Lessons

Lesson Info

Mixing Vocals

The vocal part is interesting because Spencer, our singer, is an extremely talented vocal producer and he knows a lot about how to produce his own voice and he was very insistent upon being able to apply the processing to the vocals on his end and he did a lot of that on the way in, so the vocal mixing for this album is basically non-existent here. I have here what he provided me. We have a main vocal track, a backing vocal track, and ambiance track, and I created a reverb effects track where I sent something through a reverb, but for the most part, it really just consists of three tracks. Now, I can show you a little bit about how his vocals sound and we can talk about the general techniques for mixing vocals, but I can tell you I'm pretty sure he tracked this through a hardware chain involving two microphones. He was using a Neumann KM 84 small diaphragm condenser alongside a Manley Reference large diaphragm condenser and they were run separately, one through an LA3A and one through ...

an 1176 compressor and combined through an EQ, so they're pretty much mix ready from the word go. I'll take off what I did apply to them. That's just cutting the volume of them a little bit, so I'm going to leave that in tact, but this is what the vocals sounded like when they came to me. ♪ It's always better on the other side of that ♪ ♪ Wanting what you never had ♪ ♪ It is a goddam broken state of mind ♪ ♪ Share the world that we've been living in ♪ ♪ It's never ♪ So, you can hear on there there's a fair amount of doubling going on. He's also got quite a different kind of ambient sound towards the beginning. There's quite a roomy sounding ambience he's using on his voice, which kind of surprised me, but it works really well in the mix. For the vocals for this album, he really wanted to go heavy on the layering and you can really hear the amount of detail put in to make sure that all of these doubles are super in tune and lined up. He's a really, really great singer and it was only knowing his skills as a singer and knowing his skills as a vocal producer that will allow me to let him provide files at this stage. It did make my life extremely easy when it came to mixing, but essentially, we can have a listen through to some of the different sections and hear the various effects that he's applied to his vocals. I really like how he's got a lot of articulation on the consonants of what he's saying. They're really kind of sticking out, they're really spit-y, and that's part of his delivery as well, but a way that you can enhance that is to not use the fastest possible attacks time on a compressor or maybe use a second compressor after your main compressor with a slightly slower attack time to really exaggerate a little bit like with the guitarist to get that very hard front end to the words. I'll just play a different part, I think this is the chorus here, and then, we'll compare it with some of the effects we've got later in the song. ♪ Stop bleeding on the inside ♪ ♪ We are alive ♪ ♪ And it's more than enough ♪ ♪ to paint the walls that are white ♪ ♪ Can we show the ones who are blind ♪ ♪ Stop bleeding on the ♪ So, it's interesting to note that I'm pretty sure he's triple tracking the vocals. We've got hard panned backing vocals out to the side with a center vocal that's kind of doing majority of the lifting work there, and if I bring in his ambience channel down here, which is kind of a printed version of all the delays and reverbs that he used during tracking, we'll get an idea of how he sets up the ambience on his voice. ♪ Stop bleeding on the inside ♪ ♪ We are alive ♪ ♪ And it's more than enough ♪ ♪ To paint the walls that are white ♪ ♪ Can we show the ones who are blind ♪ ♪ Stop bleeding on the inside ♪ ♪ We are alive ♪ ♪ And it's more than enough ♪ ♪ To paint the walls that are white ♪ So, he's using quite heavy use of quarter note delays there and he's also rolling off a huge amount of top end in the delay sends and that's something that I definitely do a lot when I'm mixing vocals. And, are those delays printed as well? Yeah, that's literally a printed track down here, the one that's labeled ambience ♪ Stop bleeding on the inside ♪ ♪ We are alive ♪ ♪ Can we purge this ♪ ♪ And it's more than enough to paint the walls that ♪ You can hear it's also very saturated. He's actually distorting the delays quite a lot and the combination of that with the top end roll off is giving me a kind of warm layer underneath the vocals that isn't conflicting with what he's saying, but he is also providing a bed of sustain and creating a slightly elongated sound. At the same time, he's not relying on it super heavily in the mix if you listen to the vocal ♪ Can we show the ones who are blind ♪ It's really not conflicting with the vocal. I could probably mute the ambience in the mix and probably not lose too much of the vocal sound if I were to play that now with none of the ambience. I meant, in the mix. ♪ Stop bleeding from the inside ♪ I need these plugins to be active ♪ It's more than enough to paint the walls that are white ♪ ♪ Can we show the ones who are blind ♪ He's sounding very quiet for some reason. I'm just going to