A couple of weeks ago I produced another track for assessment. It’s by a songwriter called Kevin Reeves and I found raw multitracks of his song The Game on Mike Senior’s website (definitely recommended for mixing practice!). Mike is a renowned engineer who runs Cambridge Music Technology – an outlet offering studio services and educational resource. He also writes for Sound on Sound and wrote Mixing Secrets for the Small Studio.
Here’s how the track sounded from a straight bounce after importing the audio files:
Before attempting to tackle this mix, I listened to his other work to gain an idea of his influences. Then, I was able to compare to other songs by bands, and research into their production. I found he had elements of ELO, Supertramp, Queen and a general Beatles-y vibe – particularly in this track. This led to about 20 minutes of reference listening so I build on (nick) ideas for my own mix. Dave Pensado (The Dude) provides some fantastic advice towards referencing.
Looking more in-depth, I began studying those behind the sound of Kevin’s influences and sounds. Interestingly, I found Jeff Lynn, longstanding member and producer of ELO, also having worked with The Beatles. The other man responsible for ELO’s work is producer/engineer Reinhold Mack – who’s also just as famous for his tenure with Queen! This is all too big a coincidence – I sheet you not, I definitely made that call before this discovery!
“Triple H’s new entrance music written by Mika?” (Liam Flaherty 2012)
I’ll point out a few things which I thought defined this mix –
- It became a matter of modernising the sound of Kevin’s dated inspiration, and is clearly a pop artist with a rock backline. I enjoy the tightness of the drums in tracks like Queen’s Dragon Attack, however to bridge the gap between these genres, I took heed of pop rock producers to note their work on live drums. Max Martin and Dr. Luke‘s production on Behind These Hazel Eyes by Kelly Clarkson is a typical example of this as it involves MIDI triggered samples. I enjoy the prominence of the kit and the impact it gives.
- Using a subtle amount of parallel compression sprung the drums to life. I’ve learned a great deal since my blog on the technique as my examples are unconvincing and tasteless. By carefully selecting what drum channels (not just all) to send to the parallel bus, and by mixing in the Sends individually (i.e. not all 0dB), I resulted in achieving a way more impacting-but-not-impeding sound.
- In terms of the Wurlitzer-type sound from the start and the clean broken-chorded guitar part, I thought it had Mr. Blue Sky written all over it.
- The lead vocals I think were (as if Michael Bublé became a rock singer) best left with no effects until we get to the guitar solo. To make it sound ‘epic’, I tried to do something along the lines of Brendan O’Brien‘s production on Chris Cornell’s delayed vocals on Audioslave’s Revelations, Shape Of Things To Come in particular (arguably one of the best rock voice sounds ever). I used the standard Extra Long Delay II plugin to achieve this, as well as double-tracking the vocals, one slightly offset. The parallel compression used on the drums in this track sound awesome too and also acted as a catalyst for mixing mine.
Should I even talk about the backing vocals? You guessed it, I ripped on Bohemian Rhapsody (2:54). Just like Kevin did. But we can’t be that narrow-minded. Suited a lot of the mixing of the BVs as between that and a lot of Supertramp’s material. I was going for a distant feel, but at the same time quite present. I just went for it, and shelved at 4.49kHz. Adding a bit of reverb helped create the distance and a
I think overall the placement of the vocals sits well – maybe a bit too ‘in your face’ at times. But it’s definitely a vast improvement in terms of vocal control from my mix a while back, Signe Jakobsen – What Have You Done To Me?, which was facking awful and sloppy. In Signe’s mix it got completely lost in the pre-/choruses and the intelligibility overall is non-existant. I’ve learned heaps off the back of that, and found the reasons for the improvement were definitely better control over the guitars (i.e. compressed), and I didn’t go bananas with the cymbals.
- An effect to mention is on the middle section vocals (2:34). I started out with wanting something tape-echo-y like this. Not sure if I was happy with what came out of it though. I tried to see what George Martin was doing in Sgt. Pepper… but didn’t arrive at the desired effect. After a few plays though, it began to settle with me. Was aiming for that ‘typical breakdown in a song, then builds up’ kind of thing. I mix-automated the Send for the FX channel and the dry channel to intertwine, so on the last syllable of “life” the clarity is back and in time for the ‘epiphany’ feeling (got to believe your own bullskit sometimes…).
