Mixing Clockwork Clown - A fHumbling Journey - Part 6 - The Vocals, THE VOCALS!
by, 07-22-2012 at 07:29 AM (2879 Views)
This is THE LAST(!) installment of my blog on mixing the winning mix of the 1st round of the Slate Digital Cup. Here is the previous installment: Mixing Clockwork Clown - A fHumbling Journey - Part 5 - Atmospheres & Transitions
... & here is the link to the actual mix: cLockWork cLown - fHumble mIx
The RR regulars among you may notice that the title of this installment of the blog is somewhat familiar - I was intending to call it: "Mixing Clockwork Clown - A fHumbling Journey - Part 6 - The Vocals, The Vocals, THE VOCALS!", but I ran out of space... Why? Well you might recall this thread Brandon posted after the "Sweet Sixteen" round of the April Slate Cup mix: The Vocals The Vocals The Vocals
... Now, Brandon never confirmed or denied that he was talking about my April mix, but that seemed to be the consensus... In any case, you can see that this was a hard lesson learned, & you can read evidence of me chastising myself severely later on in the same thread... The reason for my earnest audio-asceticism is pretty self-evident though - Without wanting to be self-aggrandizing, in my March Clockwork Clown mix, (I felt) I got it (THE VOCALS, that is) very right - so, I definitely took a GIANT leap backwards in April.
From the moment I heard Iain's (Lazy E's) vocals on Clockwork Clown, I knew they were something special. Let's face it, the guy can not only sing, write & play, he also has a beautiful, rounded, honeyed vocal tone, just begging to be presented in the most flattering way possible.
Intriguingly, one of the subjects that consumed hundreds, if not thousands of key strokes during the course of our "peer critique" phase of the March Slate Cup was the pitch issues of the supplied vocal track, & how these were handled/not handled by the participants. Pitch corrected vocals seem to be not only expected, but almost demanded on any modern mix - I've used it many times myself - but my goal was always to try to make it "invisible"....Just to be clear, I have nothing against "invisible" pitch correction - With tools like Melodyne, this is very much an attainable goal. The thing is, if it is truly invisible, no one will even be aware of it's use.
However, it seems in modern pop especially, & many other genre, "hard" tuned "Perfect" pitching, with it's characteristic "robot-like" artifacts is a sound that seems ubiquitous, unavoidable even. This is a sound I actually HATE with a passion.
Now, I realize, that's a pretty strong opinion to voice, & it pretty much labels me as an "old fogey" who hankers for the days of those serious, yet tone deaf, tuneless warblers who abounded during my childhood years, but hang on - this is nothing to do with sentimentality... I'm sure no one enjoys listening to off-key singing (myself included)... No, the thing that really irritates the hell out of me about this style of pitch correction is how boring, unimaginative, & pedestrian it all is....
It reminds me of the '80s when big gated reverb & big snares came into vogue - at first it was interesting, quirky, different, compelling even. Then it became the standard. Rather than coming up with something new/ different/ compelling/ original, the hordes of zombified brain-dead slavish imitator-engineers out there fed hungrily on decaying corpses of superior, artistic/original intellects & turned it into some kind of crazy reverb & snare arms race. It just got stupid - taste went out the window... yes, we listen to it now & laugh - yet it was oh-so-serious back then.
If you take a moment to peruse modern recorded music history, it doesn't take long to see a pretty clear pattern emerging - Technology has almost always been the handmaiden of landmark artistic changes in direction.
Interestingly though, what is often even more evident is that it is the subversion of technology that has often provided the real watershed moments.
Take the electric guitar & amplifier for example - originally conceived as a way to simply increase the volume & provide clear, bell-like tones for an instrument that was typically lost in the standoff between the noisy audience & even noisier band-mates during the rigors of live performance before the days of sophisticated sound re-enforcement... Suddenly, guitarists discovered that abusing their amplifiers - & turning them up to levels never dreamed of by the designers - produced not only a terrifying sonorous roar, but also a beautiful, soaring sustain that turned the instrument from a humble supporting rhythmic role into a a mighty, melodious sword of power, with magical abilities to create shards of blistering feedback that morphed into glistening notes, & vice versa.
... & so it continues - Read anything about the Beatles recordings & you see renegade musicians & engineers pushing the equipment past it's limits, creating new sounds in the process.
Quite ironically, even pitch correction, originally conceived to simply ... erm... correct pitch, became a "sound" unto itself when an innovative engineer discovered it could turn the none-to-tuneful Cher into a singing fembot with a worldwide smash hit in "Believe"
Check out the article on it's making & the belated "Historical Footnote" about the "secret" of Autotune: Recording Cher's 'Believe'
... & then the zombies came out again...
It seems that, somewhere in between the Beatles & Cher, people forgot that new ideas, lateral thinking & serendipitous mistakes are some of the most valuable commodities in music. Today, instead of everyone trying to push the unbelievable levels of technology we have available to us into compelling new sounds, everyone is being sold the "modelling the old stuff will make us great" myth.