figure out why that is. Oh, we have this playing. ♪ Stop bleeding on the inside ♪ ♪ We are alive ♪ ♪ And it's more than enough ♪ ♪ to paint the walls that are white ♪ ♪ Can we show the ones who are blind ♪ I'm really not getting much of a sense of these delays kind of conflicting with the reverb tool when everything's being played together. I think it's really important, as with lead guitars, to try not to rely on ambience to carry your vocal sound. I think that's really going to send it straight to the back of the mix. If you can, just focus on getting super tight track parts with really careful use of compression to both, flatten the vocal, but also give it a sense of excitement and try using, actually, no reverb or no delay, perhaps, and see what you can get with that. I think, when you listen to professional productions, you might actually be surprised at how little reverb is sometimes used. I was interested to watch Joey Sturgis's class that he did at Creative Live and see his take on the vocal ambience, and he used a very minimal setup and would actually split up the delay on reverb onto different parts of the vocal mix, so maybe the doubles would be treated with reverb or the lead would basically have nothing on it, while another double might have the delay. I thought it was a very interesting way of achieving that texture without just washing everything out with delays and reverbs. I'd highly recommend people checking out that course. I got a whole load of ideas for ways that I would like to treat vocals. In addition to as many vocals, we have all sorts of parts going on which are supplementary. They might be harmonies or they might be full on choral parts, and we can check that out. Just going through the song to hear a few of those different instances. ♪ Little time we have ♪ ♪ Stop bleeding on the inside ♪ ♪ We are alive ♪ ♪ And it's more than enough ♪ ♪ To paint the walls that are white ♪ ♪ Can we show the ones who are blind ♪ So, that's some interesting kind of fifth harmonies and fourth harmonies on top of the lead vocal there. ♪ Separating ♪ ♪ Drive me crazy ♪ ♪ Separating minds ♪ ♪ Fly past the withered trees ♪ ♪ Drive me crazy ♪ ♪ Separating minds ♪ It's really cool for me to hear all the detail that he put into some of these parts. I was mainly focused on the guitar tracking and he was tracking the vocals on his own and sort of self-producing those, so he really had a chance to go nuts with these layers and, while we don't have them all separated here, I know that his sessions had outputs of 50 vocal tracks or something. Now, that would have been a real pain for me to have to mix from scratch and he ultimately saved me a lot of time by providing them like this. With most projects, as I say, I probably wouldn't let a producer provide me with such scant parts here. I did create some extra reverb for this part. I just kind of copied his backing vocal part onto a new track with a 100% wet reverb. This reverb patch is one that I really like within Logic. It's called Splendid Verb, although it seems to misspell it within the actual GY. There, it's Spendid Verb. I don't know if that's the case with anybody else's copy of Logic, but it seems to be the case with mine. I really like this as a general reverb. I eventually got it as a saved preset just so I can always get to it quickly, but it's one of the very large haul presets. It's six point eight seconds long and I've been using this since before Juggernaut and I still reach for it all the time for really big reverb textures, whether it's for a really impactful snare hit or for a guitar that needs a really, really long cathedral-style reverb and I really like this method of copying the raw track, the dry track, onto a new track with the reverb set at 100% wet because what that allows you to do is do things like here, where I faded the dry signal and so, the reverb is getting an increasing amount of signal being sent to it and it's just a very quick and easy way of doing that without having to rely on a lot of automation. Sometimes, I'll then print this whole channel down to another stereo track and just get rid of it. It's only if I'm creating any kind of in-depth vocal effects where I really want to, say, distort and radio effect a part, I'll copy it onto a new track, apply everything 100% wet and then, blend that in with the original. So, this was the effect of running that through the reverb. ♪ Drive me crazy ♪ ♪ Separating minds ♪ You can hear it. It's a six point eight second reverb, but it kind of doesn't hang around that much. I think you can see from the shape of the wave here it kind of decays quite quickly and then, hangs around before decaying further, so it's a really good patch. If you're a Logic user, I highly recommend checking this one out for all sorts of uses. It can be really good on some vocals, even as a main reverb if you've got loads and loads of space for that. In general, I really rate Logic's own plugins. We were talking earlier about compressors and, especially in the new version of Logic, there's some really, really good sounds to be had with those compressors. I've done a couple of sessions recently where I actually ended up using the Logic stock