There’s a great SOS article all about vocal effects and its creative uses. I gained a lot from it, particularly about delays and the tape echo idea. You can find it here.
- There came a point where I had to stop somewhere, and I’ve been practicing this recently to build on ‘mix-confidence’, as the inner perfectionist needs taming (going 8 hours without food/caffeine is not advised). After deciding to stop, I began mixing from stems which cleared the air and allowed for more global adjustments. After bouncing a few (i.e. ‘vocal up’), I started playing back on different setups to see how it sounds. The mixing process overall involved a lot of objective thinking i.e. gating spill, however I’d say the most part is down to experimenting systematically and working out what suits the genre you’re working to. There’s obviously no hard and fast rules, so it’s down to experience and good decision making at the end of the day.
“If it was an exact science everybody would be doing it and would become filthy rich.” (Reinhold Mack, iZotope interview)
Some really good advice I got on testing any mixdown came from Alex Fenton who runs Swanfield Studios in Leith. I worked with him before doing gigs at the Wee Red Bar and I found him in his car one day listening closely to some of his recent work. He said to be as broad as possible with listening back – To play it on as much consumer products available: car stereos, iPods/Phones and their (rubbish) earphones, docking stations for mono compatibility, home hifis, larger systems, etc. Basically anything which brings the mix out of the studio, as not everybody owns prime studio monitors.
To further-review my mix, I went into work early and played it back flat through the system. Here are my findings:
The general width seemed a little tainted through a larger-scale setup, and sounded different to headphones/home monitors. Perhaps my panning wasn’t too hot in this case, with phasing playing its part as I found the snare sounded a lot different, if a bit muffled. Although, the depth still appeared to emanate throughout the room representing all frequencies evenly.
Thinking back to early first year (October 2009) after being set our first studio task, I think I’ve come a long way since then in terms of sound production. Taking into consideration I didn’t even know what bounce meant, I’ve definitely learned a lot, however I’m still very much a rookie and have got a long way to go until I can hang with Pensado…
Part 2 of 2 of a mixing assessment from my Production & Professional Practice module at Edinburgh Napier University.
It’s by German gothic synth/electro rock trio Girls Under Glass. I chose to mix this track because it differed greatly to What Have You Done To Me? and it had a lot of interesting layered rhythms, basslines and atmospheric sounds. It also allowed me to develop my sidechaining abilities and vocal delay which is what I want to discuss.
You can hear the synths compressed throughout (1:01) and the guitars pumping to the beat in the chorus (1:31)(how old do I sound?), with the vocals having syllabic offsets delayed at times after the choruses (3:15). I also did a manual EQ sweep on the D-Command console before the second verse (2:06) which was cool as I felt like propa DJ lyk.
Producing this track was a bit of a challenge as I’d not really looked too much into mixing rock electronica (I’ll admit here to not knowing where to look for inspiration). I typically thought about Jacob Hellner and Rammstein when I heard the heavy guitars and German accented vocals with the layer of choral atmosphere. Then with the electronic side, I thought of The Prodigy’s Music For The Jilted Generation, only ’cause it carries a similar electro-industrial sound which Liam Howlett and Neil McLellan pull off so greatly, incorporating live band instruments like distorted guitars.
Slightly unrelated, but this link is so cool. Remember seeing this a couple years ago and was fascinated at the amount of samples used and the techniques applied towards the production of Smack My Bitch Up (explicit).
One of the main interests of this mix involves quite heavy use of sidechain compression. I wanted the verses to flow as standard, but the choruses to sound loud and pumping. Some of the synths and layers were compressed to the beat (1:00) which sounded good and added a greater dynamic to the piece. The guitars were lacking in impact when it got to the choruses, and by sidechaining to the pulse, it gave new life to them, creating an interesting throbbing effect. In electronic music, it’s important to utilise techniques such as sidechaining as it bonds the track together in ways you can’t using standard methods.
