"Multing" the mix

A description of an unusual mix technique in response to a post from another thread:

ORIGINAL: jayhill

Yep, when you get a chance sometime, if you would consider expanding on this paragraph. There's a lot of information in here and while we do understand what you're getting at in general, taking it to the next level of detail (on just a little more of the methods you're using to achieve those 4 cloned tracks for vocals) would be much appreciated.

Regards,

Jeff

First off, I don't have any "4 clones for vocals" rule per se, I'm just making generalizations.

My absolute favorite thing about computer recording is unlimited tracks and effects. "Multing tracks," as I call it, is something I started to get into even before I started with computer-based multitracking, and when I got the box, I hit the ground running with my favorite effect-- the compressor.

There is an old vocal trick called "Motown style" compression (I think) that was developed to solve the dilemma of keeping those soulful, dynamic, lively vocals without squashing the life out of them, but still getting some kind of consistency so the lyrics would be intelligible in a dense mix on radio broadcast.

Basically, you take the vocal track and clone it, and then you apply severe bandwidth-limited compression to the cloned track, zeroing in on the upper presence range. You basically end up with a cloned track that consists solely of a very up-front, present, and clear signal that only contains presence range so you can hear the words and another track that has the whole vocal performance in all it's glorious dynamic motion and power. Mix the two together and you have the best of both worlds-- the musicality of the singer's performance is all there, and so is the clarity and impact that we expect from pop music.

I basically took to this style of processing like a fish takes to water when I first stumbled across it, and I started using variations on it on lots of things, but this was back in the days of limited track counts and finite numbers of effects boxes and so on. With computers, I have gone absolutely hog-wild with it and may end up with 80 tracks and 15 aux busses just on a simple rock combo with a couple guitars, bass, vox, and drums.

When I first start a mix, the first thing I'm going to do is pick through it, casually, taking mental notes, soloing various instruments, sweeping eqs around, looking for stuff I like, stuff I don't, and most of all, for useful information. For instance, where is the "thump" of the kick drum? Is it at 70Hz? 100? 53? How wide or narrow a boost can I get on the thump before it starts to turn muddy or boxy or ringy? How steep of a boost can I get before it starts to sound fake or distracting? Where is the "click" of the kick? What happens to the sound if I get rid of it? If I boost it? These kinds of things-- I look for the nasalness, the trashiness, the boxyness, the tubbiness, the muscle, the mud, the air, the attack, the bleed, the sweetness, the tangyness, the decay, all of that. This typically occupies about 10-15 minutes, and I'm paying particular attention to the kick, the snare (finding the "crack" and "boom" are critical) and the lead vocal (where I want to identify a role for almost every single frequency band). I will also seek out anything particularly cool or annoying in any instrument. I am usually going to end this with a whole bunch of eqs with little bumps and cuts, mostly just to remind me later of where the different elements are and how wide they go and what I liked and didn't on first listen. Nothing is set in stone, or even sketched in pencil yet-- just taking notes, conducting preliminary studies before the excavation. Laying out your parts and tools, right?

As I start to do this, I begin "multing" the things that really jump out at me as something that I know I want to have clear and precise control over (actually, I'm usually multing at the tracking stage, but I'll get to that later). One example that usually happens very early in the process is the bass. Usually, I have two raw tracks of bass-- one from the amp mic and one from a direct in through a DI box or Y cable. The miked sound is going to be my basic "instrument sound" in the mix. It is usually going to have a very broad, shallow low cut right up into the mids, and a steeper rolloff (usually done at tracking) to take out most of the stuff below 40-80Hz or so, depending on the situation. It is going to be primarily a midrange sound designed to complement the other rhythm instruments and anchor them to the lows.