compressor as a main snare drum compressor, as a main vocal compressor. Not to sound like a spokesperson for Logic, but-- You can add a fab filter. Fab filter is great too. Definitely don't overlook the plugins within your own door. They don't need to be a limitation. If you look at the session that we're working on now, there's a huge amount of Logic's own stuff, whether it's the Channel EQ, things like the Silver Verb plugin, we've got a load of the Channel EQ, we've got some instances of the compressors, I'm pretty sure, somewhere through here. There's one there. Not really proving my point, but the Tape Delay is a really, really good plugin. A lot of these actually have advanced functionality that you might not be aware of just from looking at it. If you click this arrow, for example, and the Tape Delay, you get access to a distortion parameter. I quite often reach for this before I'd reach for more of a boutique plugin like Echo Boy by Sound Toys. And, I think that goes for most of the doors out there. Definitely don't feel like you have to be using the ultra boutique plugins to be able to get good sounds. Often times, I'll use a Logic plugin during tracking just to get the point across and end up keeping it for the final mix. So, just definitely no need to be snobby about what plugins you use for that stuff. This is kinda going to revisit compression for a short while, but something that can be really interesting with vocals and works really well on bass guitar and stuff like that is to stage compression across several stages. I don't have a raw vocal track to do that on right here, but if you were to imagine using something like an 1176 kind of compressor that we were talking about a lot today to really kind of cut down all of the really high peaks of a vocal track, that then allows a second compressor to not have to react to anywhere near as wide a dynamic range and you can really kind of get a far more dynamic and consistent sound without having to squash really hard with one compressor to achieve that. I'm pretty sure that's what Spencer did on his own vocals, but that would be my general way of compressing a vocal. To go back to Joey Sturgis for a second, his plugin, Gain Reduction, is a kind of vocal compressor plugin, which I've used, actually, a little bit in recent sessions and I've really enjoyed using that and it almost feels like cheating because I think it has a multi-stage compressor circuitry going on within it's software algorithm. But, I would highly recommend that to anyone getting into producing vocals now. It's like a one-stop shop for a really, really good aggressive vocal sound. On top of compression, I often find that some De-Essing is necessary on any vocals, and here I have used it on Spencer's vocals. Again, to talk about Fab Filter, I really like this De-Esser and this is kind of typical of the settings that I would use within a De-Esser. I generally find that ducking the Essers by up to seven or eight dB is a really good idea. What the De-Esser essentially does is duck the level of the whole vocal every time an S becomes overbearing. You could set it to split band mode, which would just duck certain frequencies. To me, De-Essers always sound best when they're working as a general ducking of the whole level of the vocal. If I find a part which has some sibilance, we can hear what that's really doing. ♪ The withered trees ♪ ♪ Ascend to holy air and breathe ♪ ♪ Fly past the withered trees ♪ ♪ Ascend to holy air and breathe ♪ ♪ Breathe ♪ I find that one section where he just sang with a lot of sibilance. Let's see. It was around here, wasn't it? ♪ Fly past the withered trees ♪ And then, that without the De-Esser. ♪ Fly past the withered trees. ♪ Those S's are just kinda really coming out and that's a result of the amount of compression he's using, but that's very, very standard. He might have done some De-Essing on his end as well, but, for me, De-Essers are almost always necessary when I'm working with a vocal track. As far as EQ goes, I really try not to have to do very much to a vocal. You can see here, this was my EQ on his already EQ'd vocal, but generally, I find that the less you can do to vocals, the better. The human ear is extremely sensitive to changes in vocal sound. We're so used to conveying emotion and understanding subtext through hearing people's voice that are talking to us that we've become extremely sensitive as creatures to kind of strange sounding-- If you over-EQ a voice, for example, you might end up with something which just sounds very unnerving to the listener and they might not necessarily know why, so generally, I find that if a singer's been right up close on a microphone, there's always some kind of chest resonance that might be exaggerated when you compress the vocal. You might find yourself having to duck out something kind of around the three, 400 area. Sometimes, as well, just from that same proximity, you might have a buildup of mud, kind of like seven, 800 Hertz-ish area, but I try not to have to do very much from there, right the way through to the very top of the vocal range. By mixing into the brightening that I had from our top down mixing, that occasionally can create