This article by Paul White proved very useful when it came to researching sidechaining. He cleverly states the difference between using SC gating and compression to create the ‘ducking’ effect (i.e. my guitars and synths are ducking from the signal caused by the kick beat).
“The higher you set the ratio, the less the amount of gain reduction will be affected by the absolute level of the side-chain signal, as long as it is high enough to cross the threshold. For this reason, a compressor with a ratio control that goes right up to hard limiting is the best choice for ducking.” (Paul White, Sidechaining in the Software Studio)
Paul confirmed for me that I was going about it the right way, as my ratio ended up at 100:1. I tried using a gate, but it didn’t have as good an outcome (maybe I was doing it wrong?).
A lot of time was spent negotiating a suitable release time and threshold position. Because the tempo remained 125BPM throughout, I could have used a timing calculator to accurately pinpoint the release of the gain reduction, however I thought it was best to just stick with what sounded good and in time.
I think more has been achieved since my last post about sidechain compression. A bit more experimentation hopefully led to placing the technique more effectively in context with the genre. The SC compressed synths work well in that part, too – It was about training the ears to relax and listen more passively for the effect, having it subtly placed. (Tip – take a break from mixing every hour or so – your ears will start deceiving you!)
Another thing I only realised until tutor Dave Hook showed me was to simply automate the vocal channel Send to the auxiliary channel with the effect. This was in attempt to delay the vocal on “Alright!” but only capture the “…right!” ending up with “Alright!…right…right…right…” which can be heard throughout from c. 1:51. This time, the delay calculator was used to build the effect in 1/4 notes, which came out at around 480ms, nicely in time with the music.
The beef I have with this track is reverb on the vocals. During the verses when they’re a bit more exposed they tend to drag the rhythm out than the tight sound I aimed for, to compliment the bassline. I thought by gating it, it would stem the tail of the reverb and it did, but I realised the problem was the amount used. It sounds a bit sloppy and there’s too much reflections happening for it to still sound in time, never mind the vocals being in yo’ face.
Here is part 1 of 2 of my mixing assessment for my Production & Professional Practice module.
Signe Jakobsen is a Danish-born singer/songwriter and I mixed her track What Have You Done To Me?. After extracting the files from Mike Senior’s multitrack library, it was good to notice others had posted their remixes the song. This became homework listening and acted as references for my own mix. I gained from ideas which I thought were useful, but mainly from parts of the mixes which I thought didn’t work i.e. too much reverb, bass too prominent.
It was my first real attempt at mixing a pop/rock tune, so I relied on Mr. Dave Pensado (the legend himself) to explain to me a little more clearly some of the key issues and differences when mixing for a specific genre:
Just look at his cheeky smile!
Here’s a wee clip of a test I ran at work on the PA. It was really useful to hear that mix on a larger system as it exposed the width I had created. I also ran the whole track in mono in order to identify issues with phase relationships and their coherence. Luckily, it didn’t suffer too greatly, however I noticed things like cymbals sounding a little mushy.
After a class on the subject, I have discovered a new love within the mixing process. Most famously used with drums, parallel compression is a fantastic tool that usually involves mixing unprocessed information with heavily compressed signals. Also known as upward compression or New York trick, it brings out some of the weaker energy signals as opposed to reducing more transient signals. At the same time, signals with more attack are sustained whilst raising the level of quieter features. It is often referred as ‘NY‘ compression simply because mix engineers in 1980s New York began developing the technique.