My low end is going to come from a clone of the DI track, usually (this is all made up as we go), with a steep low-pass filter and absolutely ruthless compression and gating to achieve exactly the low end envelope shape that I want for this track-- If I want a slamming, dancehall, head banging low end, then I want that waveform to look like a brick with a transient, and I want this to be just a solid, hard, steady, low-end pulse. If I want round, mellow, full swells of power behind a ballad then I am going to remove any trace of attack. I am basically treating this like a sound designer would create a synth bass sound, using my dynamics controllers to create the envelope shape and eq and maybe phasing or modulation to sweeten or manipulate the sound as desired (this is early stages, very early, and everything is subject to change). Once I knock that into rough shape I'm going to bus it to a "low end" bus along with some kick for later processing of the two together. The kick sent to this bus is often also going to be a clone with mostly just the "thump," again, manipulated appropriately to the vibe of the song.

Next, I am going to look for anything particularly interesting or useful in the bass sound-- is there a cool stringy pick attack, or some nice fingering or growl from a fretless or a neat burpy aspect or clakety fizz that I like? This may come from either the DI or the miked sound, but I am especially interested in any useful presence-range element that can be used to add definition, clarity or attack is I need it later. If I find it, you guessed it-- time for some stem-cell research. A-cloning I go and pull some of that motown sh!t on this thing-- strip out the body and just keep those snarly nasties and compress the hell out of them to get 'em right out in front. Then take this and my "basic" sound and send them to my second bass bus- "bass."

So now I got six faders for bass: 1. The original "miked" sound-my basic bass; 2. The synthy lows 3.The "low end" bus with some kick 4. The "bass bite" clone 5. The basic "bass" bus 6. The original DI track (just muted for now-- I might need it later). The relative levels of everything don't much matter now-- they're gonna change later. I'm just in the advanced stages of "laying out my tools." I have given these tracks colorful names that reflect my assesment of what role they may be called upon to play. I am also appending my text file, begun whilst tracking, to note what tracks are where and what they are for an what effects and settings (in rough numbers) I am applying. Total time "multing" the bass-- about the time it took you to read this.

Now what on God’s green earth am I gonna do with six bass faders, you ask? I’ll tell ye. When it comes time to actually mix, I’m basically gonna stick to my busses, and treat the two things-- “lows” and “bass” as separate elements or instruments. One is sound of the INSTRUMENT-- the sound of a four-string guitar detuned an octave-- that is going to be where the bass player’s artistry and expressiveness are showcased. The other, the “lows” fader, is usually going to be pretty low in the mix-- it is the ROLE of the bass, the tonic foundation, as the BASE, as the GROUND FLOOR. It can also be used as an instant “HELL YEAH” button when needed to get the speakers pumping. The ratios of the two may change during the song, and there may be additional processing applied to each.

The individual “bite” track may be bumped up a dB or three during bass solos or cool licks to really push the bass out front, or pulled back during mellower parts. On occaision I have even flipped the phase of it to take the “bite” OUT of the bass, but it was weird.

Have you ever seen an orchestra?

Do you see what I am doing?

I am breaking the “instruments” down into “elements”-- those are what I care about. When the band is going full throttle and drummer rips out some wild fill and I want every hit to feel like punch in the chest, I want the PUNCH, the IMPACT, the ATTACK to knock you on your ass. If I just reach for the drums bus and turn it up 6dB then I need to either pull everything else down for that section (not good) or turn down the whole song 6dB (even worse) or run a limiter across the main bus (Heaven forefend!) BUT, if I have a nifty little slider that ONLY turns up a few narrow ranges on each of the drums that contain the IMPACT elements, then I can go hog wild and hardly make a blip on the RMS levels while still bashing your goddam head in with the tom fill.

And that is exactly what I want to do. I want to BASH YOUR GODDAM HEAD IN while still maintaining a neatly conservative range of RMS variation suitable for broadcast and keep lively and interesting musical dynamics.

The most important instruments in the mix are usually, in descending order of importance: The vocals, the snare, the kick drum, and the bass. These are the structure and essence of the song. The rest of the stuff is the flavor, the personality, the stuffing and gravy (isn’t that always better than the turkey? But you’d never just eat stuffing and gravy, right? See where I’m going?)

