RECORDING THE SIMPLE SONG DEMO
Farm to Market Recording Two Tracks At A Time
By Greg Forest
I received quite a few emails regarding my review of Reaper last month. A recurring theme in the messages was a request for a walk through of a basic songwriter demo. There are many ways to skin an audio cat – each producer and engineer has their own methodology that reflects their experience and gear. Here’s the process I use to spit out a quick song demo on an easily attainable 2-track recorder.
IT STARTS WITH YOU
If you are the songwriter, you know where you are trying to go. You can hear the finished demo in your head before you have recorded a single note. The fastest way to get a tune from brain to CD is to play all the parts yourself if possible. As we are talking about a song demo, getting a super slick and sometimes overproduced recording can sometimes be counterproductive. Keep in mind as you build the musical complexity of the demo, you close off avenues of interpretation to the listener. The more complex and slicker the arrangements, the more unlikely that you will appeal to the person you are pitching to. Keep your demos simple, short and sweet. When crafting the 2-minute blazing guitar intro, keep in mind the old Nashville publishing adage, “Don’t bore us. Where’s the chorus?”
Looking at the entry level for audio input and output I am focusing on the M-Box 2 which is reasonably priced, small enough to throw into a guitar case, and comes bundled with Pro Tools and a few plugins. The M-Box is your most basic input/output (I/O) with only two balanced/unbalanced inputs record simultaneously enabled. You can stack up dozens of tracks – but only two at a time – so a bit of advance planning is in order.
I start with the tune itself. I have always found that things go much faster if you know the song you’re dealing with. If it is something I have composed, that is pretty easy but if it is someone else’s piece, I sit down and chart the song measure by measure using the Nashville Number Method (see my earlier article about the system). Knowing what is coming up the next few bars of a song make things go pretty quickly.
BRING IN THE ROBOTS
As a matter of course I record just about every studio session using a click track or metronome. Keeping consistent time at the beginning of the session makes things go much more quickly and leaves you many more options later in production. If the song is at a uniform tempo throughout the performance, you can cut and paste verses and choruses to your heart’s content and still keep everything in time. If the tempo drifts during the song, you won’t be able to easily move segments around without having to use some time-tempo stretching widget which will extrapolate the tempo over time and “fix” the timing irregularities. It is much easier to just cut a basic track in time.
My first robot on the job is my trusty keyboard – an eight year old Yamaha PSR-730. The keyboard has a good palette of drum patterns in most musical genres. The reggae patch, for example, may not be exactly what you are looking for but it will give you a basic groove to build on if that is the genre you are working with.
I plug the right and left output from the keyboard into my M-Box channel 1 and two. I find the rhythm pattern patch I want to use then modify the tempo until I have the beat at an tempo that is easy to perform to. When I figure out which patch I’m going to use and how many beats per minute the song is rolling at, I make a note of if either somewhere in the software or just jot it down somewhere for later reference if necessary.
Since I know the song from start to finish, I get the ball rolling by recording the Yamaha drum track first. I get the pattern setup on the keyboard and punch the play/start button. Using the knobs on the M-Box, I adjust my input levels until they are nice and strong without clipping. When all the levels are set, record enable the first two tracks, start the recording on your computer then hit the start button on your drum machine or keyboard.
I let two bars of the pattern go by so I will have a long enough count-in for later tracks. Then I sing the song while the drum machine does its thing. The drum machine can also act as a “meter minder” for you. If a section of your song is not in time, it will be glaringly apparent pretty early. Drum machines and keyboards set up their patterns in 4, 8, 12 and 16-bar segments. That is how pop music works generally. If you’re song sounding weird against the drum machine pattern, you might look at your song’s timing. The two beats you drop from a bar might seem cute but you just put somebody’s feet on their dance partner’s toes. Generally speaking if there is something wrong timing wise when using a robot drummer, the fault lies with the mapping of the song, not the machine.
