In this post I am describing highly advanced compressor techniques that are beyond pro level.
When we work with core processing like volume fader, eq and compression, we can do this at multiple scopes - A) individual track, B) group and C) mix.
The frequency density increases from A to C and from a compression standpoint that is key to be aware of.
Upon applying compression to the mix, we have to have a track routing model that allows for a good compression distribution.
So this means that rather than forcing as much frequencies straight from A to C and apply compression on C, we need to go through B in a clever way.
Furthermore, we need to add the compression in stages from tight to relaxed.
This is done by scaling out B both broadly and deeply and then temporarily add lots of density as early as possible where the density is very low and on top of that dialing this in at the first chorus where the frequency density naturally is very high.
Let me give you an example.
Let's say we have the following sound sources in the mix:
- Vocals
- Background Vocals
- Kick
- Snare
- Hi-Hat
- Percussion
- Cymbals
- Overhead
- Room
- Bass
- Piano
- Organ
- Electric Guitar 1
- Electric Guitar 2
- Acoustic Guitar
- Pedal Steel Guitar
In this case, the B groups/aux routing could be like this:
Group Layer 1
- Group A: Kick, Bass
- Group B: Vocals, Snare
- Group C: Hi-hat, Percussion
- Group D: Piano, Electric Guitar 1
- Group E: Organ, Electric Guitar 2
- Group F: Cymbals
- Group G: Overhead, Room
- Group H: Acoustic Guitar, Pedal Steel Guitar
- Group I: Background Vocals
Group Layer 2
- Group J: D, E
- Group K: H, I
Group Layer 3
- Group L: Center
- Group: A, B
- Group M: 50% Side
- Group: C, J
- Group N: 100% Side
- Group: F, G, K
Now we add self side chaining on the first layer. (please do not confuse this with self side chaining in the A or C scopes, we do that for other but partially similar reasons)
Then we add intra side-chaining on the first layer.
Then we add inter side-chaining on the first layer.
Once we have achieved great dynamic relationships, we apply very hard compression on the groups in the first layer after the side chains, on the other layers we add compression medium relaxed on layer 2 (RMS) and relaxed on layer 3 (Peak). What hard compression means here is an aggressive threshold at a medium ratio with lots of make up gain. In other words, we force very high frequency density very early where it is naturally low and has been made extra low due to the side chains.
During the mixing process we gain stage the mix towards a louder than final level, say to -7 integrated LUFS, which is a very loud configuration and use a monitoring solution that makes that easy to achieve without lots of fighting with distortion.
We make it sound the very best it can sound with this type of dynamics. Once that is done, we begin to relax the compression on all of the temporarily hard compressed compressors in the first layer so that the mix now gets a loudness of around -8.5 integrated LUFS.
Once this is done, we again gain stage towards the final loudness at say -7.8 integrated LUFS.
The reason why this works is that because the underlying frequency density at layer 1 is naturally so low and sweet and we begin to relax those compressors, we introduce a very healthy (logarithmic) amount of lower mix density and more sweetness and can do so at a very good isolation level and very precisely within the stereo context.
Since we have so little compression where the frequency density is naturally high, we can now maintain good stereo separation between elements in the mix. Because we dial in the compression at the song's peak frequency density, the sweetness from the dynamics and resonance that we are able to achieve is set as a minimum for the mix as a whole.
A few notes on this technique.
- Focus on maintaining stereo integrity
- Focus on lowering the total mix density for a light soft pleasant final master
- Distribute compression effectively
- Do things in the right order so that we can stay as mild with the processing as possible
- Work with all of the dynamic dimensions as necessary
- Provide a good amount of room and options for gain staging late in the process
Some complementary notes of importance relating to the technique.
- We do all of this aligned to our quality goals for the short and long term dynamics within B groups
- If we have any mix coloration goals, we apply that on the A scope (individual tracks, dial in on C and apply to A)
- Any transient designers or dynamic processors should be applied in between the self side chain and the intra side chain
- Any dynamic mix sound archetypes/sound types/broad stroke multi-band compression/additional saturation should be applied on layer 2 at around 50% of the hard compression on layer 1 so that it falls into its sweet spot when you relax the compression on layer 1 (dial in on C and apply to B2)
- Place the "light" de-essing in front of the processing in the point above
- Have a true peak non-clipping condition on all groups in layer 1
- As an option for stereo enhancing, use multi-band expanders on layer 3
When we work with core processing like volume fader, eq and compression, we can do this at multiple scopes - A) individual track, B) group and C) mix.