If you can get a good mix going with just the bass, the backbeat, and the vocals, then you have built the framework of a house and it’s time to call in the queer eye for the mix guy crew to fill it out and make it home for some hip little band that needs your help to save them from looking like the dopey mouth-breathers they are.

Here are your elements, broken down by instrument:

Organ: You can do anything with this-- it is the answer to the fat “lows” portion of the bass-- it fills out the gap between chords and melody and you should feel free to go absolutely nuts with whatever kind of processing you want on all the different frequency ranges it occupies to fill out the sound in the places where you want it to. Saturate the hell out of it and run it through a delay and flanger and a Leslie emulator if you feel like it. Do it when the organ player’s not around and he’ll thank you later if you do it right. Gating combined with medium-attack, medium-release, heavy compression adds attack that you never thought an organ could have. Turn the envelope around and you have some awesome wailing pad action that puts a smile and a guitar player-style “bad smell” happy wince on your audience’s face like nothing else.

Piano: As a rhythm instrument in a dense mix- Shoehorn that box of strings into it’s proper role as a tonal percussion instrument. Left hand=some mean bass content. Right hand=what the hi-hat can’t do and the guitar won’t do and what the singer wishes he could do. Work it, baby. Find the plunk and put it to work. Make the resonance give it up for the people. As a lead-- Just clean it up and let it do it’s thing. Don’t mess with the piano if they player’s got it right, just make up for the mic and the medium and let it go.

Background vocals: You know what to do. It might hurt, but it has to be done. They are called “background” for a reason.

Acoustic Guitar: In a dense mix, think “shaker”-- no lows, heavily compressed attacks, no sustain. In a lead role, this highly expressive instrument has four elements-- the string noise/finger sound, the midrange resonance, the low-end depth, and the attack of the pick/finger. These are separate instruments. Get rid of the ones that weren’t invited and treat the others as individual components of your mix.

Electric guitar: what’s to say? The guitar player might think that he is playing chords and notes, but he’s really playing sounds. Whispers, roars, howls, creaks, screams, machine-guns, chimes… the great thing is, they’re all loud, and they’re all full-frequency. Which means that this is your playground. If you tracked it right, you can do no wrong. Take out anything, anything at all that you don’t like or that steps on other’s toes. It takes a licking and keeps on ticking. Do this last. A good guitar sound fits wherever you put it and still rocks the house ten ways till Sunday.

Get the idea? This is not a conventional approach to mixing, it is an exaggerated version of a certain kind of special technique. Would I use this for a jazz ensemble? Probably not. For a chamber orchestra? Definitely not. But for either of these, I wouldn’t be doing separate, one-at-a-time, mono tracks to a click.

More later?

Maybe.

Cheers.

this should be a sticky.......

Great post Yep!

ORIGINAL: jeffers_mz
On step one, are you soloing tracks more often, listening to a full mix more often, or somewhere between most often?

both, and flipping back and forth a lot, too. Also a lot of listening to the mix real quiet with individual tracks turned up louder through an aux bus, sort of like a solo "fader." I'm basically going back and forth, listening to each track, evaluating what I might need or want it to do and what I don't, and checking to see what happens to it in the mix.

I might not split the vocal track at all, or I might split it into several different components-- the throatiness, the articulation, the "lower presence" proximity-effect frequencies, the air and breath sounds, the musical tonic of the melody, the wet, textural "mouthiness," the rasp or gravel, the organic resonance of the singer's natural vibrato-- these are all elements that I may want to play up or play down at different parts of the song in different ways. Some of them can be easily manipulated on a single track with changes in eq or bandwidth-limited compression or whatever, and some of them are ones that maybe I feel like I need to pull out of the track and handle seperately-- this is particularly the case with certain kinds of details that are a really cool part of the sound but that get lost in either the dynamics or the ambience that I also want to achieve.

