Apple Podcasts loudness without artifacts
This is a practical guide for creators and producers who want speech that sounds natural and passes review. You will learn how to measure loudness, how to set safe margins, how to clean without swirl, and how to export mixes that survive transcodes. The approach is simple. Use AI where it speeds the boring parts. Keep a senior ear on the final choices. Protect tone at every step.
What success looks like
- Your show feels clear at low volume on a phone
- Loudness sits in a steady range with no pumping
- True peak remains below the red
- Breath and room are present but not distracting
- Intros and ads are not jumping above the host
- The file sails through review on the first try
Why loudness matters
Listeners ride the volume knob when loudness swings. Review teams reject files when peaks clip. Platforms apply their own processing which can add new artifacts if your mix has no headroom. Good loudness targets are not about volume for its own sake. They are about consistency and headroom. The best mixes feel relaxed and confident.
Basic ideas with plain language
- Loudness is an average of how loud the program feels to a person
- LUFS is a unit for that feeling
- True peak is the highest moment the signal reaches between samples
- Integrated loudness reads the whole section
- Short term loudness reads smaller windows
- Set a target for integrated. Watch short term to keep speech from spiking
Targets and safe margins
Stereo Shows
Often sound great near -16 LUFS integrated
Keep true peak near -1 dBTP
Mono Shows
Often sit well near -19 LUFS integrated
Keep true peak near -1 dBTP
These are steady ranges that test well. If a network gives you a spec, use that spec.
Signal chain overview
- 1. Start with cleanup that is gentle. Aim for small corrections that add up
- 2. Shape tone with subtle EQ
- 3. Control dynamics with light compression and limiting
- 4. Confirm translation on phones and small speakers
- 5. Print at a stable format
- 6. Listen back to the rendered file
This seems slow. It is faster than fixing a reject.
Proprietary methods from AI Audio Expert
Anchor Talker calibration
Pick the most present host phrase. Twelve seconds is enough. Measure integrated loudness on that phrase. Place the anchor one to two LU below your final target. Now mix the rest of the show by ear toward the anchor. This reduces chasing meters later. The anchor keeps you honest when guests arrive with different mics.
Swirl Guard 30 rule
Add the total strength you apply from denoise and de-reverb. Keep that sum near thirty percent for dialogue. If the room is worse than that, pause. Solve the problem with manual edits or scene choices. Over that line artifacts rise faster than benefits. This single rule prevents most metallic tails in podcast speech.
Dialogue Window check
Pair two views while you work. Watch integrated over one minute. Watch short term over five to ten seconds on bright and sibilant phrases. If short term sprints more than six LU above your minute read, do not slam a limiter. Ease sibilance a decibel. Reduce compression a touch. Let consonants breathe.
Phone Sum test
Play the mix through one phone speaker at low volume. Sum to mono. If words blur or vanish, narrow wide effects, remove phasey ambience, and align stereo sources. Meters do not catch this early. The phone test does.
SNR Ladder method
Measure the room at a silent moment. This is your noise floor. Measure a typical sentence from the anchor talker. Subtract the floor from the sentence level. You now have a speech to noise ratio. Push cleanup only until you reach a comfortable rung on the ladder. For speech, fifteen to twenty five dB tends to feel natural. This method stops you from chasing a silent floor that kills tone.
Advanced techniques
Limiter Triage steps
If the limiter grabs on laughs or plosives, step back. Lower the input by one decibel. Lengthen release. If that is not enough, move the work earlier. Use a small clip gain ride on the offending word.
Breath Budgeting
Breaths give pace and life. Loudness workflows can lift breaths too much. Before you chase LUFS, trim or fade heavy breaths. Keep a simple rule. The biggest breath in a section should feel quieter than the smallest word.
C Weight Translate
If you use a meter that can switch weightings, check a pass with C weighting. For bright voices, C can show issues that K lines do not show alone. If C shows harsh spikes during S sounds, handle de essing before you print.