too bright a sound in the vocal. I did actually find that to be the case because of the fact that Spencer had pre-EQ'd his vocals, so what I did in the instance of this, and I've done this on other mixes too, was create a separate instrumental bus over here. His vocal is actually ignoring all of this vocal chain here. It's going pretty much straight to the limiter just because of the fact that he did add a whole load of EQ. I think he boosted the top end quite significantly because he wasn't mixing into like a top down approach. If I had a completely raw vocal, I might well be able to leave that EQ vocal in place, but this is just an example of what you can do if you find that you have too much top end from a track. Once you've run through this EQ on the master bus, you can create an instrumental bus that then leads into the actual output from your door and you can essentially bypass any extreme processing there. You can choose, maybe you can leave your bus compression and your tape saturation on that bus if it doesn't affect whatever it is in a negative way, but you can certainly move the EQ onto a different bus, which is treating everything apart from what you're having a problem with. In general, my normal approach, if something wasn't pre-processed, would be to try and cut those frequencies at the source and not have to create a routing nightmare for myself. But, in this case, just because of the nature that things were done, that was necessary. Can we talk about using multiple compressors on a vocal again real quick? Yeah, sure. Is there a certain amount of gain reduction your would try and get at each stage or using one compressor more than the other? I would say, generally, if you're mixing vocals for rock and metal stuff it's typical to really pin them with a huge amount of compression. You could be dealing with outputs of ten dB of gain reduction with each. Each? Wow. With each. I would normally have the 1176 doing the bulk of the work. I might do ten, 12 dB of gain reduction. Actually, when I track vocals with a vocalist, I have them run through a hardware compressor. I use a de-Stressor at home and I will generally be looking to get six up to maybe ten dB of gain reduction, depending on the kind of singer. For one, I find that that really helps them when they're performing, having more consistent volume. If they're moving around on the microphone, it's not going to make too much difference to what they're hearing in their ears, but that will often then serve as my first stage of compression. Then, with the second stage, maybe I could be going as much as five or six dB. If it's really slammed, then I'd happily go above ten dB. I know I've seen some top producers talk about compressing vocals by the time it's kinda gone through a whole analog chain, five or six times even, and sometimes, in total, you're talking about 40 dB of reduction or something like that, but then, some other guys will tell you compression is evil and what you should do is automate every word to be even in volume instead of relying on an automated process to do that. Personally, I think I sit somewhere in the middle. There's a certain sound you get from really compressing a vocal. It brings out the character in their voice, it can make them sounds really aggressive, especially if they're screaming. I really like when I hear overly natural vocals within metal. It kind of breaks the fourth wall a little bit for me. It doesn't kind of match the intensity of the rest of the instruments, but certainly, if it was more-- In terms of dynamics, you mean? Well, it's dynamics, but also the raspiness that you get from compressing, really bringing out the detail and the kind of throat of the singer, especially what it does to the consonants, like really exaggerating those hard consonants to make the really sound articulate and spit-y, I want to say, which sounds like a negative, but for some reason, I like that sound. If you listen to a lot of things produced by John Feldmann, you can really hear that. If you listen to early, The Used records or something like that, that's a really extreme example, but, to me, there's so much energy conveyed by his use of compression there. So, it really depends on the singer. I try not to, as I say, completely rely on compression. I'll go in and automate a fair bit, as well, to achieve a more consistent mix, especially from section to section, but I'm not afraid to, if the vocalist needs to come across as super aggressive, then I'll happily smash it, upwards of 20 dB or something like that. And, you're talking about using, generally, a faster compressor first and a slower compressor next. Yeah, the ideology is that the really fast peaks that might be present on consonants or when we're talking about bass, some notes might have a really huge spike on them if you pick a certain way. That can really throw off a more smooth compressor with a slower attack and release time and you'll end up with a very obvious pumping effect that might take awhile to stabilize. By using a fast compressor first to kind of trim all that stuff down, you're making its job so much easier, the second compressor, because it's seeing something that's way more dynamically