You can achieve this kind of effect in Pro Tools by the following steps:
- Create two stereo Aux Input tracks and name appropriately to distinguish them i.e. DRUM SUB, NY SUB
- Solo Safe these Auxes [COMMAND + CLICK] so they don’t interfere when you’re soloing separate tracks
- Select these Aux tracks’ inputs to available stereo buses i.e. Bus 1-2 and rename them as above
- Select all drum tracks and change their outputs to DRUM SUB (Bus 1-2)*
- Decide which tracks you’d like to send to the parallel compression bus. Send them to the NY SUB via Sends until your mix window looks like this:
- Remember to bring up the faders of all selected Sends and select PRE for pre-fade*. This will ensure your signals are feeding regardless of the track’s main fader position
- Add a compressor to your NY SUB track and to notice immediate effect, draw the threshold fully back and mash the ratio and make-up gain straight up (NOTE: Increase the gain gradually if you’re already playing back. Don’t whack it straight up then play, or you may become infertile)
- To avoid possible phasing issues before mixing in your NY SUB, ensure Delay Compensation is active in the Options menu
- Play your tracks back using your DRUM SUB submix. Begin to bring up the fader of your NY SUB track with compression applied. Notice the difference in sound as you introduce this. Mute the DRUM SUB track to hear more clearly what the compression is doing to the same signal.
The mix you aim for is entirely subjective, however you must experiment by ear to achieve a suitable balance of original and compressed signals. It’s very much about:
- Adding a suitable amount of the compressed track – Too much can sound unnatural and gimmicky
- Selecting what drum tracks you want involved – You may find you like the effect on the hats and overheads, however not the snare and room mic.
- Setting the compressor itself – Play around with the ratio, attack and release as the ‘thrash the pants off’ approach may not appeal to the type of recording
*Hold [OPTION + SHIFT + CLICK] when delegating outputs/zeroing faders/enabling pre-fade. It’s the ‘do to selected‘ shortcut and will save you a lot of time.
I really enjoy experimenting this technique because in the past I have often struggled to create dynamically impacting drum mixes. Using it helps to fatten the sound of drums (or bass or vocals), adding more interesting sonic content. Being a big fan of pumping (jokes aside) in dance and rock music, I particularly like the effect parallel compression has to offer.
Here are some prepared examples of using varied amounts of parallel compression:
Original dry signal:
Original dry signal mixed with a healthy amount of NY:
Original dry signal mixed with an insane amount of NY:
Leaving that drum take unmixed and pushing all the tracks in to the NY bus has made me realise a few things. The spills from mic-to-mic had horrible consequences when using the NY compression. It made it sound very crowded and undefined. By separating some tracks into their own NY buses, you can have more control over the sound each grouping makes. Even by using a mixed product, EQing some of the compressed sounds i.e. snare rumble can clean up the mix greatly. Finally, not all drum tracks need to be used when applying NY compression. Be selective and find out what works best for you.
A great way to add dynamic flow to your project would be to gate something using a DAW’s sidechain functions. Classic uses of this go back to early 1980s dance music, where you’d find kick drums controlling gate signals generated at sub bass frequencies, or a synth being chopped into 16ths from a gate receiving signal from a beat. This technique is extremely useful when keeping a track in time (which is why it’s heavily used in EDM) as it can also add qualities and relationships within the tracks you didn’t know existed. It can be used in a number of ways, methodically and creatively.
I found this video is particularly useful as it explains the process clearly how to set this up. This SOS article helped too in understanding other ways in which sidechaining can be applied when mixing
This technique is also popularly used to tighten up bass guitar tracks by gating them and feeding their key inputs with the kick drum — you can at least prevent the bass note from starting before the kick drum. Just make sure you set the hold and release controls to suit the type of sound you want. (Paul White, Advanced Gating Techniques, Part 1)
The sine kick trick in Pro Tools:
- Firstly, take the track you want to control the gate i.e. KICK
- Choose a Send (A) and select an available Bus (1)
- Select the Send and click PRE for pre-fade. Bring the fader up to 0dB [OPTION + CLICK]
- Locate the sidechain enable option and press it
- Once you’ve set the ratio, tweak the attack and release times until you have something that appeals to your taste. Adjust the threshold throughout to control the gate. Watch your times though, as you can begin to get an undesired (or desired!) crunching effect if you don’t leave enough time between opening/closing.