Fooey to that! We're modern, we have modern gear, let's be modern....
When I mixed Clockwork Clown, I wanted to see if I could do something new, & different with Autotune - even if it was in a small way - I wanted to stare my own (irrational?) prejudices straight in the eye & deal with the issue that I hated it so much... Was it simply because of the "no-brainer" approach that was being applied to it's use? Was it because I just have a bloody-minded tendency to always go against what appears to be the inevitable landslide of popular opinion. It might sound hypocritical, but I really wanted to use it! However, in doing so, I also had a burning desire to subvert the existing paradigm. I deliberately set out with a goal to do something that I'd never heard done before. (I'm not saying it hasn't been done, I'm just saying I haven't heard of it being done).
Now, I'm not trying to sound superior here, because I've bought into the "older is better" myth at various times as well. It's just like anything in life though - you start out thinking that everything has to be done a certain way - the way "the pros" do it, or have done it. Gradually, you realize you have to find your own way, or your stuff will come out sounding like everybody else....
Anyway, more on that later - first, the "sensible" stuff...what of the vocal?
The vocals were quite dynamic, so wanting to preserve their natural timbre as much as possible, whilst still retaining control, I started out by compressing the vocals via the "Backwards Compression" method that I outlined in this tutorial here: "Backwards Compression" - Have you tried it? . As it is a "destructive" process, I didn't keep the exact settings I used. From memory, I think I used UAD2's Fairchild 670 compressor, compressing pretty hard. When I use this method, I usually hi pass the vocals mildly at the same time, to avoid the low end triggering the compressor too much.
Below, you can actually see the compression in the screen shot: The top waveform is uncompressed, the one below is what it looked like after being backwards compressed:
... here is the original dry vocal:
... & here is the backwards compressed version:
So that was a good starting point. There was a lot to like about the vocal as it stood, but there was a certain "woolliness" about the vocal that needed to be downplayed to give it a more attractive sound. It seemed that it was being generated from slightly excessive low end created by the proximity effect. I tentatively applied a high pass filter with a slope of 12dB per octave - Moving it's turnover point up to progressively higher frequencies as I played the vocal in context revealed that at around 80hz, I managed to achieved a nice balance of low end girth & low-mid clarity. Instantiating Waves CLA LA-2A compressor, I found I was able to get a pleasing, additional "detail" lift in the texture of the vocal. I really wanted to get the vocal sounding "prettier" though. I use that adjective because that's the quality I imagine that the high frequencies bring to a vocal sound. The low end & mids push the power & strength, but the high end brings the detail & finesse that are critical when a vocal is right up front, as this one was. With that in mind, I grabbed UAD2's Neve 88R Channel strip. Using only the equalizer section, I gave a gentle shelf lift at around 7.5k, & a fairly hefty boost around 11k to really bring out the breathiness. Things were sounding really fine now, but there was some sibilance that needed to be tamed, so out came Waves' R De-esser to do it's usual excellent job.
Here is the screenshot of the Lead Vocal insert chain:
This is the sound that resulted:
Now that we had a nice sounding dry vocal. What about send effects?
I really wanted to emphasize the w-a-r-m-t-h (an emotional description, not a sonic one btw, Brandon!) & humanity of the vocal - I wanted it to enfold the listener like a cozy blanket, so my first step was to create some vocal widening. I know there are plugins that do this in one step, but I've been using this complicated little method, because it allows me to use the stock plugins in Sonar to achieve excellent results. I send the vocal via an aux send to 2 busses - each panned hard righ & left respectively. On each buss I insert Cakewalk's "Pitch Shifter", with one set to shift -0.11 & the other +0.11. Both these busses are high passed @ 500hz, & have Waves' R-De-esser set to destroy the sibilance. These both then feed into another stereo buss with Cakewalk's "Channel Tools" plugin set to the "stereo widener" preset.
Here's the screenshot:
The result really spreads the vocal subtly across the stereo spectrum. Here's the widening on it's own:
Next the reverb: Most of the time, I like to create a "common" reverb that creates the impression of all the elements in the same room. However, as I have already outlined in earlier episodes of this blog, Clockwork Clown was a little different in that it takes the listener into different virtual "rooms", depending on the dominant character influence on each respective portion of the song.... Ultimately, I took a slightly different approach with the reverb on the "hi fi" "human" voice we're discussing here. Rather that trying to place the voice in a specific environment, I used reverb simply with the goal of subtly enhancing the smoothness of the vocal. I chose a plate reverb model preset from IK Multimedia's Classic Studio Reverb plugin. As you can see from the screenshot below, I eq'd the signal coming into the reverb with Sonar's Pro Channel eq, rolling off the lows around 330hz with a fairly gentle 12db per octave slope. I found the reverb was still creating some low-mid "cloudiness" when combined with the vocal, so I dipped out 4.2dB at 489hz as well. The reverb also seemed to resonate unpleasantly centered around the 3.7k mark with the vocal, so a gentle dip there of 2.9dB assisted the high mids of the vocal in coming through with great clarity & definition.
