Audio Processing

I am quite fond of using a simple compressor and a subtle noise filter, that doesn’t fully remove all the background noise to process production audio.

Here is my workflow in Audition:

Export audio

1. Solo the microphone audio layer from the timeline and set your in/out points.
2. On the top-right of Premiere, press the share icon, which will open the Quick Export dialogue.
3. Select the WAV 48 kHz16-bit preset and choose a File Name & Location - usually the Audio folder in the root of the job folder.
4. Press Export

Import & Backup:

1. Once exported, Right-click on the audio file in the Files panel and select Duplicate to create a backup.
2. Drag/drop the exported audio file into audition.

In Audition

Normalise Levels:

1. Highlight the entire audio waveform.
2. Go to Effects > Amplitude and Compression > Normalize (Process).
3. Set the dB level to -3dB and click Apply.

Noise Reduction:

1. Highlight a short section of the audio where only the background noise is present. (avoid breathing, and wait for a few beats after they finish talking to avoid echo)
2. Go to Effects > Noise Reduction / Restoration > Capture Noise Print. Or use the shortcut - Shift + P.
3. Then, highlight the entire audio waveform. You can use CMD/CTRL + A
4. Go to Effects > Noise Reduction / Restoration > Noise Reduction (Process) or use the shortcut - CMD/CTRL + Shift + P

Noise Reduction settings:

Standard noise reduction

• Noise Reduction: between 50 - 70%. The higher you go, the more you loose the detail in the voice and it sounds like a cheap speaker.
• Reduce by: 15 - 25 dB. Same effect as above, but you can play with these values.
• Spectral Decay Rate: between 30 - 60%. The lower it is, the more “tinny” the audio sounds.
• Smoothing: Between 7 - 12.
• Precision Factor: 7
• Transition Width: 6 dB Noisy environment

If all else fails, just highlight a short bit of dialogue and listen to how the settings above affect it.

Dynamic Range Compression:

1. Highlight the audio waveform.
2. Go to Effects > Amplitude and Compression > Dynamics.
3. In the compressor section, a good starting point is with:
   • 1.5 - 3 ratio,
   • -25 dB Threshold,
   • drop the attack to 0.25 ms,
   • release to 500 ms.
4. I also use the Expander with a -20 dB Threshold and a 1.5 - 2 Ratio.

Normalise to -3 db before exporting

I have found that -3db is the best middle ground for voice so that you still have some wiggle room with music etc. We’ve been setting all mic audio to -3db since 2023

Export:

1. Press CMD/CTRL + S to save the file over the original.