The solo button is really useful to see what you're losing when you drop the sound back into the mix-- You know when you have a guitar sound that sounds awesom by itself, but in the mix it just turns into nasal fizz? Well, the first thing to do is to get rid of that fizz and the second thing to do is to figure out what elements of the sound you DO really want to keep and how you are going to fit them in the mix. A couple of simple, quick eq rips might be all that you need, but you might find that being able to treat the pick attack or the "chug" seperately opens up new possibilities.

The whole thing is to play with it. Instead of thinking of the tracks as prefab puzzle pieces, think of them as raw materials that you can cut down, weld together, throw away, bend, stretch, shrink, pulverize, whatever.

Cheers.

Thanks Yep. I might start a new thread. But I think my questions already answered. Like most "sounds" just play to you get it right and keep notes. But you can not blame me for seeking the secret of things, can you..........................JDW

That's some good stuff yep - nice of you to share it. What crossovers do you use for your band-limited tracks? I've been using Wide Boys FreEQ for really steep crossovers (> 60db/octave) which isolated the cloned bands more - I'm not sure if I like it better than gentler slopes yet, it depends what I need I think... just curious.

FreEQ: http://www.kvraudio.com/get/986.html

Someone said something about the SUN sound - IMO a dirty slapback echo can go a long way. There's an analogy sounding slap back delay in this suite from Voxengo:
http://www.voxengo.com/product/analogflux/

You might be able to get some of the same feel if you have the old Cakewalk Tapesim hangin around then throw a digital delay (slap settings < 60ms I think) either before or after it. Probably across the master buss. I visited Sun last fall - a site to behold (historically)...

Very valuable info. Much appreciated!

ORIGINAL: jeffers_mz

BTW, do you have the exact settings for the bad smell sneer effect you mentioned for guitar solos, or is it more of an application specific kind of thing?

Hmm...
I think I actually meant that a good organ sound could put a look on the face of a listener like guitar players have when soloing...

As far as guitar solos go, here are some of the things I look for, roughly (when it comes to leads, organs and organ-like synth sounds are pretty similar to distorted guitar, incidently):

The octave roughly between 40 and 80 Hz is going to be where the big volume swings happen on heavy-metal chugs or palm-muted punk rock chords etc. Take them out, and you can get a lot more loud before clipping, but at the cost of losing your power and making your guitar sound like a casio keyboard. Leave them in, and where does the bass live? Trying to control those lows is gonna be a nightmare.

From 80Hz up you have two or three octaves leading into the dreaded midrange. Very murky waters. This is where the meat and body of the sound are going to live. You have to decide how much you want and WHERE YOU WANT IT. The overdriven guitar is like a swarm of locusts that will destroy everything in its path. Put it onstage in a controlled environment in front of screaming teenagers and its power is awesome to behold. Allow it to run riot over everything and and it's just depressing, annoying, and financially ruinous. The range from about 400-1k is particulary challenging for beginners, because it can be tough to hear what's going on down there and how your changes are affecting things, but it makes a huge difference in how your mix is going to fit together. It is the lower boundary of the midrange and decisions about it cannot be made in isolation. This range overlaps every other instrument in a rock band and getting it wrong will leave your mix sounding lurchy and fizzy and honky and disjointed and quiet and lame. Get it right and everything glues together and gets your motor running even when listening from the next room with the door closed.

Just above that it another octave that is much easier to hear but harder to know what to do with, from a little below 1k to just above 4k. Resist the temptation to just get rid of it or to pretend that it isn't there. This is the guitar's presence range, fortunately just under the singer's, with a lot more room to work. This is going to be where most of your pick attack happens-- maybe a little lower or a little higher. This is critical. It is different for every guitar, amp, pick, string guage, and player. There are at least two and sometimes three different components to the pick sound-- The first is the "guh" sound, lower in the range. Make a sound like a hard G. Do it. Feel that little lump in your throat? That's the low part of the pick attack. That's what you look for. The high part of the pick attack sounds like "chp." Now, sweep a little eq bump around until you find them both. Found them? Good. Now do you want them? Tough question to answer. Maybe one, maybe both, maybe neither, maybe lots, maybe little. It matters. Pray that you got it right at the tracking stage.