The frequency density increases from A to C and from a compression standpoint that is key to be aware of.
Upon applying compression to the mix, we have to have a track routing model that allows for a good compression distribution.
So this means that rather than forcing as much frequencies straight from A to C and apply compression on C, we need to go through B in a clever way.
Furthermore, we need to add the compression in stages from tight to relaxed.
This is done by scaling out B both broadly and deeply and then temporarily add lots of density as early as possible where the density is very low and on top of that dialing this in at the first chorus where the frequency density naturally is very high.
Let me give you an example.
Let's say we have the following sound sources in the mix:
- Vocals
- Background Vocals
- Kick
- Snare
- Hi-Hat
- Percussion
- Cymbals
- Overhead
- Room
- Bass
- Piano
- Organ
- Electric Guitar 1
- Electric Guitar 2
- Acoustic Guitar
- Pedal Steel Guitar
In this case, the B groups/aux routing could be like this:
Group Layer 1
- Group A: Kick, Bass
- Group B: Vocals, Snare
- Group C: Hi-hat, Percussion
- Group D: Piano, Electric Guitar 1
- Group E: Organ, Electric Guitar 2
- Group F: Cymbals
- Group G: Overhead, Room
- Group H: Acoustic Guitar, Pedal Steel Guitar
- Group I: Background Vocals
Group Layer 2
- Group J: D, E
- Group K: H, I
Group Layer 3
- Group L: Center
- Group: A, B
- Group M: 50% Side
- Group: C, J
- Group N: 100% Side
- Group: F, G, K
Now we add self side chaining on the first layer. (please do not confuse this with self side chaining in the A or C scopes, we do that for other but partially similar reasons)
Then we add intra side-chaining on the first layer.
Then we add inter side-chaining on the first layer.
Once we have achieved great dynamic relationships, we apply very hard compression on the groups in the first layer after the side chains, on the other layers we add compression medium relaxed on layer 2 (RMS) and relaxed on layer 3 (Peak). What hard compression means here is an aggressive threshold at a medium ratio with lots of make up gain. In other words, we force very high frequency density very early where it is naturally low and has been made extra low due to the side chains.
During the mixing process we gain stage the mix towards a louder than final level, say to -7 integrated LUFS, which is a very loud configuration and use a monitoring solution that makes that easy to achieve without lots of fighting with distortion.
We make it sound the very best it can sound with this type of dynamics. Once that is done, we begin to relax the compression on all of the temporarily hard compressed compressors in the first layer so that the mix now gets a loudness of around -8.5 integrated LUFS.
Once this is done, we again gain stage towards the final loudness at say -7.8 integrated LUFS.
The reason why this works is that because the underlying frequency density at layer 1 is naturally so low and sweet and we begin to relax those compressors, we introduce a very healthy (logarithmic) amount of lower mix density and more sweetness and can do so at a very good isolation level and very precisely within the stereo context.
Since we have so little compression where the frequency density is naturally high, we can now maintain good stereo separation between elements in the mix. Because we dial in the compression at the song's peak frequency density, the sweetness from the dynamics and resonance that we are able to achieve is set as a minimum for the mix as a whole.
A few notes on this technique.
- Focus on maintaining stereo integrity
- Focus on lowering the total mix density for a light soft pleasant final master
- Distribute compression effectively
- Do things in the right order so that we can stay as mild with the processing as possible
- Work with all of the dynamic dimensions as necessary
- Provide a good amount of room and options for gain staging late in the process
Some complementary notes of importance relating to the technique.
- We do all of this aligned to our quality goals for the short and long term dynamics within B groups
- If we have any mix coloration goals, we apply that on the A scope (individual tracks, dial in on C and apply to A)
- Any transient designers or dynamic processors should be applied in between the self side chain and the intra side chain
- Any dynamic mix sound archetypes/sound types/broad stroke multi-band compression/additional saturation should be applied on layer 2 at around 50% of the hard compression on layer 1 so that it falls into its sweet spot when you relax the compression on layer 1 (dial in on C and apply to B2)
- Place the "light" de-essing in front of the processing in the point above
- Have a true peak non-clipping condition on all groups in layer 1
- As an option for stereo enhancing, use multi-band expanders on layer 3
Last edited:
