The piece that I have been working on most recently is Poker Face by Lady Gaga. The track was originally recorded in first year as part of an introduction to recording methods and as a result the quality of the audio files and recording/mixing techniques are rough at best. I decided that this presented a great challenge for me to try and show not only the progress I have made over the last few months but also that of the last 4 years at University in terms of mixing and production.
Upon reopening the session after 3 years my first decision was to rename all of the tracks to something more familiar to me, as well as cutting and deleting all regions with silence. The tracks were put into groups depending on their instrumentation and also stemmed via stereo busses to separate ‘sub’ mix, stereo aux tracks – something which I have done religiously since the technique was made aware to me during my third year. The drum tracks featured kick, snare top and bottom, floor tom and overheads. All except the overheads were treated with subtle EQ to ‘warm them up’ as they were all sounding very thin and also had a few unwanted overtone frequencies removed. The kick drum was mildly compressed to level the transients making them more consistent, without overdoing it, as I had planned ahead to use some parallel compression later on. To do this the kick drum and snare top and bottom were bussed to separate aux channels and fed with heavy compression which was blended with the dry drum sub to create the parallel compression. This can be seen here on the left:
The snare drum required a lot of work to try and make it sound even reasonable a I set up another aux channel and bussed the snare top to it and gave it some very aggressive EQ to give it a very different texture. This was then blended with the dry track and the snare bottom track to help enhance the sound of the drum. This can be seen here to the right:
With the overall kit sounding much better I treated the drum sub with the Oxford Equaliser & Filter plug in to brighten the whole drum kit as it was beginning to sound a little muddy behind the other instruments with all of the alterations, as without them the whole kit was thin and ‘tinny’ despite cutting through.
I moved onto the bass guitar next. It required a little bit of elastic audio treatment to align one or two of the transients with the click as it played slightly out of time in sections. By placing the event markers and shifting the warp markers forward this issue was resolved. Both the bass, acoustic guitars were all mildly compressed and only required subtle warming with EQ and the AIR Enhancer as their audio quality was surprisingly good. I also found that the electric guitars responded extremely well to the same treatment, only this time I had to be very aggressive with the EQ, left, to completely reshape the sound and mellow it significantly. With this complete they were re-amped through a PSA-1 to create their final tone.
All of the female vocal tracks (melody, harmony rap and counter melody) were given moderate compression and EQ. The rap and counter melody were later given more surgical EQ to deliberately alter the sound for effect, right.
Reverb was later added to the female vocals to help it blend a little better with the rest of the track, but by doing so they also needed a De-Esser to remove the undesired frequencies.
The male vocals were heavily compressed as they were very quiet in the mix and had to be brought much further forward, as was the intention for effect. The three male vocal tracks were panned left, centre and right to provide a full stereo backing vocal, while also taking phasing into account should it be mono-tested. A gate was required on one of the male vocal tracks because it had two vocals on it, over-ran significantly and needed cut out without affecting both vocals.
Finally the track was mastered using the ‘layback’ technique taught in Recording Studio Theory & Practice 3. All of the sub stems were bussed to a separate aux and the mastering plug ins on the master fader were copied to the aux fader. Next the aux track output was sent to a new stereo audio track which was record enabled and by hitting record, the session was ‘laid back’ to this audio track in real time.
The main issues encountered with this song were the electric guitars battling with the vocals for space and similarly with the male vocals and bass guitar. These issues were solved by chopping space with surgical EQ and panning the male vocals away from the centre to allow the bass to sit in the middle of the stereo field. The tone of the guitars was horrible at best and this has been resolved with similar EQ and re-amping. In addition to these instruments, the drum kit needed completely reworked. I had looked into the possibility of using triggered midi samples to layer onto the existing live drums. However, because of the tone of the overall track and the instrumentation I decided that such triggering would potentially make the drums sound artificial and quite jaded alongside the rest of the instrumentation.
Having been presented with the session ahead of taking it on as the assessment it was decided that the production should follow a similar guise as the original non-produced version. Instead of potentially creating a dance track out of the raw audio files I have kept i more like a ‘Live Lounge’ version, something which its tongue-in-cheek’ nature is suited to. It is worth, however, exploring the possibilities of midi triggering with live sounds. A short example can be heard on Dave Pensado’s Video from one of my previous posts.
Over the past few weeks the Production & Professional Practice module has provided me with invaluable knowledge and experience with which to take forward to future recording and production sessions. Initially some of the techniques such as parallel compression and side-chain compression were great tools to play around with and experiment with how far I could take them to create different sounds. But as the module has progressed I have learned to use them more subtly and, through practice, how to really enhance the sound of an instrument tastefully. One thing I have completely taken on board and will continue to use in every session is stem mixing. I have found it to be a most useful way of organising my sessions and mixing more tastefully and have found a new appreciation for the sonic differences between minutely variable mixes.
I thought I’d give an update into what I have been up to in the studio recently. As part of my Major Music Project I am recording singer-songwriter Kyle Farrell and will be producing a 4-track EP for him. His music is heavily acoustic based and this is something which I will be focusing strongly on during the production. With this in mind I have decided to make use of the middle and side stereo widening technique to really stretch the stereo field given that the acoustic guitar will be so prominent in the mix. At this stage the sessions are still in the recording phase but I have a couple of examples from my sessions to show just how effective this technique can be.
The microphones I have used for the acoustic are a Neumann TLM102 condenser for the ‘middle’, an Electro Harmonix EH-R1 ribbon for the ‘side’ and an AKG Perception 400 as a room microphone and also to record the guide track which at this stage I am keeping.
This short excerpt shows how the opening bars of the session sound without enabling the middle and side technique:
Now, by duplicating the ribbon track, flipping the phase on the duplicated track and panning the ribbon tracks hard left and hard right, the following results emerge:
Other than this stereo technique, no other production has taken place. With this in mind I am very pleased at how the session is sounding so far!
As the middle and side stereo technique will feature prominently in my Major Music Project recordings I thought I’d give some brief insight on it. Middle and Side is essentially a way of widening the stereo capture of an instrument, in my case, an acoustic guitar. Two microphones are used: one ‘middle’ (usually with a cardioid pattern) and one ‘side’ (a figure of 8).
These two audio tracks on their own will not produce any results. In order to turn them into a wide stereo pattern the ‘side’ track should be duplicated. The phase is then flipped on the duplicated track which should silence them both as one is completely out of phase with the other and will cancel each other out. The trick is to pan them hard left and hard right. What you will now hear is what has been picked up by either side of the ‘figure of 8’ pattern, stretched fully left and right. Bring in the ‘middle’ track to fill in the gap in the centre and by adjusting its volume you can vary how wide or how narrow the stereo field appears. While researching the technique a little further I came across this great tutorial video which practically explains how to set up and implement it and also gives audio examples throughout:
Not that I plan to use it for my project (yet!) but I also came across a very intriguing video on Mid-Side plug ins and their use on bounced WAVs during the mastering stage. Check it out!
Following up on my last post regarding the North to Alaska track, I also completed a remix of the Halo 4 soundtrack as part of the same assessment presentation. Compared to the North to Alaska mix this initially seemed to present much more scope for experimentation. In my 2nd year at University I had produced a dance track as part of a Music Technology module using Logic Pro so I found it an intriguing prospect to produce something similar using Pro Tools.
To begin I had the original stems including the drums, orchestral instruments, percussion, piano and synthesiser. I also had an array of drum breaks at my disposal. The first thing I did was to listen to the stems as a full mix to get a feel for the track and then dissect each individual stem and pick out the sections I liked. Once I had decided on a few sections with good hooks and rhythms I cut them from the stems and created new audio tracks for each. I determined the tempo of the entire track and set the session accordingly in ticks as I would be using the grid. The process was then repeated for the drum breaks, choosing the ones which I thought could prove useful and would compliment the sections of the Halo mix that I had cut out. With the grid in place I used the elastic audio function of Pro Tools to stretch the drum breaks to the same tempo as the session so that I could copy and paste any of my breaks around the arrangement page and they would fit perfectly. I was then a case of building the remix from the selection of regions I had given myself.
Arranging the regions into something useable was the easy part. Turning it into a good product was the fun part! I created new tracks featuring a bass line, extra synth and a sweep using the XPand2! plug-in. I side-chained the bass line to the synth which created a dotted pattern between the instruments once I compressed the synth – making use of side-chain compression which was discussed in a previous post. I created a similar effect between the bass and string section which gave a nice pulsing relationship between the two.
As with the North To Alaska track in my last post I also side chained a sine wave to a midi kick drum to boost the lower frequencies and really get the song pumping. I coupled this with two additional, yet very different, kick samples to give the kick some top end punch. Finally I programmed a sweeping synth to help each section blend together a bit better. One aspect of this track is that the sections are quite quick to change and the sweep minimises any jarring.
With this track I wanted to try and limit myself with the number of options I had – in this case the stems and additional midi – to try and enhance my creativity. By focusing on developing the production on the original stems and the derivatives thereof while restricting the number of programmed midi tracks I feel I have achieved something close to that aim.
As part of one of the assessments I sat a few weeks ago in my Production & Professional Practice module I have produced a mix of ‘All the Same’ by the band North to Alaska. Initially it was an in-class task but as I spent some time on it I decided to use it as part of my production presentation. With just the original raw audio files to work with it was a case of importing them all into a new Pro Tools session and making a start. Here is the finished article, whether you wish to listen while you read or read afterwards:
The first port of call was to try and make sense of the tracks, so all tracks were renamed to something which I found easier to use – kick, snare, toms etc – rather than audio tracks numbered from ‘1’ to infinity. So that I could begin routing the tracks to aux subs I firstly grouped them into their instrument types, mainly drums, guitar riff, power chords, guitar left, guitar right, guitar solo and vocals. Once these were established I then began routing the outputs of each track to similarly named stereo aux tracks via stereo busses.
With all of my routing and track titles as I wanted, I then set about turning the session into something sonically pleasing.
The drums were the tracks that I tackled first. The sound of the kit when soloed was already sounding good before any tweaking took place. So apart from some mild compression on the kick, EQ to boost the low-mids and some surgical EQ to dampen the ring from the snare drum not a lot needed done. The main production components for the drums were the addition of a sine wave, side-chained to the kick drum to give it more punch due to the song being so guitar-heavy and potentially lost underneath them. A mono bus to a new aux track and a gate to control the signal generator from the kick put this in place.
I also used parallel compression on the kit. From listening to the song though, it was clear that the level of cymbal ‘wash’ meant that only kick and snare should be routed to the compressed track. Anything else would muddy the sound a little. So the kick and two snare tracks were stereo bussed to a new aux track and a heavy amount of compression applied.
To give you an idea of the enhanced sound the parallel compression creates here are two short examples:
With the drums sitting nicely I turned my attention to the guitars. I added a chorus effect to the opening guitar riff and also fed the second guitar through a guitar amp plug in and smothered it in reverb to make it sit back in the mix and sound a bit less like a DI. The chorus guitars were causing most of the problem in the mix as they were simply too overpowering and producing troubling overtones. This was resolved with some very aggressive EQ-ing to settle them down which produced a very altered but tasteful sound. As there were four tracks of guitars and their respective DI signals I created a 2+2 set up – two guitars plus their DIs going left and the same on the right – and panned them around the stereo field.
I used a similar technique when mixing the vocals. The chorus features four separate takes, all usable, so I panned them around the stereo field to widen the chorus a little. The verse had two vocal tracks and were panned hard left and hard right, ensuring compatibility with mono-testing. Compression was added to bring all the vocals forward in the mix with the verse benefitting from slightly more compression for effect.
The closing sections which feature the guitar solo were also causing frequency issues with the vocal part. The guitar solo needed EQ-ing to allow space in the mix for the vocals without the two battling for space.
With all individual sub groups mixed, the stems were then adjusted according to preference. The main consideration in the session was not to allow the guitars to control the mix during the chorus and still allow the drums and vocals to punch through. I feel this was achieved although in hindsight the track maybe feels slightly more ‘pop’ than ‘rock’. It was my intention to have it this way as an experiment in keeping heavy guitars at bay.
Leading on from my last post about stem mixing, I thought I’d upload this video of Dave Pensado explaining some processes involved in mixing instrument levels. It’s short and sweet this time but I figured it gives great insight into the issues that can arise and what considerations must be taken when mixing levels of instruments, and in particular, the levels of several stems together.
Hope it helps!
Stem mixing or ‘sub mixing’ is a useful tool for the producer or mixing engineer. Essentially it allows the mixer to consolodate tracks into single ‘stem’ groupd making it easier and more efficient to mix sessions with large numbers of tracks.
To set up a session using stem mixing you should first decide on how to group the tracks. For example drums would consist of kick, snare, toms and overheads. In the mix window route these individaul tracks to a new auxiliary track by selecting the output from the I/O and chosing a new stereo bus. These tracks can now be controlled from the new single aux track with respect to volume, panning and any extra plug-ins. By doing this with similar groups sucj as guitars and vocals the session can be controlled by a handful of aux tracks. If the tracks being routed to these ‘sub’ tracks are also linked with the group function it makes it easier to navigiate the mix using faders, solo, mute etc, and it is also a good idea to ‘solo-safe’ each aux track. Here are a couple of images from a recent session which show my sub routing. Notice I also include some parallel compression on the drums and route it to the drum sub aux as well:
When it comes to actually mixing the subs make sure that each individual group is ‘internally’ mixed. Once this is done it saves time giong back to tweak things here and there while mixing the stems. This technique allows each group to be subtly adjusted in volume and panning (if not already panned internally) using single aux tracks.
I first came across this technique during my 3rd year at University when completing my 5.1 surround sound project. My task was to convert a stereo track into 5.1 and before I could do so I had to dismantle the stems so that I could route the individual instruments to the 6 different outputs.
Effects can be added to both the individual tracks and the stems. Personally I prefer to add my plug-ins pre-stem or bus them to separate aux tracks and blend them with the dry ones and reserve the stems purely for volume adjustments. It is all down to preference.
Parallel compression is a great tool for engineers to use to enhance the sound of an instrument, or ‘fatten’ it up, without losing the important transients which give it attack and feel. Also known as ‘upward compression’ it brings up quieter sections and really beefs up the sound.
As a drummer I have often found that some recordings and mixes that I have worked on have lacked the sound I was wanting to achieve with the kit. This can be particularly noticeable with the snare drum which often loses most of the ghost strokes beneath everything else that’s going on. If I was to pile a compressor on the snare it would lose most of the attack from the initial transient. Parallel compression gets the best of both worlds.
If you bus the drum tracks (usually only kick, snare and toms to avoid unwanted ‘wash’ from the overheads when compressed) to an auxiliary channel and add thick compression to it you can then mix this sound in with the dry drum tracks. At this point I should point out that it is useful to send the outputs of all the individual drum tracks to a stereo auxiliary channel so that you are working with a single stem for the drums – stem mixing will feature in an upcoming blog. This simplifies the process by only having two auxiliary tracks to work with when blending the dry and wet signals, which should be mixed by ear. And when doing so the blended sound retains the attack and fast transients while also gaining some of the quieter sounds from the compressed track which give it a much larger sound.
To help explain in a practical sense I have found this great tutorial video detailing the process:
I thought I would post some information about more techniques we have been learning. Side chain compression is a technique heavily used with dance music. Because it is again ‘side-chaining’ it follows the same practice as side-chain gating in that one track controls another through key inputting. Essentially, the send from one track influences the compressor on another.
One of the benefits of this is to give more feel within the track and to get tracks pulsing and complimenting each other. For example you could have four on the floor with a synth on the off beat, bus the kick to the synth’s compressor and instead of a straight question and answer between the two tracks, the rhythm can be adjusted to create a dotted rhythm or simply just space between them. By playing around with the attack and release you can adjust how much space you want or how much ‘whoosh’ to build in the synth.
It is also used frequently away from dance music particularly to make room in a mix as I mentioned above. When you play around with the gain and threshold you can create more of a relationship between two tracks by having one track hide a little and let another have more prominence in certain points in the mix. This could be between vocal harmonies, a bass/kick drum groove or even a guitar solo to find room within chords. But in most of these cases it should be used very subtly so that it doesn’t create a disjointed effect and still allows tracks to flow and not stutter.
I will be uploading more on this subject when I reveal this process in a practical example.
Over the next few weeks we have been delving into new production techniques as part of the Production & Professional Practice module at Edinburgh Napier University. I thought I would begin by talking about side-chain gating as I have some experience of using this technique with a previous University project. Particularly useful with dance music and also to enhance the kick drum within a mix.
An auxiliary track should be created and the kick drum track sent via a bus. Next, add a signal generator to the auxiliary track and set a sine wave to 50-60 Hz, the same as a kick drum produces. Add a compressor/gate plug-in to the aux track and activate the ‘key’ icon, followed by selecting the ‘key input’ icon and choosing the bus which the kick drum track is routed to. Mess around with the gain and threshold depending on taste. The kick drum track will now trigger the sine wave and the overall sound of the kick will be enhanced. The following is the assignment I worked on which features this technique:
Something to note at this point is that it will be useful for some projects to create two auxiliary tracks, one for the sine wave and the other as a duplicate of the kick drum track or as a separate instrument track with a programmed kick. This gives the advantage of not being restricted by the original kick’s pattern when developing the sine wave, as you can create any kind of sine wave pattern with a programmed midi kick.