Right above the pick attack is about two octaves of stuff that is going to play a vastly different role in most leads than it plays in the rythm section. Fizzy, present, squealy, noisy, weak, aggressive, raspy, annoying, screechy-- this is where your leads happen. If you don't want people to say bad words about your leads than get it right.

Finally, the home stretch. Once you cross the border north of 8kHz, there's not usually much up there that you need, and you can open up a lot of space for the instruments that need it. Cymbals, vocals, acoustic instruments, room mics, synth pads and the like will all make much better use of the highs than an electric guitar usually will. Which is not to say that you should just kill it mercilessly-- just look for anything particualrly interesting or annoying, and if it's not there, try rolling off the highs a litle and see if hurts.

Of course, all of the above totally depends on the guitar, the key of the song, the player, the pick, the amp, the strings, the mic, and so on. And once you know what you like and what to look for, you can often get most (or at least half) of the way there at the tracking stage.

Cheers.

Hmm...
I think I actually meant that a good organ sound could put a look on the face of a listener like guitar players have when soloing...

I know, but bad smell sneer deserved recognition.

Good stuff on the guitar spectrum. Reckon I can fit a power spike for guitar in between kick and bass at say, 80 to 125?

I know, I know, try it and see...that's what unlimited tracks are all about yes?

ORIGINAL: name1432

yep, i'm new to the techniques forum but already know your posts are well worth the read; thank you for them. If I could hear audio examples of your techniques it would help me as I try your advice on my own mixes; if I've missed links to any of your audio productions, would you mind directing me to them?

some terms, like "trashiness" and "boxyness" fly over my head, which has little experience mixing. Even still, I'm understanding enough so that what I read is very helpful; thanks

I don't have any music on the interweb. I have some customers who do, but you can also hear what I'm talking about on the radio. Maybe one of these days I'll put together some before and after examples.

As far as terms like "boxiness," that's not a technical term, it's just a sort of aesthetic description of what might be a dull resonance in the lower mids that happens to sound to my ears like a cardboard box sort of sound-- not usually something I want, but never say never, you know...

You might want a different sound profile than I do, you might hear things in different ways. My purpose was not to tell you this frequency is bad or this one if good or your guitar should or shouldn't sound nasal or murky or boxy or edgy, just to describe the way I think about all the different components that make up the sound.

Cheers.

Within the last couple months, Micheal Japan wrote a classic reference piece on some of the slang terms and where they live in the EQ spectrum.

It is in this forum, (Techniques) and I believe it lives about halfway down in a thread about school programs for mastering engineers.

It's here, and not buried too deep, well worth the read.

ORIGINAL: name1432

yep, thanks for helping me with that term, and I do get what you're saying about your analysis and sythesis technique.

I got the book-with-audio-examples idea while reading bob katz's book and not being able to hear in my mind what he was writing about.. but I understand you better than I did him. cheers

Thanks for the compliment. I am pretty sure that that is the first time my work has ever been favorably compared to that of Bob Katz.

FWIW, "Mastering Audio" by Bob Katz is an outstanding book, and is rapidly becoming a standard-issue classic, along with Yamaha's "Sound Reinforcement Handbook" and Huber and Runstein's "Modern Recording Techniques." The three of those put together will tell you pretty much everything you need to know about anything audio, although most people probably won't get most of it without a lot of practice and re-reading.

Bob Katz is a mastering engineer, and an audiophile. Both good things to be sure, but most record producers and audio engineers that I've met are not even close to being half as obsessive or fidelity-minded as Mr. Katz, whose most acclaimed mixes are not even mixed, per se, just exceptionally high-quality audiophile stereo recordings of live music. His book sets the bar-- it does a phenomenal job of showing in broad strokes how to make records for the most discriminating listeners.


