OK, so in this pro mixing tutorial I want to share a few really simple pro techniques that instantly just make any mix simply better sounding. This is stuff many of you might be aware of but it can help sometimes to highlight things like this and give it some attention. Of course you don't have to do any of this if you don't want to, do what works for you and create the character that is you, I'm just giving some hints that generally speaking I've found these things help...
Optimize the side component before narrow stroke compression, before broad stroke compression and after broad stroke compression
This technique is all about maintaining an optimal signal on the mix side component when applying the dynamics processing. A way of understanding this is that you want to ensure optimal gain levels on the side component so that the side is as full and sweet as it possibly can be. Full in this case means that it has very high quality attack, modulation and decay. This in turn means that the mix feels and sounds nice.
Lower noisy sound sources in volume dramatically early on
The thing is that when you apply the dynamics processing and gain staging you are to some degree going to reduce the volume of the individual tracks in order to balance the mix towards the target loudness level and the louder the target loudness level you go for the sooner the noisy sound sources are going to start to add more to the perception of the mix, meaning more of the noise is leaking into the experience of the mix. This issue is not as dramatic when you go for a lot less loud mixes, say something like -13 or -14 LUFS in the chorus, but those mixes on the other hand tend to lack vibe and vitality at normal playback levels in certain genres comparatively speaking. Also, with the right type of compression you can dramatically reduce the issue, but the main point is that the louder you make the mix, the more likely it becomes that you are underrating the loudness level of the noisy sound sources within the mix. The noise just leaks into the perception and you become a bit more aware of it with fresh ears, but it is when you hit the mute button when you go: wow. Typically this noise can stem from various drum, vocals, guitar and fx tracks. So watch out for that, remove it almost entirely and then at the end bring it in as icing on the cake.
Go for more harmonics in most genres
I see this all the time, engineers go for 90% groove and 10% harmonics. A mix cannot become emotional until you get enough of the pad element to sing out enough of the harmonics of the chord progressions. This is why you should have a volume fader for rhythm and a volume fader for pad, so that you can easily balance the ratio between rhythm and pad. Usually what is the problem is that the pad element is assign a lot of signal but very little work has been done on the harmonics, then on top all in all the combination of the input track volume faders have much too much volume level reduction, this then becomes muddy because the modulation becomes so thin and the sound sources sit more towards the null point of the stereo 3D sound field and hence the combination makes them harder to separate in the stereo field, hence the experience of the mud. So think about having enough volume on all tracks, enough volume of the pad element relative to the rhythm element and enough harmonic excitement on the pad element.
Level dominant peaks with multiband compression
It is nice to have a lot of natural dominant peaks in all sound sources, after all that is what is their natural state when played in certain ways and in certain contexts. But the issue is that in many cases a lot of the dominant frequencies are also to some degree the result of some amount of frequency build up/resonance at those frequencies, relating to the acoustics of the recording space, the dynamics of the playing and other things. What this does is that it makes it more difficult to end up with a balanced sounding mix that translates well because you have a lot of loud pings at various frequencies popping up at various locations in the mix. This is what multiband compression is good at dealing with. So I find that during the initial stages of the dynamics processing you should sweep out the dominant frequency "pings" with multiband compression in solo on each track as needed and soften those pings a little. I usually set the monitoring loudness slightly above the average playback level when I do this, in this way I can grab just enough of those pings to level them out and also then mess with the signal little enough. Do this in zero-phase mode. Avoid this on very broad stroke, it tends to reduce the quality of the stereo image.
Start with slow compressor attack, then tighten
The issue with very fast compressor attack times is that they attack the initial portions of the attack very aggressively which causes a "snap" or "thickening" of the transients especially when you push the compressor. Although this can help define sound sources in the mix and make them more distinct it also makes you do wrong gain staging and dynamics processing moves due to those fast attacks. This is something that can really damage a mix very quickly and in the background. To avoid this I find it is way more effective to start with slow attack times to keep enough of the peak softness/thinness and modulation qualities present in the mix, then tighten up selectively as needed by making the attack times faster without doing additional gain staging moves. Fasten the attack times fairly late in the process when your mix is already very near the final target loudness level. It can be a very good alternative to gain staging or high low frequency gaining. But keep in mind that it will add some amount of tightening and when that tightening becomes too broad or heavy, it will take away a lot of the beauty of the mix.
Let drums slam
Drums that slam and are noisy, that's awful. But drums that are really soft and sweet and slam, that gives the mix a really good amount of vitality, especially from the cymbals. Slam in this case means that the transients of the drums are "free" among the other transients within the mix. If you let those drums into tons of broad stroke compression, much of that natural drum vitality present in the slam is reduced. What this means is that when you have groups of tracks for center and side, ensure that the drum tracks are in their own slots/groups and that you have very minimal or no compression on those group tracks. Then on the more narrow stroke compression ensure that all drums duck all other transients in the mix, that's a way of ensuring the drums slam in the mix. Now, because the drums cut through the mix, you don't have to tighten them up on the individual track level as much either, meaning you can have slower attack times and still have enough bite, but now they are relaxed too. Can sound awesome.
Add gain at 200Hz and 2kHz on the monitoring path
You will always to some degree compensate for something through the moves you make. The important thing is to understand the difference between positive compensation and negative compensation. When you understand that, you can take advantage of positive compensation and end up with much better final results than if you did not do so (positive compensation is essentially the result of great monitoring, you get a more relaxed sound). By expanding at 200 Hz and 2 kHz on the monitoring path you will amplify frequencies that often will add harsh attack to the mix, in this way you positively compensate those frequencies and can avoid some of the tightness in those frequency areas that comes with compression. I usually combine this with other stuff I have on the monitoring path.
