Quote:
Originally Posted by ErBird
! Thanks a lot guys for your feedback. I was dismayed to find that the songwriter went with someone else's mix. I thought mine served the song better, but oh well. The guitar tracks were so clipped it's a wonder the mix sounds good at all.
I'm wondering how I could make it less static at the mix level. Other than volume or EQ automation, I can't think of what that would entail. I didn't play any of the parts on this one.
|
Well, almost by definition, "motion" will require automation, unless it's built into the arrangement and orchestration, which in this case it probably isn't. Without having the raw tracks, it's impossible to say exactly how much the recording "breathes," but even some very basic, mild automation work can add a lot of life and pop to a mix. Even gentle 1-2dB volume rides on the band mix in the verses can help separate them from and add apparent energy to the choruses, for example. Another place to look would be to add some dynamics back into the vocal. In this style, it's kind of a "crush it with a comp, then bring it back to life with automation." That'll take a bunch more time, obviously, but you can use heavy compression on the vocal, then automate that track to further level it out on a syllable-by-sylable basis. Then, on the vocal master buss, use another automation lane to add dynamic accent to specific words or phrases. The first puts the vocal into a pocket, the second lifts it out, but controllably and predictably. And yes, EQ automation for that subtle stuff, adding just a bit of bite or "cut" to a specific line here and there works a treat. All very labour-intensive, but it really does put the polish on the static mix. For me, it's not uncommon to have upwards of seven, eight or even more lanes of automation on a lead vocal track.
I would go back over your mix, and start with a "V2," just for forensic purposes, with fresh tracks in a fresh session. I'm not going to get into "pitching and pocketing," for this example, we'll assume "it is what it is." First, phase-align your drum inputs. Pick a "Time Zero" (T0) for the whole kit, and slip-edit all the drum tracks to line up. That will make a more present, coherent drum "sound" right out of the box. I generally use the left overhead as T0, but others use the snare top equally successfully. Keep this same T0 for any other track that may have significant bleed, like guitar or bass if they were tracked live with the drums in the same room. Once you have the track aligned, do a basic, static mix, but without any delays or verbs added, just the raw tracks. Do all your basic bussing, EQ, compression, gating,etc in this stage. At this point, you should have a tonally-balanced, functional mix, with nothing really sounding "odd," or "off." Now, add small verbs (ambience type) to put things in a sonic space if and only if something sounds too "flat" or "in your face," like maybe a guitar, for example. At no point, should anything sound "distant" or "in a big room," we're still working on intimacy, as if the listener was in the same LIVING ROOM with the band. Not an arena, or cathedral. In particular, watch how much "room" you add to the drum mix. If you can hear it, it's probably too much. You should only really notice it if you take it away, by muting the "room" track(s)...
Next, start adding "effect" type reverbs and delays. Think in terms of snare verb, tom verb, vocal verb, vocal throw delay, things like that. One of my favorite delay tricks is to double-buss a track to a return, first send on pins 1+2, second on pins 3+4. Then add your favorite delay (1/4 note ping-pong), then a compressor. Have the compressor track pins 3+4 on the sidechain, and you now have a ducking delay! When the singer is singing (screaming?) the comp slams the delay output into nothingness, but when the singer shuts up, the delay echos "magically" appear and float out the ends of the phrase or line. It takes a bit of work dialing in the comp (VERY long hold and release times!!) but once you have that tuned to the basic tempo of the song, you'll have a pretty automatic phrase throw delay. Very easy to automate the levels on the delay return track to avoid muddying up anything.
At this point, you should have a pretty polished-sounding, clean, clear, present, in-your-face kind of mix going, that sounds decent no matter where you "drop the needle" on the track. What you're shooting for isn't mix perfection, but the ability to hear EVERY instrument clearly, where each has its own distinct place in the mix.
Now, it's time to automate! This will be like building the mix all over again! I like to start with bass and drums on a rock production, so mute everything else. Focus on the kick and bass, and do volume rides wherever they tend to stand out or disappear. Follow that up with the snare, then toms, then finally the metal in the kit. Now your rhythm foundation should be SOLID. The kick is consistent, in your face, and doesn't interact with the bass in any negative way. The rest of the kit sits right in the pocket with the kick and bass. Now, add the first guitar track, and do the same kind of leveling. Then move on to the second guitar, then keys/horns/accordion/banjo/what-have-you. At this point, your mix should be dead-solid all the way through, with each part breathing a bit, allowing the listeners focus to jump from instrument to instrument as their part becomes "important." Rhythm guitar accents should jump to the forefront just a little bit (EQ automation works a treat here), the tom rolls should sing, and the snare should never get overbearing. Just as you have de-conflicted the kick and bass, you should have the same thing going with the two guitars. Almost there! Work in the vocals. First pass, level it all out. Second pass, EQ automation to eliminate any weirdnesses in specific words or phrases or pitches (throat resonances, 1k6-2k2 edge in the vocal break-range, chest rumbles on slides, etc.). The goal is a consistent end delivery of the vocal. It should just sit right on top of the mix, not jumping in and out either in level OR tonal weight. Finally, automate the phrasing to accent the important bits.
Your final step will be to go back over the mix on a VERY high level, and MAYBE push and pull the sections a bit for impact. As I mentioned previously, a little 1-2dB drop in the verse levels in the band mix can let the vocal ease to the forefront and then add weight to the chorus when the level comes back in at the section change. Finally, A-B your original mix against the "V2" mix, and see what the differences were.
All that being said, IF the song is properly arranged and orchestrated, and the performers are on top of their game, all that stuff is pretty much done for you... In this case, unless you spent a TON of time intentionally pulling out all the dynamics, it sounds like a "turn it up to 10 and hit record" session, so you'll have to create all the dynamic motion from scratch. It's a challenge, but you can learn quite a bit about how music "works," and how various frequencies impact different instruments.