- Set Audacity to use the External mic.
- Ensure you're recording in 16-bit PCM + 44100Hz.
- Do a sound check, record everyone in the room speaking. You want the lighter part of the meter bar (the green or red bar at the top) to bounce between -12db and -6db while people are talking. It's hard to achieve this... so just settle for as close as you can get, without the bars going off the scale :)
- Play it back and tweak levels until it sounds good @ 60% speaker volume.
- Save the Project BEFORE recording live :D (prevents quality issues)
Edit for Content
- Join recordings to one track.
- Listen to it + Edit bad content out.
- Save this copy to Google Drive, or another backup spot.
Edit for Sound Quality
(Low Pass and High Pass should happen AFTER the Compressor)
- Low Pass Filter 10,000Hz @ 24 db roll off. Reduces 'Ssssss' and other high pitched sounds.
- High Pass Filter 30Hz @ 24 db roll off. Reduces low hums (like a fridge) and gravely voice grumbles.
- Equalization 200Hz @ 6db boost, 4,000MHz @ -3db loss.
- Compressor = Threshold -24db, Noise Floor -50db, Ratio 3:1, Attack Time 0.2s, Release Time 1.0s, check both 'Make up gain' and 'Compress based on peaks'. Threshold needs to go lower (more negative) if the resulting track is jumping to different volumes. To set the perfect threshold, listen to 30 seconds of your softest track while watching the dB meter. Use the lowest sensible value. If background noise is getting amplified, you need to raise the noise floor.
- Add intro music.
- Use the Envelope tool to soften music volume.
- Normalize -1db (this should ALWAYS be your last step).
- Amplify makes a track louder. After Amplify you want the light blue part of the track (RSM) to be at exactly -0.5 and 0.5. This is ideal listening volume (corresponds to about -6db final volume).
- Compressor adjusts volume up and down to be more consistently in the middle. Default settings are generally good, usually just need to tweak the threshold. Threshold controls loudness. Loud sounds are softened.
- Normalize adjusts the volume center around the loudest peak. If the final track sounds too loud, tweak this.