Key Of Grey

November 13th, 2009 – Hmmm, Friday the 13th is not a particularly auspicious day to try something new, but I’m kind of excited to try this out. Late last night I was listening to some of my old stuff in an attempt to figure out why some songs I’d worked on were good, and why some weren’t. During that process, it dawned on me that some people out there on the interwebs might want to know how and why I did certain things to a track.

So this post will be the first in a recurring segment on this site where I examine actual songs that I’ve produced and step you through the process. Or at least as much as I can remember anyways. The particular track we’ll be looking at today was created almost 5 years ago. You may not be a fan of the style, or the particular song, but maybe there’ll be something interesting in here that you can apply to your own music.

Lastly, I’ll go over some of the things that I’d change if I were to work on the track today. Without further ado, here is today’s track entitled “Grandfather Clock” :

Backstory

First, a little history on the track and why it was recorded. At the time, I was working at the small studio you see in the picture above (ugh…I wish there was one without me in it I could have used instead). My job was to write, produce, and record songs for clients that my boss would bring in. Most of the clients were teen girls that wanted to sound like Kelly Clarkson or Avril Lavigne (remember, this was 5 years ago). As much as I liked writing songs, I hadn’t really written anything for my own enjoyment.

This song came out of my attempt to get away from the genres I wrote for my work, and something that I’d want to listen to on my downtime. At first, the song was meant to be an instrumental. However, as it progressed, it became clear that the song wanted vocals. At the time, I had only worked with one hiphop lyricist, and he seemed like a good fit for the song. Along with him came a friend of his that could sing the hook. For those of you who are wondering, the song is half in Japanese and half in English.

Creating the Instrumental

The instrumental on the track was created using Logic 7. The first thing you’ll notice, is the vinyl static. Although you really only notice it in quiet sections, it loops and runs through the whole song. Modern hiphop is a little too clean for my tastes so, depending on the style I’m going for, I will sometimes simulate the old school flavour using techniques like this.

The kick, snare, and high hats are sampled from the Korg Triton’s “hip hop” bank. They are run, and EQ’d, through a Universal Audio 2-610 pre-amp in mono. From the UA 2-610, the audio is converted using an Apogee Rosetta AD converter that then routes via SPDIF into a DIGI 002.

Through Protools, they are then routed through a Yamaha O2R digital mixing console where additional EQing, and compression are added. The digital audio is routed out back through the Apogee converter and then into a Sony CD burner. Phew…that was a super convoluted path. Why did I go through the bother?

Well, I write almost all of my music from home, and the studio was where the nice gear resided. The drum samples I had created out of the studio sounded fantastic, so I needed a way to bring them home with that quality. Once I got them into Logic I arranged and copied the waveforms manually to create the beat. Smart people import the audio into samplers and then play them with a MIDI controller, but for some reason, even though I’m a musician, I prefer hand arranging drum samples.

The rhythm instruments consist of a traditional Japanese string instrument processed using an EQ and reverb, as well as the EVP88 instrument in Logic routed through a stereo delay. The reverberating bass frequencies are handled by samples of a Japanese bowl, which is an instrument I’ve never seen, but it sounded cool. There are also soft horn punches to accentuate the chorus.

That’s it for the instruments. When I first started writing music, I used to add way too many instruments. It cluttered the production making it sounded muddy. Worst of all, it didn’t leave any room for the vocalists to fit. At the time I wrote this track, I had almost gone “zen” about the way I was producing things utilizing as simple arrangements as possible.

The Vocals

We went back to the studio to record the vocals. They went through the same signal chain as the drums, except that I did all the editing and cleaning up of the recorded tracks in Protools. I can’t remember the exact microphone but it was a pretty simple AKG.

For the vocal arrangement, we recorded several takes of the whole song. As we recorded, I noted which sections I liked. After the leads were recorded, I quickly comped together the best parts of each take. We then recorded accents. These are sections of the lead vocal that are sung exactly the same, but layered to thicken them up. It is a technique that is immediately obvious in hip hop music, but not so much in pop music. However, in every single released on the radio, regardless of genre (even folk), it is used.

You might accent certain words, or you might accent the last couple of words for every line. You are effectively, selectively doubling the lead vocal. You need to do this at least twice for each accented section. This is because in the mix, you will leave the lead vocal centre and pan the doubles each to the left and right. When producing pop songs, I’ve used up to 6 tracks of doubles to create 3 pans on each side, each at different pan positions. Trust me, once you use this technique, you’ll always use it.

During the chorus, you will want to record as many doubles, harmony, and doubles of harmony as possible. You want as many possibilities to work with when mixing. The chorus is the part listeners will remember the most, and since they will be repeated, you want them to have the most impact as possible.

