How to Fix Uneven Podcast Volume: A Practical Step-by-Step Guide

Uneven podcast volume usually appears in one of three ways: one speaker is quieter than another, a speaker moves closer to and farther from the microphone, or occasional laughs and emphasized words jump far above the conversation. The solution is not simply to normalize the entire file. Normalization changes the overall level, but it does not reliably correct the relationships inside the episode.

A better workflow is to repair large differences first, control the remaining dynamics, and set final loudness only after the mix is stable. The steps below work in any capable audio editor or digital audio workstation, even though effect names may vary.

The Short Answer

To fix uneven podcast volume:

1. Put each speaker on a separate track when possible.
2. Use clip gain to correct obviously loud or quiet sections.
3. Automate track volume where a speaker changes position or delivery.
4. Apply moderate compression to smooth smaller variations.
5. Balance dialogue, music, and inserted clips by ear.
6. Measure integrated loudness and true peak on the complete episode.
7. Master and quality-check the final mix.

Save an untouched copy before processing. It gives you a reliable way back if a setting introduces pumping, distortion, or audible artifacts.

Step 1: Identify What Is Actually Uneven

Listen once without changing anything and mark each problem. A waveform can reveal large peaks, but it cannot tell you how loud a passage feels. Perceived loudness depends on duration and frequency content as well as peak level, so use both meters and your ears.

Classify each issue:

• Speaker-to-speaker imbalance: the host is consistently louder than the guest.
• Changing microphone distance: one voice fades in and out during the recording.
• Short transients: laughter, plosives, or table bumps create isolated peaks.
• Segment mismatch: the intro, remote interview, ad, and outro were produced at different levels.
• Noise mistaken for quiet speech: raising a low recording also raises its room noise.

This diagnosis matters because each problem needs a different tool. A compressor can smooth speech, but it is a poor substitute for correcting a guest who is 10 dB quieter for an entire interview.

Step 2: Separate Speakers Before You Balance Them

If every participant was recorded to an individual track, edit and process those tracks separately. Lower the loud speaker or raise the quiet speaker until normal conversation feels balanced. Solo each track briefly to find noise or distortion, then make the final decision while listening to everyone together.

When all speakers are baked into one file, you have less control. Split long passages into clips and adjust clip gain around each speaker's turns. Add short fades at edit points to avoid clicks. If people talk over one another, prioritize a natural transition rather than attempting a drastic level change inside the overlap.

For future recordings, multitrack capture is one of the most useful safeguards you can add. Our recording tips for clear podcast audio cover microphone placement and other source-level improvements.

Step 3: Correct Large Differences with Clip Gain

Clip gain changes the audio level before it reaches later effects. Use it for whole phrases or sections that are clearly too loud or quiet. This lets a compressor receive a more consistent signal and reduces how hard it has to work.

Start with broad adjustments, then refine. If a guest gradually gets quieter over five minutes, divide that section into a few natural phrases or draw a gentle gain curve. Avoid dozens of abrupt changes; they can make breath and room tone shift unnaturally.

Treat very loud plosives or laughs individually. Reducing only the problem event often sounds more transparent than applying heavy compression to the entire conversation.

Step 4: Use Volume Automation for Changing Delivery

Volume automation is useful when level changes happen continuously. Draw gradual moves that follow a speaker leaning away from the microphone, becoming animated, or dropping their voice at the end of sentences.

Make adjustments while listening in context. The goal is intelligibility and comfort, not a perfectly flat waveform. Natural speech needs some contrast. If every syllable is equally loud, the result can become tiring and emotionally lifeless.

Step 5: Apply Moderate Compression

Compression reduces the gap between louder and quieter parts once the signal crosses a threshold. For spoken word, begin conservatively:

• Choose a modest ratio, such as 2:1 or 3:1.
• Lower the threshold until ordinary louder phrases produce a few decibels of gain reduction.
• Use an attack that does not blunt every consonant.
• Set release long enough to avoid obvious level flutter between words.
• Add makeup gain only after comparing processed and unprocessed loudness.

These are starting points, not universal targets. A calm narration and an energetic roundtable need different settings. Bypass the compressor frequently at matched listening levels. If the processed voice pumps, lisps, or sounds boxed in, reduce the amount of compression or revisit clip gain.

Step 6: Manage Noise Before Raising Quiet Audio

Boosting a quiet track also boosts hiss, fan noise, and room ambience. Clean persistent noise gently before making large level increases, and listen for metallic or watery artifacts. A noise gate may help between phrases, but an aggressive gate can cut soft words and make the background switch on and off.

If noise is the main obstacle, follow the more detailed podcast background-noise removal guide. Severe noise that overlaps speech cannot always be removed cleanly; accepting a little consistent room tone can sound better than over-processing the voice.

Step 7: Balance Music and Inserted Audio

Check the intro, outro, ads, listener messages, and archival clips against the dialogue. Do not judge music only during an empty section: set its level while speech is present. Use automation to lower a music bed under dialogue and raise it during transitions.

Listen to each boundary from several seconds before the edit through several seconds after it. Sudden tonal changes may be as distracting as volume changes, particularly when remote speakers used different microphones.

Step 8: Normalize and Limit at the End

After the internal balance works, measure the entire program rather than a short excerpt. Integrated loudness describes the episode over time; true-peak measurement helps reveal peaks that can cause problems during encoding or playback. A limiter can catch remaining peaks, but pushing it too hard may add distortion or make speech feel dense.

Loudness targets depend on the publishing workflow and whether the file is mono or stereo. Treat platform guidance as a delivery consideration, not a replacement for a good mix. Read our podcast loudness standards overview for the relevant terms and a measurement workflow.

Step 9: Quality-Check Outside Your Editor

Export a test file and listen from beginning to end, ideally on headphones plus an ordinary speaker or phone. Check:

• Can every speaker be understood without touching the volume control?
• Do laughs and plosives remain comfortable?
• Does background noise surge when a quiet speaker begins?
• Are music and inserted clips consistent with the conversation?
• Are there clicks at gain edits or automation points?
• Does the exported file match what you heard in the editor?

Fix structural and speaker-level problems before mastering. The distinction is important: editing and mastering solve different podcast problems.

Measurement References

The measurement terminology in this workflow follows ITU-R BS.1770-5, the current ITU recommendation for programme loudness and true-peak level. Apple Podcasts' audio requirements provide a current podcast-specific example of applying that measurement before encoding. These references define measurement and delivery guidance; they do not replace a full listen to the exported episode.

Finish a Balanced Episode with SONE

Once your dialogue, music, and edits are combined into a stable final mix, SONE podcast mastering can handle the finishing stage. It is designed to optimize the completed file rather than rebalance isolated speakers inside a multitrack session, so make the major volume decisions first.

Every account receives three free MP3 mastering credits each month; unused free credits do not roll over. One credit is reserved while a job is pending or processing and deducted only after the mastering completes successfully; a failed job releases the reservation without consuming the credit. Free credits can be used only for MP3 output, while promotional and premium credits can be used only for WAV output. That makes a short test episode a practical way to compare your mix before and after mastering without committing your production workflow upfront.