*Remember if you are going for a sine kick effect, you need to create an Aux track and select Signal Generator
An absolute in-sine (ho ho ho) example of this comes from amazing dance producer TEED (Totally Enormous Extinct Dinosaurs). Orlando Higginbottom (I know…) came out with Bournemouth in 2009. I first heard it from my friend Roy, through two of his twin 18″ L-Acoustics subs. I sound like a Skrillex fanboy, but, I veeerry nearly had to go to the toilet. It was ridiculous.
Anyway, I call around 45Hz sidechained to that kick beat (1:10). It sounds like it’s double-sidechained so that there’s a continuous ‘doo, doo, doo, doo…’ sent to one bus, and also a ‘swelling up’ warpy effect on another (next blog on parallel sidechaining?!)
A cool thing about sidechain gating is that by controlling the time for a gate to open, you can have one instrument trigger another without immediate effect. I’ve made the bass here swell up from the kick beat to achieve a more interesting sound. It also defines the transients of the kick beat too, and this is important in a track like this – to maintain a strong beat.
Here is a MIDI triggered bass along with the beat:
With SC gate enabled, and a slow attack to create the swelling effect:
Hear how the bass does not impact the kick beat. It is allowing more space for the beat to manoeuvre in what would be a busy mix. This swelling technique is (over)used in a lot of (‘filthy‘) dubstep production, to help towards that ‘wob wob wob’ effect.
Using a tone generator can add extra character to an existing signal. Here I have routed the bass into the gate plugged in to the tone frequency track. Without the gate sidechained, all you’d hear is the tone playing throughout, regardless what the bass is doing.
Here is the tone gated:
Just the beat and synth:
The beat with sidechain gated synth (with the original track thrown in):
I gave all this a spin in Studio A at uni. Sorry for the phasing, I should really get Elluminate Live or something…
These are all extreme examples, however when used subtly and by creating track relationships it can definitely improve the sound. It opens up so many more options, and can appeal to just about any genre. It completely depends on who’s mixing and for what style, so it’s a matter of balancing a sound which appeals to the rest of a track. When using the gate controls in more depth, you can also notice creative changes to the entire feel of the music.
You may want to view the sister post to this entry, Sidechain Compression, where I run over similar issues but with the power of using compression.
I’d like to discuss a very interesting technique known as sidechain compression. What this method entails is the ability to control a compressor by the feed of another signal. There are many uses for this in mix production such as using the transients of a kick drum to compress a heavy synth to create a pumping effect. In doing this, each beat from the kick will compress the synth, squashing it down for a desired time, until the compressor allows the synth back up to level.
A step-by-step video demonstrates simply here how to achieve such an effect:
- Firstly, take the track you want to control the compressor i.e. KICK
- Choose a Send (A) and select an available Bus (1)
- Rename the Send appropriately i.e. KEY COMP
- Select the Send and choose PRE for pre-fade. Bring the fader up to 0dB
- Apply a compressor the to track you want compressed i.e. SYNTH
- Locate the sidechain enable option and enable it
You will notice that the sound of your compressed signal may have changed. Experiment with the threshold first to find at what point the Send signal (KICK) is compressing the affected track (SYNTH). Play about with ratios, attacks and releases to find a desirable level of compression. The affected signal can change dramatically with subtle changes to the plugin, so get used to what function serves what!
For example, here is a compressed hi-hat track being controlled by the snare track signal:
It’s exaggerated so you can hear what the hats are doing. They’re ‘ducking’ to allow the snare transients to stab through, creating the illusion of it becoming louder. This can benefit tunes where the snare really needs to poke through the mix, as it momentarily decreases the impact of the hi-hats
The following examples are extracted from a session for my Production & Professional Practice module at Edinburgh Napier University. They are all associated with the track Word I Never Said – Lupe Fiasco, produced by Alex da Kid.
This is just the beat and synth together, no trickery applied:
Compressed synth sidechained to drums signal, listen for when drums are muted:
You can notice a definite change in the synth information. It becomes more dynamic, and for that reason it allows the beats to continue with their solid rhythm, as the range of the synth fluctuates. Hear when the drums are muted, you can sense the whole character of the synth has changed. It takes a lot of experimentation to reach a desired effect, and you need to make sure it fits the music – gimmicks will ruin a great track.