Mixing, stereo, balance and mastering

A rhythm guitar and a synth pad occupy the same space and clash. Give one way to separate them in each of the four dimensions.

1. 01Level

   Set one clearly as the supporting element and pull it a few decibels below the other, so one leads and one supports rather than competing at equal level.

2. 02Frequency

   EQ them complementarily - carve the pad where the guitar's body sits (and vice versa) - so each owns its frequency region.

3. 03Stereo

   Pan the guitar to one side and the pad to the other (or wider), so they occupy different parts of the stereo field.

4. 04Depth

   Put more reverb on the pad to push it back and keep the guitar drier and closer, separating them front to back.

Result: The clash is resolved by separating the two in level, frequency, stereo and depth - the four dimensions of the mix - not by level alone.

Outline how you would build the level balance of a vocal-led band mix, and explain why automation is then needed.

1. 01Set the reference

   Start with the lead vocal (or the kick and bass) at a sensible level as the reference for everything else.

2. 02Add elements against it

   Bring in the rhythm section, then harmony and background parts one at a time, setting each relative to the vocal so the hierarchy is clear - lead loudest, supports below.

3. 03Check gain staging

   Ensure tracks and the mix bus peak below 0 dBFS with headroom, using clip gain or trims so nothing overloads.

4. 04Automate

   Ride the vocal up in the dense chorus and pull instruments back under solos, because a single static balance will not hold through the whole arrangement.

Result: Build the balance from the vocal reference downward with healthy gain staging, then automate levels to keep the balance right through the changing sections.

Explain, using the panning law, why a vocal panned to the centre would sound louder than the same vocal panned hard left if no compensation were applied, and state what the law does.

1. 01Hard left

   Panned hard left, the sound plays from one speaker only at full level.

2. 02Centre without compensation

   Panned centre, the same sound plays from both speakers; two sources contribute, so it would be acoustically louder (by up to 3 dB in power) than the single-speaker case.

3. 03The law's fix

   The panning law attenuates the centre by about 3 dB (feeding each channel at 0.707) so that the total acoustic power - and hence the perceived loudness - stays constant as the sound pans across.

4. 04Result

   With the -3 dB constant-power law, the vocal seems equally loud whether centred or panned to a side, so panning changes position without changing loudness.

Result: Two speakers contribute at the centre, so without a law the centre is louder; the -3 dB constant-power law attenuates the centre (0.707) to keep loudness constant across the pan.

A stereo-widened synth sounds huge in stereo but almost disappears when the mix is checked in mono. Explain the cause and how to fix it.

1. 01Diagnose

   Many stereo-wideners create width by making the two channels partly out of phase; the out-of-phase content is what the ear reads as extra width.

2. 02The mono problem

   Summing to mono adds the two channels; the out-of-phase content cancels, so the widened synth thins or vanishes.

3. 03Check in mono

   Regularly monitor the mix in mono (a mono button) to reveal such cancellations before they reach a mono listener.

4. 04Fix

   Reduce the amount of widening, use a mid-side approach that keeps a solid centred mid, or achieve width by other means (panned doubles, stereo source) that stay mono-safe.

Result: The widener's out-of-phase content cancels in mono; check in mono and use mono-safe widening (less widening, mid-side, panned doubles) to keep the synth present.

Explain how a drum group bus with gentle compression makes the kit cohere, and why it differs from compressing each drum separately.

1. 01Route to a group

   Send all the drum tracks (kick, snare, toms, overheads) to a single drum group bus, so one compressor can act on the whole kit together.

2. 02Gentle bus compression

   Apply a few decibels of gain reduction with a moderate attack and a tempo-related release, so the whole kit moves and breathes together on the hits.

3. 03Why it glues

   Because one compressor responds to the summed kit, the elements pump and recover together as one instrument, which reads as cohesion; separate compressors would each react independently and not lock the kit together.

4. 04Restraint

   Keep the compression gentle - a few decibels - so the kit is glued and energised without being squashed and losing its dynamics.

Result: One gentle compressor on the drum group makes the kit pump and breathe as a single instrument - the cohesion that per-track compression cannot give.

A producer masters a track at about -8 LUFS to be as loud as possible. A streaming service normalises to about -14 LUFS. Explain what happens and what has been lost.

1. 01Normalisation

   The streaming service measures the master at -8 LUFS and turns it down by about 6 dB to reach its -14 LUFS target.

2. 02No loudness gained

   After normalisation the track plays at the same loudness as a track mastered to -14 LUFS, so the extreme loudness gained nothing on that platform.

3. 03What was lost

   To reach -8 LUFS the master was heavily limited, crushing transients and reducing dynamic range; that damage remains after it is turned down, so the track sounds flatter than a more dynamic master played at the same level.

4. 04Better approach

   Master to a sensible loudness near the target with a safe true-peak ceiling (around -1 dBTP), preserving dynamics that survive normalisation.

Result: Normalisation turns the -8 LUFS master down to -14 LUFS, so no loudness is gained but the dynamics crushed to reach -8 LUFS are lost - the loudness war is self-defeating.