Notes from the Shore

Work, marriage, kids and music

Song Production from Begining to End

There are quite a few steps that make up taking a song from initial concept to completed production, ready for public consumption. Since I don’t have any real progress on my own music to report, I thought I would entertain myself outlining those steps for potential recording songwriters.

Writing. The first step is simply writing the song. This consists of taking an original concept, either lyrical or musical, and developing it into a cohesive whole. The result is usually a lead sheet, which consists of the melody line, harmonization (usually in the form of chord symbols), and lyrics (if any). Sometimes a piano reduction outlining important harmony lines and forms will be included. If you have much more than this, then you have probably crossed the line into…

Arranging: This next step involves taking your basic song and figuring out what will play when. Will the intro be fingerstyle guitar? Solo piano? Will the drums count off and the whole band jump in full scale? Where will the strings play? What will the strings play? How many times to repeat the chorus at the end? These are all questions that get answered during the arranging step.

Tracking: Here’s where the fun begins for many. Tracking is the process of recording the final performances of each part settled on during the arranging process.  Sometimes a band will track the entire band at the same time, all on different channels. Not many home studios are set up to do this. There are a number of interfaces, though, that will get you 8-10 channels at once, such as the M-Audio 1010LT or the Alesis iO|26. Recording multiple instrumentalists at once can capture that “live” feel, as the musicians play off of each other.

If you’re recording alone, then playing all the parts at once isn’t really an option. When I track, I start with the primary rhythmic/accompaniment instrument, which is almost always the piano. I usually play this in free, i.e. not to a metronome, and then go back and sync the tempo to the performance. Then add in the other parts one at a time

Now, this is important. Many home recordists tend to blur these first three steps. They are arranging and tracking while they’re still writing the song. Arranging while you’re writing isn’t a huge sin, but I promise you that you will get better results if you keep things separate. If you need a scratch arrangement while writing to keep the song moving forward (which I do, sometimes), don’t marry yourself to it! Have the freedom to junk the arrangement and start over if you decide to go in a different direction.

Comping: Comping is sort of a sub-set of tracking. It doesn’t refer to accompanying a jazz solo, but rather constructing a single track out of multiple takes – a compilation track. I was doing this just last night on a vocal track. In my case I had four vocal takes to choose from, and I selected the best verses and choruses (sometimes just the best line) from each take. When I’m done, I’ll roll these all into a single track so I can easily apply the same effects and volume envelope to the whole performance.

Mixing: Mixing is an art as well as a science. Much has been written on-line and in innumerable magazines about mixing, but at it’s root mixing is about making each member of the band work with the other members of the band. This often involves changing volume and equalization so various instruments don’t step on each other’s frequency spectrum. Modern DAW’s also allow for automation of various parameters, including volume, which will let you emphasize certain aspects of certain tracks when it’s musically appropriate. For more reading on mixing, look here.

When I mix I use track envelopes to set and automate my levels. I know some people really prefer a control surface with sliders for every channel, but I didn’t cut my recording teeth in the analog era, so I have no emotional connection to physical faders. Plus I make significant use of automation, which I would have to record in real time if I used a control surface.

I will generally tackle a song one section at a time. I’ll set my levels for each instrument for the introduction. My mixes so far have been simple enough that I usually don’t reach for EQ. I generally only use EQ if I can’t get some aspects of the mix to sit nicely together no matter what I do to the levels. A good example was a recent mix with several instruments that had too many low frequencies. I chose which one would carry the bass, and rolled off the rest with the channel EQ. It cleaned up the mix nicely.

At the transition between each two sections, I’ll drop two adjacent nodes onto the volume envelope. This way I can grab the volume anywhere in the section and adjust it up or down without messing with the volume in adjacent sections.

Effects: I don’t mix with effects until the very end, unless that effect is fundamental to a particular sound, such as a delay. The other exception is channel compression, which I will apply as necessary. I recently had a recorded instrument that was just too wildly expressive to sit in with the other instruments. So I compressed it a bit until fit well. The Sonitus Compressor that comes with Sonar is really nice in that it has a graphical display of the compression curve and where the signal is on that curve in real time. You can see exactly what’s happening to the signal. Anyway, channel compression should be applied while mixing because it changes the way the channels relate to each other, which is what mixing is all about to begin with.

Last I add my reverb in. Since I record a lot of acoustic music, I usually only use one reverb on the master output. I especially like SIR convolution reverb, and use the 40×60 or 50×80 live room impulses from Noisevault. I add in just enough reverb so that I can hear the effect, and the back off about 3 dB.

If you’re recording a pop song, then the application of reverb becomes much more complex, with different reverbs on voice, drums, guitar, etc.

Mastering: This is the most controversial aspect of this list. I believe there are now two types of mastering. The first is mastering in the traditional sense. I’ve heard it best described this way: You mix a song with tracks, you mix a CD with songs. In other words, mastering is all about making the songs work together on the CD the way mixing is about making the tracks work together in the song. The primary tools of this kind of mastering include very high-end EQ’s and compressors. Occasionally other sorts of effects and signal processing are used, but pretty rarely.

The other form of mastering is a result of our download-a-track-at-a-time world. Tracks are not often heard in the context of an entire project any more. This is something that makes me sad – the single died in the 80’s with the advent of CDs, and the CD-length project is dying in the 00’s because of the downloadable single.

This type of mastering doesn’t care about the context of the entire project, but rather the context, or more correctly, the competition, of the entire music world. Making the song stand up against other loud songs is the goal. While many people (especially Mastering Engineers) rail against this practice, it’s a simple fact of life that if you want radio airplay, you have to bring up the relative volume. I’m toying with creating not just a radio-edit, but a radio-master in addition to the full dynamic version on my CD.

1 thought on “Song Production from Begining to End

Leave a Reply

Your email address will not be published. Required fields are marked *

Enter Captcha Here : *

Reload Image


Copyright © 2019 Notes from the Shore