This is just the way I personally see the mixing task so take it with a grain of salt as everyone has different vision of technique. Also, everything I explain below is really depending on the type of music (pop, rock, metal, electro). The approach I’m explaining is mainly for pop/rock music. When I’m mixing a song, I'm trying to bring the spotlight on the right elements, at the right place and the right time. I'm trying to make the song 'work' and bring it to the next level. The most important key for me is the vocals and making them sit on top of the rest is priority. I always build my mix from bottom to top (drums -> bass -> guitars -> keys, arrangements -> vocals).
I see the song as two separate elements that need to be glued together and complement each other: the music & the vocals. The music supports the vocals. Most of the time, what people tend to remember when they have a song caught in their head is the vocal melodies or any catchy melodic hook. You need to bring those hooks as evident as possible.
The drums create the rhythm section, the punch, the ambiance. The drums are the foundation of the mix. If the drums are too weak, everything else will suffer and matter how you balance the other instruments, the drums will always sounds lost in the background without any punch. On the other hand, if the drums are too intense for the song, or have too much attack, not enough ambiances, you will have problem fitting them with the rest. The choice of the kick and snare is really important. Try to mix a ballad with a clicky kick and with a piccolo snare and you’ll know what I’m talking about. Of course, if you’re mixing a song you didn’t produce, you can always rely on drum samples. When doing pre-production and tracking, I spend much time on selecting the right drums, proper tuning for the song. It makes a big difference. If you’re mixing a fast song, you need more attack and less ambiance/decay on the drums. If it’s a slower song, you need more bottom, more room/ambiance. Try to use ambiance and decays that suits the tempo the song. The playing is also very important. Inconsistent drummers suck, period. Yes you can augment with samples but the overhead and room mics never lie. You can quantize the drums as much as you want, the drummers in the end is still the one that controls the dynamic of the room and cymbals. In modern music, kick and snares have more and more samples blended with the real tracks and some are simply totally replaced. What makes the drums still sounds real and good are the overheads and room mics. Don’t focuse or simply getting the punchiest kick or snare, try to come up with a drum sounds that complement the song.
The bass instrument for me is a link/bridge between the rhythmic section and the musical section. The bass instrument also creates the low-end/bottom and supports the guitar tone. Dynamic control of the bass is really important. Having a bass that pops in and out of the mix is one of the common problems. Obviously, it starts with a good player but compression and automation, EQ and distortion are your friends.
Guitars, they have to blend with the bass instrument and act as one solid piece tied together that fill the gap between the rhythmic section and the vocals. The guitars need to be balanced volume wise but also EQ wise. You can have the best sounding drum, bass and vocals, but if the guitars sounds harsh or have too much top end, it kills the mix. Another common problem is finding the sweet spot between too dark and too bright without making the guitars sound too muddy or too harsh. Dialing the right amount of gain/distortion is a key here. Also, the relation between the cymbals and the top end of the guitars is really important. Guitars that are too bright or harsh can make you boost too much top on the cymbals and make the mix sound amateur. Same thing happens with the kick and bass. Too much low-end can make the mix sounds too dark and forces you to add unneeded top end and you end up with a mix that sounds too scoop (smiley curve EQ). Once I have the drums, bass and guitars dialed in, I move on the arrangement; keys, synths, strings, etc.
Once the instrumental mix is in pretty much good shape, I bring up the vocals. Depending on the vocals arrangement (doubled lead vocals, back vocals, harmonies, adlibs, etc); I listen to how the vocals interact with the instrumental. Then in my head I hear what type of spatial/ambiance effects will be needed for that particular song. At this point, I know I will need for example a ¼ delay during the chorus or a slap echo during the verse. I start by selecting the right amount of compression for each section of the song and this is achieved by automating the threshold/input of the compressor. By this I mean that the vocals might be more compressed during the choruses than the verse but the volume fader won’t move. The doubled lead vocals add the thickness and volume needed to make the vocals cut during the choruses. When I mix the vocals I listen at very low volume on my NS10 and often switch to mono. If the vocals seems to be at the right volume but I’m still having problem to hear every words, most of the time it means that they need some EQ’ing. Controlling the ‘esses’ is also very important. Sibilant vocals can create the illusion of the vocals popping in and out of the mix and can also make the vocals sound harsh. On the other hand, improper de-essing can lead to lisp issues.
Once the song sounds good, everything is pretty much balanced and sitting at the right place, I start creating movement and dynamic in the song. Obviously, this is based on the type of music. Nowadays, modern rock or metal music has so little dynamic and everything is right in your face through all the song. Some people like, some others don’t. I’m personally only care about making the song work and have the biggest impact on the music listener. But for music that still has dynamic, I try to create a big difference between the verses, choruses and bridge so that it always keeps the attention of the listener. IHMO, the arrangement count for 75% of the dynamic of the song and volume is 25%. If you have tons of lead guitars, keys and back vocals during a verse, it’s harder to make the chorus sounds bigger, even if you turn the master volume up. Having fewer elements during the verses is the key. I often need to automate the volume of the drums between the verses and the choruses though.
I consider my mix done when I press play and I’m able to listen to the song without having any technical concerns. Once you press play and listen to the song in an artistic way or like a music listeners would do, not an engineer, then you have done your job. When you’re able to listen to the song and feel the essence of it, understand the message of the lyrics, when there’s a vibe to it and you can just close your eyes and imagine the band in front of you playing the song in a live room, you’re done. Then it’s a matter of making sure you get the same feeling when listening in your car, with headphones, with cheap iPhone headphones, on your television, etc. If the mix is constant at low volume or high volume and the balance stays the same on different listening system, you’re good and you can move on.
I really hope that can help someone.
LT