After all the doubles/accents are recorded, I will quickly comp and clean them up, and move onto recording ad-libs. These are the little vocal lines that make the vocals interesting. In R&B/pop music they can be the vocal “riffs” or “runs” that you hear or they can be something as simple as a little giggle in response to something in the lyrics. Whatever they are, they separate amateur vocal arrangements from the professional.

Even an artist like Damien Rice who records live off the floor with very little “production”, adds these bits of vocal interest by playing with where he comes in on the beat, or a heartfelt cry during an instrumental interlude. When you are producing someone else, especially an artist who isn’t experienced in the studio, you want to record these last. This is because they are often the most nerve-wracking things to record for someone who isn’t used to doing them.

You want all the “meat and potatoes” of their performance recorded before they head to this so they are excited about finishing and are more relaxed about the process. Also, being able to hear the rest of their performance, will give them a scaffold to sing around.

Mixing

I prefer mixing in Logic because it’s comfortable for me. You can do many of the things I do in Logic in Protools. I like mixing in the box because I don’t like using physical faders to set track levels. I like the precision that drawing automation curves provides. Even when routing through an SSL console, I would rather automate the volume changes through the DAW. I like big knobs and switches for pre-amps though. I know, kind of contradictory.

Since Logic wasn’t available in the studio, I brought the vocals back to the original session on my home set up. To do this, I pre-mixed the vocals. By this, I mean I mixed only the vocals routed through the Yamaha O2R. I added my EQ, compression, and reverb through the console, and then recorded 3 separate tracks to CD : lead vocals, doubles and harmonies, and adlibs. Each of these tracks were stereo.

I re-imported them into Logic and further mixed them with my instruments. I hear some of you cringing at the thought of pre-mixing the vocals. While it’s certainly true that I wouldn’t be able to go back and fix a mistake in Logic, I was careful and there weren’t any glaring errors. Besides, if something was really bad, I could always have gone back to the studio and fixed it.

Some might wonder why I even bothered to bring the vocals back into Logic. If I pre-mixed with Logic, and I had already mixed an instrumental track, why didn’t I just mix those together in the studio and be done with it? Well, I’m a little picky. It is my opinion that when tracks are mixed with vocal and instrument pre-mixes, there is a noticeable separation between the vocals and instrumental, almost like a karaoke recording.

Sure you can send both to an auxiliary track and apply compression, EQ and reverb to the sum so that they sound better, but I prefer being able to independently mix the relationship between each individual instrument and the vocals. I believe that the mix sounds better, although I’m sure some would disagree. Maybe it’s all in my head.

Mastering

Because I don’t have the money to purchase specialized mastering gear, and the studio I worked for didn’t have any, I master mostly in Logic. Sometimes I try out the mastering presets (you can find them on the master fader). Often they don’t sound very good and fight my mix, so I’ll use three different plugins.

For making small changes to the EQ overall (any big changes mean I need to go back and examine my mix) I use Logic’s Linear Phase EQ. I will sometimes use a Multipressor to selectively compress certain frequencies if I feel the track needs to tighten up a bit. Finally, I use the Adaptive Limiter to bring up the gain. I tend to like a lot of dynamic range, so many of my tracks won’t be as loud as they could potentially be. I always find them loud enough, and if I want them louder, I figure I’ll just pay a professional mastering engineer to work with the track.

In Conclusion…

Wow, I knew I’d write a lot, but I didn’t expect to write this much. In the future, the explanations will be shorter as you get to know my process and I don’t need to provide as much primer. I’ll probably come back and edit this article as I remember things, or as I read over it in the future.

Going back and reviewing older tracks is something that I do quite a bit when I’m writing new music. Knowing what works and what needs improvement in your own music, helps get you psyched up and prepared for the new task at hand. You’ll also save time by remembering how you did things before.

Even after all this time, I like this song quite a bit. Of course, there are some things I’d go back and change. For one, I feel like the vocals are too quiet overall. This was likely due to a my bias; I wanted to hear all the cool little things I did musically, so I mixed with the music being more important. The music isn’t wildly overpowering, but if I was mixing this for a wider audience, the vocals would definitely need to come up.

The vocals also need to be evened out a bit. The rapper was moving quite a bit around the microphone so there were times when he was a little far from it. I could compress the vocals a little more to compensate for this or I could manually automate gain changes. Since he does it quite a bit, I think compression would be the best bet.

I also would need to work on the Japanese bowl. While I like its overall volume and the balance it brings, it seems to be causing the compressor to pull gain on the vocals.

Do you have any questions about the process I used? Have you done things another way that worked well?