How do I fix bad audio levels after video processing?

Bad audio levels after video processing usually come from one of a few issues: clipping, normalization, codec changes, or accidental gain changes during export. Here’s a practical way to fix it.

1) Identify the problem

First figure out what “bad” means:

• Too quiet → raise gain/normalize audio
• Too loud/distorted → reduce gain, remove clipping if possible
• Volume jumps around → use compression/limiting
• Audio out of sync → re-mux or re-export with correct timing
• Different loudness on platforms → target a standard loudness level

2) Use audio normalization

If the audio is simply too soft or inconsistent, normalize it.

Common targets:

• Streaming / general video: around -14 LUFS integrated
• YouTube-friendly: often around -14 to -16 LUFS
• Broadcast: may require -23 LUFS (EBU R128) or -24 LUFS (ATSC)

In FFmpeg:

Normalize to a target loudness:

ffmpeg -i input.mp4 -c:v copy -af loudnorm=I=-14:TP=-1.5:LRA=11 output.mp4

If you want to inspect first:

ffmpeg -i input.mp4 -af loudnorm=print_format=json -f null -

3) Fix clipping or distortion

If the audio peaks too high, lowering volume may help, but clipped audio may already be damaged.

Quick volume reduction:

ffmpeg -i input.mp4 -c:v copy -af volume=0.8 output.mp4

If the audio is distorted from clipping, try:

• Lowering gain before export
• Using a limiter
• Using restoration tools in audio editors like Audacity, Adobe Audition, or iZotope RX

4) Smooth uneven levels with compression

If speech is hard to hear in some parts and too loud in others:

ffmpeg -i input.mp4 -c:v copy -af "acompressor, loudnorm=I=-14:TP=-1.5:LRA=11" output.mp4

A compressor reduces the gap between soft and loud parts.

5) Check for processing chain issues

Bad audio levels often happen because of:

• Exporting with wrong audio gain
• Multiple normalizations
• Mixing stereo/mono incorrectly
• Applying filters in the wrong order
• Accidentally re-encoding at a lower bitrate

Best practice:

• Keep a clean master copy
• Process audio separately if possible
• Apply normalization once at the end

6) Use an editor if you need manual control

Tools like:

• Audacity: free, good for normalize/compress/amplify
• DaVinci Resolve: good for video + audio adjustment
• Adobe Premiere Pro / Audition
• Final Cut Pro

Typical workflow:

1. Import audio or extract it from video
2. Inspect waveform and loudness
3. Adjust gain/normalize
4. Compress/limit if needed
5. Re-attach or export with video

7) Extract and reattach audio if necessary

If the video is fine but audio needs repair:

Extract:

ffmpeg -i input.mp4 -vn audio.wav

Process the audio in an editor or with FFmpeg, then reattach:

ffmpeg -i input.mp4 -i fixed_audio.wav -c:v copy -c:a aac -b:a 192k output.mp4

8) Verify with loudness meters

Use meters to confirm the fix:

• Peak level
• Integrated LUFS
• True peak
• Dynamic range

If you’re targeting web video, make sure:

• Peaks stay below about -1 dBTP
• Integrated loudness is near your target, like -14 LUFS

Quick recommendations

• Too quiet: normalize to -14 LUFS
• Too loud: lower volume or use a limiter
• Uneven speech/music: use compression + normalization
• Clipped/distorted: repair if possible, otherwise re-export from clean source

If you want, I can give you:

1. a FFmpeg command for your exact case, or
2. step-by-step instructions for Premiere, DaVinci Resolve, or Audacity.