Sometimes, though, our listeners are not the most discriminating, and in the rough-and-tumble, low-paid, highly-skilled world of popular music recording, absolute fidelity is often the furthest thing from anyone's mind.

Cheers.

The 'motown vocal' technique sounds cool. Would it be the best way of solving this situation?

I've been trying on one song to give my vocals a sort of very ambient yet nearly lifeless sound (emotively it has to sound sort of dead/lost and a bit angsty). So I compressed the living daylight out of them, Eq'd the lower end out like there was no tomorrow, and then chucked on some reverb, tailing it off really quickly (I started with the Sonitus Reverb on the "short and sweet" preset and tweaked it a little). I'm guessing some of you are pulling your hair out at this thought, but I don't really have a clue what I'm doing so I guessed.

The track was a rough “get the ideas down” version to test the waters and see if it was going anywhere. It consisted of my vocal, a gentle arpeggio-style guitar melody and an acoustic rhythm. As far as the idea of the song, it got a great response – I played it to my g/fs parents as her dad shares a similar music taste to me, and they both had tears in their eyes by the time it finished. But her dad later said that whilst it was a really emotive song, and the “thing I did to the voice” was really haunting, there were parts where he found it difficult to follow what I was actually singing.

He’s since hijacked the CD and has played it to everyone he knows, including a guy in his gym who is a producer (and has a friend who was involved in the making of U2’s Joshua Tree album). The producer guy gave him his business card for me to phone him, so he could find out what I intend to do career-wise so he could point me in the right direction. When I spoke to him he told me that it was a good song that had a real sense of raw emotion.

So my problem now is with this great reception I got from the poor quality demo, I’m going to have to jump up a hundred rungs of the ladder as far as my ‘producer’ abilities go because, quite frankly, sitting me in my bedroom ‘project studio’ (which has left me in a lot of debt) has the same comparative ability of putting a chimp in a space shuttle and getting him to fly around the moon with no pre-programmed instructions. Basically I’m afraid I will ruin a good song with both over-processing, or ‘incorrect’ techniques.

So I’m trying to find ways of keeping the same vibe from the vocal effects, but making them clear and crisp. The singing style, I’ve been told, is similar to Thom Yorke (Radiohead) on the Fake Plastic Trees or Street Spirit tracks on their album The Bends. So there is an element of ‘lazy vocal’ style in blending words together, which is giving me a major headache as far a reverb goes, but works really well for the song in general (and is also kinda the way I sing anyway).

The three of those put together will tell you pretty much everything you need to know about anything audio, although most people probably won't get most of it without a lot of practice and re-reading.

Totally agree! I own all these, just finnished "Modern Recording Techniques" and definetly reccomend it. LOL to tell you the truth my buddy found it in a box on the side of the road with a bunch of useless books, killer score!

about time this thread got bumped.

jonny m, in regards to your vocal style - a sample would go a long way in getting some advice.

try dipping ~800hz on the vocal and see where that gets you. you could also try a small boost at 3-5k which is the "clarity" and "presence" range for vocals. i'd go easy on the compressor unless you're getting the exact sound you want from it.

sounds like you've got a pretty light mix. i'd run high-passes on the guitars as well (less aggressive since you don't have drums and bass) and try some complimentary eq between them and the vocals. the frequency will depend on your guitars (mine happen to be very mid-heavy so i dip a lot between 400-1600hz). you should also do this in mono so you aren't fooled by hard panning.

overall you should consider rolling off the lows and highs - there's a thread here which mentions the RIAA curve, basically a pretty steep high-pass at 60hz(?) and a similarly steep one at 15khz (or was it 12?). anyway i don't use it yet as you can tell but it will help you eliminate lows robbing you of headroom and highs which can lead to a nasty harsh sound when they all get together.