackles. It works well for live streaming and podcasters who need real‑time processing with minimal latency. You can blend the processed and dry signals to find the sweet spot.

3. Spectral Editing (Visual Removal)

For stubborn crackles that automated tools can’t catch, spectral editing lets you see the problem. Open the audio in a spectrogram view (most DAWs and Audacity have this). Crackles appear as sharp vertical spikes, often extending from 1 kHz to 10 kHz. You can select them and apply a “silence” or “attenuate” command. This method is manual but gives you complete control over what stays and what goes. Be careful not to select too wide a time slice, or you’ll cut consonants. A good practice is to zoom in to the waveform, select only the sample region that contains the spike (5–20 ms), and then apply a spectral repair function that fills in the gap with predicted audio based on surrounding frequencies.

Step‑by‑Step Workflow for Removing Crackles

Follow this process to clean your podcast audio without introducing distortion. Adjust the parameters based on the severity and type of crackles in your recording.

Step 1: Make a Backup

Always work on a copy of the original file. If you remove too much or damage the audio, you can revert without re‑recording. Keep the original WAV or AIFF file untouched. Use a new track or duplicate the file in your DAW.

Step 2: Listen and Mark the Worst Spots

Play the entire episode at a moderate level. Mark the timestamps where crackles are most audible. Use markers or a note-taking app. This helps you target your efforts instead of applying blanket effects that might dull the whole recording. Focus on sections with silences or pauses first, as crackles are more noticeable there and easier to remove safely.

Step 3: Apply a High‑Pass Filter First

Set a high‑pass filter at 80‑100 Hz to remove low‑frequency rumble. This often eliminates the “base” of a crackle, making it easier for de‑click plugins to work. Use a gentle slope (12 dB/octave) to avoid thinning the voice. If your voice has natural low end, you can go as low as 60 Hz, but no higher than 120 Hz for spoken word. A steep filter (24 dB/octave) can cause phase issues and make the voice sound less natural.

Step 4: Use a De‑Click Plugin or Tool

Open your chosen de‑click effect (e.g., iZotope RX De‑click or Audition DeClicker). Start with a moderate sensitivity setting (around 40‑50%) and listen. Increase sensitivity until the crackle disappears, but no further. If you hear “gating” artifacts (the voice sounds choppy), back off. In iZotope RX, also adjust the “Event Separation” parameter—lower values treat short, isolated clicks; higher values treat longer crackles. In Audition, use the “Pop” detection slider for loud, short bursts, and the “Crackle” slider for continuous or series of clicks.

Step 5: Manual Spectral Repair for Stubborn Clicks

For remaining isolated pops, switch to the spectrogram. Zoom in to the event, select a narrow time range (5‑15 ms) and a frequency range that covers the click (usually 1‑8 kHz). Use the “Replace” or “Fill” function to have the software predict the missing audio. In iZotope RX, the “Pattern” mode works well for crackles that repeat rhythmically, while “Replace” is best for single, loud clicks. In Audacity, use the “Repair” effect on a tiny selection (about 10–20 samples) centered on the pop.

Step 6: Treat the End of Words and Sibilants

After de‑clicking, check the ends of words where sibilants like “s,” “sh,” and “ch” occur. These can be accidentally attenuated by aggressive de‑click algorithms. If you notice a loss of clarity, use a de‑esser plugin set to a narrow band around 5–8 kHz, or use a multiband compressor to tame those frequencies only when they exceed a threshold. This preserves the natural sharpness of speech.

Step 7: Final Check with Headphones

Listen with high‑quality headphones. Pay attention to sibilants and fricatives; if they sound dampened or artificial, undo the de‑click and try a lower sensitivity. The goal is zero perceptual difference in the voice, just a cleaner background. Also check the processed audio through earbuds and a car stereo—different playback systems can reveal hidden artifacts. Compare the processed version to the original by soloing each track and listen for unnatural gaps or metallic echoes.

Preventing Crackles Before They Happen

The best fix is prevention. Implement these practices to minimize crackles at the source, saving you hours of post-production editing.

Hardware Hygiene and Setup

  • Inspect cables regularly: Look for kinks, cuts, or loose connectors. Replace any cable that feels stiff or crackles when moved. Even a tiny internal wire break can produce intermittent pops that are maddening to locate.
  • Use balanced XLR cables: They reject electrical interference better than unbalanced 1/4” or USB cables. For long runs (over 15 feet), balanced XLR is essential to avoid hum and crackle.
  • Keep gain staging clean: Aim for peaks at -6 dB to -3 dB. Recording too hot (above -1 dB) risks clipping; recording too quiet forces you to amplify noise, which can make crackles louder. Use the preamp gain knob on your interface, not software gain, to set levels.
  • Shield your equipment: Keep audio interfaces away from power bricks, Wi‑Fi routers, and LED lights. Use ferrite chokes on power cables if needed. A simple USB isolator can clean up crackles from ground loops in computer audio.
  • Use a power conditioner: Clean power from a UPS or line conditioner eliminates many electrical crackles caused by fluctuating voltage or RF interference.