Quiet Car test
Play the mix on earbuds while walking or riding as a passenger. Keep volume low. If you can follow the story without thinking about it, your mix is ready. If you are straining, go back to sibilance, midrange balance, and short term loudness control.
Podcast Trim Map
Sketch a simple timeline map with marks for intro, host monologue, guest A, guest B, ad, outro. Write down a trim note for any section that feels hot or dull relative to the anchor.
Gear and room tips
Record in a soft room. A closet with clothes beats a bare office. Use a pop filter. Aim the mic to reduce plosives. Keep consistent distance. Capture ten seconds of room tone.
Production checklists you can print
Capture
- • Record in a soft room
- • Use a pop filter
- • Keep a steady mic distance
- • Capture room tone
- • Name files by speaker and segment
Prep
- • Select an anchor phrase
- • Build a trim map
- • Gather notes on hot spots
- • Save a clean copy before you start
Cleanup
- • Apply small denoise only if hiss is steady
- • Add small de reverb only if words mask
- • Keep total near thirty percent
- • Fix clicks and plosives by hand
Tone
- • High pass rumble if needed
- • Ease mud only when real
- • Add small presence if words blur
- • De ess only enough to stop sting
Dynamics
- • Use low ratio compression in two gentle stages if needed
- • Set limiter true peak near minus one
- • Watch integrated and short term windows
QC and export
- • Pass the meter test
- • Pass the phone test
- • Pass the skip test
- • Export a master
- • Export delivery copies from the master
- • Listen to the actual file
Quick start recipe
- 1. Pick the anchor
- 2. Clean gently
- 3. Match guests to the anchor
- 4. Shape with small EQ moves
- 5. Compress two to three dB across speech
- 6. Set limiter ceiling near minus one
- 7. Adjust gain to reach your integrated target
- 8. Check short term on sibilant lines
- 9. Listen on a phone
- 10. Export
- 11. Listen to the export
- 12. Send
Common problems and simple fixes
Swirl after cleanup
Fix: Reduce de reverb amount. Recover body with a small lift near one to two kilohertz. If tails sound metallic, turn the module off and handle by hand. Short edits on the worst words often beat heavy settings.
Breaths jump after loudness
Fix: Trim heavy breaths before the loudness pass. Add small fades. If breaths still pop, try one dB of expansion instead of pure reduction. The breath remains natural and less obvious.
Limiter grabs on laughter
Fix: Lower the input a decibel. Lengthen release. If the laugh still drives, cut a tiny spot on the laugh with clip gain. Then restore your previous limiter input. Do not let the limiter shape your tone.
Remote guest mismatch
Fix: Find the center of each voice. Use EQ to meet in the middle before you match loudness. Add a tiny amount of room tone under edits so cuts feel smooth. If a guest is harsh, reduce five to seven kilohertz more than you think.
Frequently asked questions
Do I need to hit a single number exactly?
No. Hit a steady range that feels natural. Consistency beats perfection.
Should I master in mono?
Mix in stereo if the show is stereo. Check mono often. Many listeners hear on one small speaker. The Phone Sum test keeps you safe.
Can I fix clipping fully?
Sometimes. If consonants vanish, a pick up line or a creative edit may beat any tool.
What bit depth and rate should I export?
Twenty four bit at forty eight kilohertz is a reliable choice for a master. Delivery copies can be made from the master.
How loud should music be under a host?
Keep it lower than you think. If words fight the chord, notch the chord. The voice wins.
Summary you can keep near the desk
Anchor one voice. Keep cleanup gentle. Watch integrated and short term. Hold true peak below minus one. Check on a phone. Export from a clean master. Write down your numbers. Repeat next week.
Need Professional Help?
Our expert team can handle complex audio issues with professional oversight and quality control. We use AI where it speeds the boring parts and keep a senior ear on the final choices.
Get a Quote