As a chorus or other dramatic change comes up, I hit the “fill” button on the drum machine to give me a kick into the next part of the song. Go all the way through the song to the end and then hit the stop button on the drum source and recorder. You now have your basic drum track as a foundation to build on. In our new found buddy Reaper, the wave files will look like this:
While tracking, be sure to turn down your monitors and go into headphones. The monitors will cause a delay (not always undesirable) and if turned up loud enough, feedback. The phones will isolate each track and reduce bleedover.
BRING IN THE HUMAN
My next pass on the song is a scratch guitar and vocal track I use for reference during subsequent tracking. I create two new tracks and record enable them with my two M-Box inputs. One input is a vocal microphone and the other a plugin guitar. I “rewind” the song and hit the red button. After the two bars of count-in scroll by, I play and sing the song from start to finish. You will know by the time you get to the end whether on not you timed your drum fills properly. If not, record the drum pattern again. If you are on the money, its time to start trackin’ and stackin’.
Next I create a new track and plug in a bass guitar or use one of the patches on the keyboard to get a bass line going. A couple passes and I have a bass track that sits nicely with the scratch rhythm section.
My next pass is again on two channels but this time I am using the M-Box’s input for the right and left channels of my Line Pod, a guitar processor with great amp and effects models. The unit has great stereo effects so I record in stereo. If I was recording a commercial release instead of a quickie demo, I wouldn’t record the guitar with effects unless I had at least one clean line in too – it locks you into a sound you may not want long term. On the quickie demo session, there will be sparse instrumentation so I print the effects on the pass.
I add tracks until my session looks something like this:
Track 1 – Drum machine right (click)
Track 2 – Drum machine left (click)
Track 3 – Scratch vocals
Track 4 – Scratch guitar
Track 5 – Electric rhythm guitar right
Track 6 – Electric rhythm guitar left
Track 7 – 2nd Guitar right
Track 8 – 2nd Guitar left
Track 9 – piano/keyboard right
Track 10 – piano/keyboard left
Return to Track 3 and record the keeper vocal and recut the scratch guitar part if necessary. You could add a couple tracks for harmony vocals and you’re set to start mixing.
When mixing a simple song demo, there are a few things to keep in mind that are different from a straight ahead CD release production. First the lead vocal – the portion of the song that does all the heavy lifting, should be more out front than on a CD mix. The lead vocal defines the song and you want the listener to be able to hear every word and note clearly. Also try to use the absolute minimum of effects and processing during a demo mix. If there is a “solo” or instrumental section, keep it brief – a prospective song buyer will have their own ideas in that department. Simple is good. Less is more.
Most audio software and hardware contains some “mastering” capabilities – most notably normalization to bring the performance volume to the highest degree possible without clipping. Use normalization and multi-band compression and EQ sparingly. Many platforms have “wizards” or presets that will be adequate for this job. If your tune has soft, quiet passages, be very careful with normalization or your subtle whisper will become a shout. When bouncing the mix down to stereo, leave yourself a bit of headroom should you want to revisit the song later for overall EQ and compression tweaking. I never let a song peak higher than -.5 db. Its not a lot of room left but you have a bit of space to make modifications without having to pull the entire performance’s volume down globally to make minor alterations.
FARM TO MARKET
Another thing to consider, especially if the song will be distributed as an MP3, is to brand the files themselves with the indexing data (Meta Data) that reveals the artist, song title and even photo of the writer or CD cover. Just about all CD duping software such as Nero will let you do this. It brands the file as yours and in the more recent CD players will display this information on the MP3/CD player as the song plays.
A simple song demo session takes me anywhere from one to three hours depending on the song and instrumentation. I am no Flaco Jimenez on the accordion so it might take me all day to come up with a part I can live with but I can slap down guitars as fast as I can record enable tracks. You will create your own formula for work flow as you gain experience and each new song will go faster and more smoothly than the previous session. The recording might not win a Grammy but hopefully will get the ears of someone who can take it there.
UP NEXT MONTH
Next month I will be sharing with you the results of my telecommuting experiment. How much does physical proximity to the industry centers count in today’s music business? Is your physical location really an issue in the days of high speed Internet? How hard is global collaboration? Firing up my old buddies Skype and LogMeIn, I will be finding the answers on my next trip to Jamaica.