Recording Environment and Technique

  • Quiet the room: Turn off fans, air conditioners, and refrigerators during recording. Use sound blankets or foam to reduce room echo, which can mask the onset of crackles and make them harder to isolate later.
  • Mind your mic technique: Maintain a consistent distance (6‑12 inches) and avoid sudden movements that could cause cable noise. Use a pop filter for plosives. Also, avoid touching the mic stand or desk while recording; vibrations travel through the stand and can sound like crackles.
  • Monitor with headphones: Wear closed‑back headphones while recording so you hear crackles in real time and can stop to fix the issue before continuing. Open-back headphones leak sound into the mic and can cause feedback that manifests as crackling.
  • Record at a stable sample rate: Set your project sample rate to 48 kHz or 44.1 kHz. Avoid changing sample rates mid-session, as this can introduce dropouts that sound like clicks.

Software and Firmware Updates

Outdated drivers for your audio interface can cause buffer underruns that manifest as crackles. Check your interface manufacturer’s website for the latest driver. Also, update your DAW and plugins—many restoration tools improve with each version. For example, iZotope’s updates often include better transient detection algorithms. If you use a Mac, check that your macOS version is compatible with your interface drivers; systematic crackles after an OS update indicate a driver conflict.

Choosing the Right De‑Click Plugin Settings

No single setting works for every recording. The key is to understand the parameters most de‑click plugins share:

  • Sensitivity or Threshold: Determines what level of transient is considered a click. Start low and increase gradually. Too high and you’ll remove consonants; too low and crackles remain.
  • Frequency Range: Many plugins let you set a band of frequencies to process. Crackles often live in the 2–8 kHz range. Narrowing the range to that band prevents the plugin from affecting low-frequency voice body.
  • Attack and Release: A fast attack (under 1 ms) captures the initial transient, while a longer release time prevents the plugin from clamping down on the following syllable. For speech, a release of 10–30 ms usually works well.
  • Quality or Resolution: Some plugins offer a trade-off between processing speed and accuracy. For offline processing, choose the highest quality; for live streaming, a lower setting may be necessary to avoid latency.

Test settings on a short section with known crackles. Listen to the processed and unprocessed versions by toggling bypass. If the crackles are gone but the voice sounds the same, you’ve found the right balance.

When to Avoid Too Much Processing

Not every crackle needs to be removed. If a pop is very brief (under 2 ms) and occurs only once or twice in a 30‑minute episode, leaving it alone might sound more natural than a processed artifact. Over‑cleaning can produce a “plastic” or “robotic” quality, especially in the upper mids. Trust your ears: if the fix sounds worse than the original crackle, back off.

Also, be cautious with aggressive high‑pass filtering. A filter set too high (above 120 Hz) will remove the fullness of a deep voice, making the podcast sound thin. Similarly, heavy de‑clicking on a voice with naturally sharp sibilants can destroy clarity. In these cases, use a dynamic EQ that targets only the crackle frequencies and only when they exceed the surrounding audio level.

Putting It All Together: A Real‑World Example

Imagine you recorded a 45‑minute interview, and during playback you hear about 30 short crackles scattered throughout. Here’s how you’d tackle it:

  1. Backup the original WAV file to a separate folder.
  2. Apply a high‑pass filter at 90 Hz (12 dB/octave) to remove subsonic rumble. This alone may eliminate 10% of the crackles that were low-frequency thumps.
  3. Run iZotope RX De‑click at a sensitivity of 50, focusing the frequency range from 2–8 kHz. Listen to the first five minutes to check for artifacts. If you hear any vowel warble, reduce sensitivity to 45.
  4. Spot‑check the remaining crackles in the spectrogram. For a loud pop at 12:35, zoom in to the waveform, select a 10‑ms region around the spike (from 12:35.100 to 12:35.110), and use “Spectral Repair: Replace” with pattern mode. For a series of three rapid clicks at 24:18, select all three and use a longer fill with “Attenuate” rather than replace.
  5. Apply a gentle de‑esser at 6 kHz to tame any sibilance that might have been exaggerated by the de‑clicking process. Use a threshold that only catches peaks above -12 dBFS.
  6. Export the cleaned file as a 48 kHz WAV (or 320 kbps MP3 for distribution) and compare it to the original by toggling mute on the tracks. The cleaned version should sound identical except for the missing crackles.

This approach preserves the natural voice quality while removing distractions. If you have more than 30 crackles in a 45-minute episode, consider re-recording sections with problematic hardware.

Additional Resources and Further Reading

For more advanced techniques, consult these trusted sources:

By combining knowledge of your equipment, careful recording practices, and the right software tools, you can produce podcast audio that is clean and professional—even if you’re working in a less‑than‑ideal environment. The time invested in prevention and targeted cleanup will pay off in listener retention and production quality.