consistent and it can kind of do a more gentle job and be more transparent that way, but often, those slower types of compressors that people use, things like LA2As or Fairchild compressors, are like vintage designs with a huge amount of saturation and coloration that they impart. Sorry, that's probably really loud. That's one of the real benefits, as well, to using that kind of compression mode is you can combine the characters of a super aggressive compressor, like the 1176, with something that's really going to enhance the tone of the vocal from the second compressor. Sometimes, you might even want to throw a third compressor on there, doing very little, just to further enhance that. It's all worth messing around with. Like when we started the mixing session, we looked at using the CLA-76 as a distortion instead of a compressor and that would be an example of the kind of way you could do that. That's actually a really good one to try on screamed vocals, in my opinion. Stage compression is definitely the way to go in terms of vocals and bass. I haven't heard of too many other ways of using it. I've heard of a couple of people trying to do that with snare drums, I think, but, to me, I don't really know that there's too much point in that. I really like the approach, to go back to drums for a second, of using minimal amounts of variables and that's one of the reasons why I've condensed things down to using buses to basically just use leveling of tracks to achieve something which is going to be close to what I want and then, processing that at the bus. Also, the same thing with the bass. Some people seem to really enjoy having a huge amount of parallel tracks and some vocal producers will have tracks that are super Hi-Passed and there's just treble and they compress those really hard and stereo widen them and there's some really cool tricks out there that people try, parallel distortions. Personally, for me and my general ethos of mixing, I find that very difficult to maintain. I like to have faders that I can just grab if I want to just adjust the level of something. I don't want to have to worry about how it's also affecting another bus, which is on ascend and disrupting the balance of multiple things just by moving one fader and I found that to be the case a lot when I was trying these ideas with lots of parallel buses. I'd find that I was always chasing what was creating something that I didn't like. I was applying loads and loads of EQ and plugins across my whole mixer and just ending up in a real knot when it came to troubleshooting. I would never know where to look or I'd always be wondering whether I should be taking that frequency out of the parallel thing or this other parallel thing or the original track and, to me, I really like to simplify things down in this way, especially with the drums just 'cuz there's so many things going on, but the same thing applies to vocals. I'll generally use maybe some element of saturation and distortion, but I won't really ever do that in parallel. If I do in parallel, it will be within the plugin with a mix control. I really like to try and limit those faders that I have. It seems like you're spreading saturation in very small amounts across the entire mix rather than parallel in one spot. Yeah, absolutely. You also hear about people maybe creating a bus just for lowering instruments, maybe busing the kick and bass together and then, compressing that really hard to get them to interact in an interesting way. I've heard recently about CLA doing that and applying, I think, R-Bass or some plugin that really enhances the sub lows of something and then, blending that in underneath. These are all really interesting ideas and I'm always fascinated to learn about them. Personally, for me, I've found that just straight up, simple, serial plugins tends to do the trick for me. I really like the simplification of that. We could have, if I wasn't explaining every step, that drum mix could have come together in probably a matter of ten or 15 minutes, whereas I used to spend hours, days, really trying to trace down that character that I was aiming for, just because of the fact that there was so many variables involved. So, if anyone else is having that problem, I highly recommend that approach and I think just applying some of the general ideas that we had in terms of the frequency balance, especially with the kick drum, kind of getting even amounts of top end information and bottom end information, getting the snare to look pretty flat across the analyzer. The same thing with the guitars, actually. Generally, all the good guitar sounds I've achieved look pretty flat on an analyzer. That's kind of told me that I have a very nice, even balance of frequencies and it means that when you put these elements together they kind of naturally sit within each other. You tend to run into problems if you've got things like guitars with a huge hump of information in one area. It's going to create issues for anything else that conflicts in that area. If everything is kind of pretty well spread out throughout the spectrum, when you overlay them on top of each other, they all tend to coexist quite happily. To me, really, one