Optimize the side component before narrow stroke compression, before broad stroke compression and after broad stroke compression
This technique is all about maintaining an optimal signal on the mix side component when applying the dynamics processing. A way of understanding this is that you want to ensure optimal gain levels on the side component so that the side is as full and sweet as it possibly can be. Full in this case means that it has very high quality attack, modulation and decay. This in turn means that the mix feels and sounds nice.
Lower noisy sound sources in volume dramatically early on
The thing is that when you apply the dynamics processing and gain staging you are to some degree going to reduce the volume of the individual tracks in order to balance the mix towards the target loudness level and the louder the target loudness level you go for the sooner the noisy sound sources are going to start to add more to the perception of the mix, meaning more of the noise is leaking into the experience of the mix. This issue is not as dramatic when you go for a lot less loud mixes, say something like -13 or -14 LUFS in the chorus, but those mixes on the other hand tend to lack vibe and vitality at normal playback levels in certain genres comparatively speaking. Also, with the right type of compression you can dramatically reduce the issue, but the main point is that the louder you make the mix, the more likely it becomes that you are underrating the loudness level of the noisy sound sources within the mix. The noise just leaks into the perception and you become a bit more aware of it with fresh ears, but it is when you hit the mute button when you go: wow. Typically this noise can stem from various drum, vocals, guitar and fx tracks. So watch out for that, remove it almost entirely and then at the end bring it in as icing on the cake.
Go for more harmonics in most genres
I see this all the time, engineers go for 90% groove and 10% harmonics. A mix cannot become emotional until you get enough of the pad element to sing out enough of the harmonics of the chord progressions. This is why you should have a volume fader for rhythm and a volume fader for pad, so that you can easily balance the ratio between rhythm and pad. Usually what is the problem is that the pad element is assign a lot of signal but very little work has been done on the harmonics, then on top all in all the combination of the input track volume faders have much too much volume level reduction, this then becomes muddy because the modulation becomes so thin and the sound sources sit more towards the null point of the stereo 3D sound field and hence the combination makes them harder to separate in the stereo field, hence the experience of the mud. So think about having enough volume on all tracks, enough volume of the pad element relative to the rhythm element and enough harmonic excitement on the pad element.
Level dominant peaks with multiband compression
It is nice to have a lot of natural dominant peaks in all sound sources, after all that is what is their natural state when played in certain ways and in certain contexts. But the issue is that in many cases a lot of the dominant frequencies are also to some degree the result of some amount of frequency build up/resonance at those frequencies, relating to the acoustics of the recording space, the dynamics of the playing and other things. What this does is that it makes it more difficult to end up with a balanced sounding mix that translates well because you have a lot of loud pings at various frequencies popping up at various locations in the mix. This is what multiband compression is good at dealing with. So I find that during the initial stages of the dynamics processing you should sweep out the dominant frequency "pings" with multiband compression in solo on each track as needed and soften those pings a little. I usually set the monitoring loudness slightly above the average playback level when I do this, in this way I can grab just enough of those pings to level them out and also then mess with the signal little enough. Do this in zero-phase mode. Avoid this on very broad stroke, it tends to reduce the quality of the stereo image.
Start with slow compressor attack, then tighten
The issue with very fast compressor attack times is that they attack the initial portions of the attack very aggressively which causes a "snap" or "thickening" of the transients especially when you push the compressor. Although this can help define sound sources in the mix and make them more distinct it also makes you do wrong gain staging and dynamics processing moves due to those fast attacks. This is something that can really damage a mix very quickly and in the background. To avoid this I find it is way more effective to start with slow attack times to keep enough of the peak softness/thinness and modulation qualities present in the mix, then tighten up selectively as needed by making the attack times faster without doing additional gain staging moves. Fasten the attack times fairly late in the process when your mix is already very near the final target loudness level. It can be a very good alternative to gain staging or high low frequency gaining. But keep in mind that it will add some amount of tightening and when that tightening becomes too broad or heavy, it will take away a lot of the beauty of the mix.
Let drums slam
Drums that slam and are noisy, that's awful. But drums that are really soft and sweet and slam, that gives the mix a really good amount of vitality, especially from the cymbals. Slam in this case means that the transients of the drums are "free" among the other transients within the mix. If you let those drums into tons of broad stroke compression, much of that natural drum vitality present in the slam is reduced. What this means is that when you have groups of tracks for center and side, ensure that the drum tracks are in their own slots/groups and that you have very minimal or no compression on those group tracks. Then on the more narrow stroke compression ensure that all drums duck all other transients in the mix, that's a way of ensuring the drums slam in the mix. Now, because the drums cut through the mix, you don't have to tighten them up on the individual track level as much either, meaning you can have slower attack times and still have enough bite, but now they are relaxed too. Can sound awesome.
Add gain at 200Hz and 2kHz on the monitoring path
You will always to some degree compensate for something through the moves you make. The important thing is to understand the difference between positive compensation and negative compensation. When you understand that, you can take advantage of positive compensation and end up with much better final results than if you did not do so (positive compensation is essentially the result of great monitoring, you get a more relaxed sound). By expanding at 200 Hz and 2 kHz on the monitoring path you will amplify frequencies that often will add harsh attack to the mix, in this way you positively compensate those frequencies and can avoid some of the tightness in those frequency areas that comes with compression. I usually combine this with other stuff I have on the monitoring path.
Last edited: