Audio Mixing Bootcamp
Notes from a LinkedIn course.
- make a copy for every mix version (optional)
- remove empty tracks
- remove / deactivate unused tracks
- group tracks together (synchronous volume changes, mute etc.)
- rename tracks with a descriptive and short names
- color code tracks (groups)
- Aux is for subgroups (applying effects to a set of tracks)
- saves system resources
- A bus is just an audio pipe, a way of manipulating streams of audio in a structured manner
- An insert is an input; e.g. in an aux channel, inserts are identifying sources of audio
- pre-delay: onset of the effect after the signal
- ~20 ms
- decay: short 1.2 s, long 2 s
- usually in stereo
- usually in mono
- delay 175 ms is favourite of McCartney
- feedback: a number of repeat
Basics of Mixing
- Groove is the pulse of the song.
- It doesn’t have to be perfect, it’s created by tension against even time.
- Doesn’t need to follow drums.
- What instruments provide the pulse?
- Find the most important element
- no contrast
- no focal point
- lacking clarity and punch
- sound distant
- inconsistency of levels
- dull and uninteresting sounds
- Mix bus level will get louder with every instrument entrance (~3 dB)
- Start with mix bus at –10 dB
- The sound of every drum will change when a new drum or cymbal is added.
- Kick outer mic should be barely audible
- We usually don’t pan drums (snares) and bass, they loose some power.
- Pan something to stay out of way.
- Background voices are good a bit to left and right, not fully.
- Panning 20–30 to left and right for spaciousness.
- Drums are panned 90% from audience perspective.
- Hi-hat panned right.
- More panning → less powerful.
- Avoid pseudo-stereo (recorded, simulated stereo).
- Panning everything hard left and right will mingle everything.
- Better pan stereo tracks across the space spectrum.
- Reverb can add stereo effect (spaciousness).
- Automated level control.
- Ratio: How much the level increase according to the input level. 4:1 means every 4 dB input goes to 1 dB output.
- Threshold: Signal level when the compression starts to be applied.
- Attack: How fast/slow compressor reacts on the beginning of amplitude.
- Release: dtto on the end of amplitude.
- Gain (output): compensation for the attenuated compressed signal
- Knee: onset of compressor
- Side-chain is linked to EQ or other filters or tracks (in el. music).
- Compressor is usually before EQ
- Applied on tracks with variations of dynamics.
- Compressor should “breathe” with the track.
- Compressing vocal is the way to keep vocal audible and comprehensive together with the band; opposite to the pop songs where vocal is above the rest
- Ratio up to 10:1 is for correcting dynamics, higher ratio creates a specific effect: limiting.
- The New York compression trick (parallel compression): second compressor on a group set to barely audible level together with the original group
Noise gates (expander), de-essers
- The opposite of compressor
- A special type of compressor: de-esser – compressor on some frequency interval
- number of bands
- gain, frequency, queue (bandwidth)
- subtractive equalization (technique)
- attenuating instead of boosting
- phase-shift is made as a deformation when boosting
- start with HMF and find the position best improving the track
- or boost and find the worst place and then subtract
- works especially wel in 200–600 Hz (lot of mic proximity effect) and 2k–4k (presence boost)
- boosts: point 1k–5k, sparkle 5k–10k, air 10k–15k; just 1–2 dB
- juggling frequencies (technique)
- e.g. two similar instruments
- boost frequencies on one and attenuate on the other
- high pass filter for cleaning up recordings
- if it sounds muddy, attenuate at 250 Hz
- if it sounds honky, attenuate at 500 Hz
- cut or attenuate to add clarity
- boost to make things sound different
- use narrow bandwidth 6–10 when cutting and 0.5–2 when boosting
last modified: 2023-02-02
last modified: 2023-02-02