Here's what the vocal reverb sounded like solo'd:
Together, all this insert & send processing added up to the following sound:
I was very happy with this vocal tone - It seemed to fit the song like a glove... but... well... more on that later...
I already had already done my "invisible" vocal tuning to the best of my ability using Sonar's in-build "V-Vocal" system. I tuned the original vocal fairly gently, just "averaging" some notes more accurately, but essentially leaving the natural vibrato & glissando of the original performance intact. V-Vocal a good system when you know how to use it, but it's far from perfect . Having recently invested in Melodyne, I can certainly now see that V-Vocal has it's limitations. However, one thing that V-Vocal does without breaking a sweat is the ol' "hard-tuned robot voice" we all know & love so much....
It's fairly obvious that I used this effect to great advantage during the sections of Clockwork Clown where the vocalist needed to portray the closeted, robotic toy clown. Here's the vocal chain I used for those sections. My rabid anti-autotune sensibilities felt unabashedly justified in rationalizing my decision to use hard tuning in this section because, after all - the vocalist was a robot (In my mind)!
Even here though, I used a parallel, or "ghost" vocal approach:
The first "Toy Vocal" track had no hard tuning, but I high & low passed it heavily, & sent it into Sonar's VX64 Vocal Strip plugin set to the "Telephone " preset, which further "grungifies/lo-fi's" the sound. For ambiance, I sent it via auxiliary send to the "Carpeted Room" convolution reverb setting I mentioned earlier in the blog.
Here is the vocal chain:
& here is the sound:
I wanted to make it really "toy robot" sounding, so rather than processing the original, I decided to do it in parallel. I hard tuned the vocal track, lo-fied it using low & high pass eq, sent it into Cakewalk's cool little "Alias Factor" plugin wihich is basically a bit-crusher. Appropriately the "Cheaper Toy" preset seemed to be just right in conveying the right sense of...well...cheapness.
Here's the chain:
... & here's the sound on it's own:
Combining the two "toy vocal" sounds together seemed to get just the right balance of "robot" & humanity I was looking for during these sections. so, here's how they sounded combined:
Now, as with any creative endeavor, one idea usually leads to another. So after fooling around with the parallel hard-tuning for the different lo-fi vocal sections, it got me thinking...
Now here's where we get to the part that I hyped up so much at the beginning. It's actually fairly subtle, but I feel it made a big difference & gave the "hi fi" vocal an "edge"... You be the judge.
Using the "hard-tuned" version of the vocal, I created a parallel lead vocal track for the hi-fi vocal. However, rather than sending it to the master buss, I inserted an instance of IK Multimedia's Classik Studio Reverb Plate with the "Wide Vocal" preset selected. I set the "Mix" parameter to 100% "wet", so that effectively, it became a reverb in itself. The big difference being that the reverb is very different from the source - hence the "ghost vocal" nomenclature.
Listen to the raw track on it's own - it's pretty "robot-like", huh?
...however, here's the processing chain that changed it into a parallel, "reverb only" signal:
...& the "Ghost Autotune" solo'd:
In addition to the high & low passing at 960hz & 13k respectively, you'll notice the fairly hefty eq cut at 2.49k. This was to get the Ghost Autotune track out of the way of the lead vocal, because of the strong resonance right in the high mids, caused by the tuning.
...& here is the final combined processed vocal & the ghost autotune together - the sum total of the final "hi fi" version of the vocal on the track:
In other sections of the track, the subtle "rubbing" of the autotune against the lead vocal is perhaps more noticeable:
Maybe not as earth-shaking as I probably hyped it up to be, & maybe you don't think it's such a big deal... but honestly, is that ever the case? In audio engineering, we often make a big deal about things that make the tiniest difference.
Nevertheless, I'm quite proud of the vocal tone on the final mix. It reminds me of a sweet, angelic "Halo" around the vocal - which is exactly the type of vibe I was trying to achieve. Ultimately, mixing is a game of inches - Thousands of little, tiny incremental changes that do actually add up to a lot in the final mix.
The larger lesson I learned from this is that everything hasn't been done yet. With curiosity, imagination, experimentation & creativity, there is no reason for us to think that we simply have to go on mimicking the glories of the past. Sure, study it all, learn how to do it all, but don't just be satisfied with that... Explore some of the new stuff & find new ways... One idea leads to another...
...just as a post script to this, & to illustrate how ideas beget yet more ideas...
With that in mind - check out the post-script to this blog here: Post-Script to Clockwork Clown Blog
... more practical, untypical uses for vocal tuning...
So that's it - A bit of a marathon I know, but I hope you found at least something of interest & use.