(and a bonus - drum group processing!)
Original Full Mix
Cordelia's version
Ngarjuna's version
Yonyz' version
Panning
I've noticed some of you have already started panning your tracks. For the purpose of this mix, however, the mix needs to be balanced when all the tracks are panned centre.
So, please, pan all your tracks back to the centre. I'll give you a moment to rescue your broken mix.
Ok, now you should have a mix that's well-balanced and maybe with a good vibe too. And it's mono. Understand that it's actually not such a bad thing to leave the panning centre for every track. You can still have a good mix without spreading various instruments across the stereo field.
Starting from mono is not about compatibility with radio/TV/telephone playback. It's not about making sure parts don't disappear when the two channels are summed. It's actually more common for the stereo field to be skewed from people placing their stereo speakers non-optimally, from hifi components being connected incorrectly, or from people sharing earbuds. It's about making sure that even if someone hears only one channel, they still hear a mix that makes sense.
To that end, the four foreground elements - lead vocal, snare, kick and bass - must remain in the centre of the stereo field. The balance between them must not change, no matter what happens to the stereo field. The background elements can be more flexible. You can actually pan these wherever you want, just so long as you remember these three things:
1) No solo. Don't solo any tracks while you're doing this. The reason for this should be fairly obvious - an appropriate panning position must be found in relation to all the other tracks.
2) Keep the stereo field roughly balanced. Try not to make one side heavier than the other. If there are two tracks that are mostly behaving the same way (playing at the same time, same tonal balance, same rhythmic/transient characteristics), then it's pretty safe to pan them roughly opposite each other. You don't, however, have to match them exactly. Depending on the differences between the tracks, and what else is happening in the mix, you might find it more useful to pan one further out and the other closer in.
3) Watch your pan law. Without going into detail, pan law describes the behaviour of a track's volume as it relates to pan position. Different pan laws will mean that the volume might change when you change the panning of a track from centre. Different DAWs have different pan laws, and some let you choose between several pan laws. For this mix, don't worry about choosing the "right" pan law - but just keep it in mind. When you pan you instruments, keep listening and be prepared to adjust the volume fader to keep the mix balanced. Otherwise tracks might become louder or softer when you pan, which will upset the mix. Also keep in mind that regardless of pan law, the PERCEIVED level of a track will change when it's panned too. Broadly speaking, background tracks that are panned far from centre will be more audible than if they are panned centre. This is because there is more energy in the centre of the mix from the four foreground tracks which masks background sounds behind them.
So long as you remember to maintain the lead vocal, snare, kick and bass in the foreground and keep the background in the background, it's pretty hard to mess this up badly.
Drum group processing
In Week 4, I wrote this:
Now is the time to address drum group processing. First, insert a room reverb on your drum group. It's important to use an insert, rather than a send, because we'll be inserting a compressor after the reverb. Usually I find I get best results by using a medium sized room, with about 10%-20% dry/wet mix. An open top end (ie, not a dark reverb) helps to add air to the kit. Sometimes the reverb can make the low end of the kick too fluffy and indistinct. To address this, either use a low cut filter or reduce the reverb time for low frequencies (sometimes just collapsing the low frequencies to mono can be enough). How you approach this depends entirely on the reverb you're using - how it sounds, and what features it has. This kind of adjusting is why I prefer an algorithmic reverb (particularly one that's very adjustable) rather than convolution - because even if I start with a preset I'll still have to tailor it to the mix. The overall goal is to make the drum kit sound like it's in a room without making the sound too diffuse and reducing the impact of the drums. Err on the side of subtlety.A couple of months ago I wrote: I usually route all the drum kit elements to a group bus, and apply some subtle processing to the kit as a whole. On Flesh And Bone, I added some subtle (stereo) room reverb and compression. The reverb adds a little air around the drum kit - without it the drums sound a little like they're being played in a vacuum. It helps the kit sit better in the mix, making it less obvious that they're samples. The compression adds just a little shaping and "glue" to bring the kit together a bit. Usually I wait until the mix is almost complete before doing this though - it's very difficult to judge the correct amount before the rest of the tracks are in place. Reverb settings are particularly troublesome - sometimes a longer bright room is appropriate, sometimes a short dark "deep" room sounds better, but I never know this until close to the end. Often the right reverb will sound quite wrong when the kit is played in isolation, but it does the right thing in the rest of the mix. What I'm saying is - send the drum kit to a group now, not don't add any plugins to that group. Just keep it in the back of your head for now.
Once you're got a suitable room sound, insert a compressor after the reverb. Again, aim for subtlety. The compressor is to provide some slight weight and glue to the kit. Start with a fast attack and release time, highest ratio (yes!) and high threshold with no knee (so at first, the compressor is doing nothing). Now, slowly lower the threshold until you can just hear the compressor. Don't go by what the meters say - use your ears. If the transients are too squashed, lengthen the attack time until you can hear some "snap". Now lengthen the release time until you can hear a slight pumping effect on the drum tails and the room sound. The trick is to use the compressor to shape the drums without further reducing the dynamic range. This is why a high threshold is so important. If the threshold is too low, the drums will be crushed, and you'll lose the subtlety in the sound and unbalance the rest of the mix.
To better illustrate my point, I've uploaded three files to the Week8 folder:
Raw drum group
Drum group with room
Drum group with room and compression
They're 24-bit files, but only 2.3MB each. It's not the whole song - just a few bars to give you a taste for how much (or little!) processing to apply to the drum group.
Reverb
I'm going to suggest you do something that I didn't do in my mix.
When I originally mixed this song, I used the process I wrote about in this post:
Recently though, I've been using a plate reverb on vocals as well as a hall. It can be a subtle difference, but generally a hall reverb will add a sense of depth and distance, whereas a plate reverb will "stick" to the sound and simply make it thicker and fluffier. I find that using a plate on vocals puts a "sheen" over them - stopping them from sounding too naked and popping out of the mix. Blending in a hall algorithm adds the sense of depth. Usually I just use two send channels - with the plate on one and the hall on another. I find that taking this approach also allows me to get away with using less reverb overall because I'm not using the hall to cover up the nakedness in the voice.A year and a half ago I wrote: If you've already got any reverb anywhere in your mix, remove it (unless it's a special effect). Add the best hall reverb you have to a send bus, and mute everything except the lead vocal. Add reverb to the vocal, and tweak the advanced parameters to get it to sit right. Sometimes the "Rich Vocal Hall" preset will be perfect, sometimes you'll tweak for half an hour. Don't be lazy. Keep going until you can honestly tell yourself "That's the sound". Read the manual if you have to. Now, set the level of the reverb for the vocal so that it's generous, but not overpowering. Next add a slight amount of reverb to each instrument. Again, remember that you're adding to the sound, not changing it. Give each instrument some depth and space, but don't (for example) transform "guitar" into "reverbed guitar". A TINY amount of reverb on kick and bass will keep them from popping out of the mix.
I've also written about plate reverbs here.
<- Week 7 Week 9 ->
-Kim.