of the fundamental aspects of mixing, or really what mixing is, is reducing the amount of contrast between things. By that I mean, creating a sense of everything belonging in the same space. If you have a guitar tone that is super dark and lacking in high end information and a really bright bass tone, for example, to me, it just doesn't gel right and I think, to most listeners, they pick up, maybe they don't know why, but there's just a sense that things are lacking in cohesion and it's your job as a mixer to go in and find ways to create similar frequency responses through all of the instruments, so that when they go together, they can be cohesive and they can glue together and they can coexist, you can hear everything, yet everything's also working together, and that's what I've found using analyzers to help put everything in the right ball park and make sure everything's nice and flat across the spectrum has really helped with, again, to reiterate what were were talking about earlier, it's not really going to make everything sound the same. If I apply the same general curve to a different kick drum, it's going to sound completely different just because of the natural characteristic of the drum or guitar turn. All of these things are made up of huge micro-variations that are constantly changing and, again, unless there's something really wrong with a character of something, I try not to touch that. I try to make broad, paintbrush kind of strokes and then, see how things sit together before kind of committing to really trying to modify one sound to sound like another sound that I've achieved in the past. It seems like using analyzers in that way, too, is really efficient when you're coming into a new situation like this. You've never recorded drums in this room before and you're getting sounds that you're happy with and, like you said, potentially, in 15 minutes. Yeah, mixing-wise, certainly. I would say, on average, I'm up and running with a drum sound, if I'm not stripping the tom tracks and things like that, having done it a lot, and Matt can attest to the fact that, while we're on the road, I will even practice this stuff. If I have a few minutes spare, I will open up a session and pull all the plugins out and just mix, especially the drums, from scratch. You just did it with Animals as Leaders' record a bunch of times. Yeah, I did that. Yeah, after it came out, just to see what I could get out of it. As my skills have developed, it's very interesting for me to revisit old sessions and try messing with that stuff. I also really would urge people to go and look on the internet for stems that other people have recorded and try working on those, even if it's maybe not quite the style of music that you work on. Working on somebody else's tracks, they may well have issues within them that you have to work around, but will really help you when you get into a situation where you might be presented with an issue within one of your own mixes. As much as we really want to try and ensure that every source tone is completely perfect within a mix, sometimes it's just not feasible and you have to have a kind of trick bag of stuff, which you can rely on and you have to have experience of working with adverse condition and still achieving a professional result, but, again, as we've been talking about all day today, great source tones is really the key to all of this. There's actually not even that much to say about things like the treating of these guitars. It's just some stuff that's really specific to this guitar tone because we tracked the tone that we wanted to hear. The bass had more processing because we ran it DI, but if that had been an amp track, we'd have been looking at an EQ, a limiter, the multi-band compression, not in this order, just the tape emulation and then, the track space. We would have been looking at five things, all of them fairly minor adjustments. The drums, sure there's some more major adjustments there, but we're working with an acoustic instrument and we're trying to turn it into something which is more analogous to an electronic instrument, really, when you break it down, but we get there with a surprisingly small amount of moves. If we look across here, the snare and kick have four plugins each, but the work is really being spread and, using this approach, I also found it to be extremely CPU friendly. This mix is sitting extremely low on my CPU meter. It really is not much strain on my computer trying to run this session and this is a 70 something track session. I highly urge people at home, if you have the chance after the session, to go and try this approach out on some recordings. We are, of course, making the kind of jam stuff that Matt was playing yesterday after we took samples, as well as the samples free for people to mess around with, and that's going to be great because, if you've looked at what I've been doing here on this exact drum track, you can completely recreate that and see how that process works for you. Definitely go and have fun with that. I'd be very interested to see if it's the kind of approach that's going to work for other people and see what kind of sounds other people come up with.