Monday, September 16th, 2019

When You’re Tired of Recording and Re-recording . . .

By Craig CliffordNovember 5, 2012


Craig Clifford
Craig Clifford of Stephenville, Texas, is a prolific songwriter who performs regularly with the Accidental Band and occasionally as a solo singer-songwriter.  He has recorded three albums with the Accidental Band and one solo acoustic album.  Currently the band is working on its fourth CD, “We Ain’t No Lap Dancin’ Band.”  Born in Lafayette, Louisiana, Clifford grew up outside of Houston, then honed his musical skills playing in the coffeehouses in Austin, Texas, in the late sixties and early seventies.  His song writing reflects both his Louisiana heritage and Texas upbringing, as well as his musical roots in traditional folk music, early Bob Dylan, and the Texas singer-songwriter tradition.  Clifford is Professor of Philosophy and Director of the Honors Programs at Tarleton State University in Stephenville, Texas.

I tend to be an old-fashioned purist when it comes to producing recordings in which real musicians have actually played all of the notes.  OK, I move a few notes around if they’re really off-beat.  Occasionally I use pitch correction.  Occasionally I used Drumagog to replace the recorded kick drum beats if all of the EQ, compression, gating, etc., can’t make them sound as good as they do live.

But sometimes you just get tired of recording and re-recording, and you just want to sit at your computer in complete solitude and make it work.  In recording my band’s fourth CD, I recently got to that point, but one of the key songs of the album wasn’t working.

A little background.  The band consists of three musicians–in our live performances that means me on lead vocals, acoustic guitar or banjo, and harmonica;   Jim on drums; and Nancy on bass.  On our first two CDs, we recorded full-band scratch tracks.  On the first CD, we kept the drums and bass, and I redid my guitar, vocals, and harmonica.  On the second CD, we kept the drums only, and redid everything else.  On the third and fourth CDs, we departed from that approach because we upgraded from 12-channel so-so converters to a LynxTwo card with six channels.  Yes, I’d like to buy the 16-channel Lynx Aurora, and I will eventually, but I’d also like to stay married.  Enough said.  To record the full band we need one mic for my vocals, five to seven mics for the drums, two mics and the pickup for my acoustic guitar, and one direct channel for the bass.  So with the new converters we decided to record my guitar and vocals first, then bass and drums last. Yes, I know that’s contrary to the normal procedure of laying down the rhythm section, but my guitar timing is usually like clockwork, and Jim, the drummer, often likes to circle around the regular beat that the guitar and bass lay down. Overall, this has worked really well on this album.  But that’s another article.

On one song, which consists of a series of verses with a one-line refrain at the end of each verse, the timing on the refrain was all out of whack.  I started moving drum beats and bass notes around to match them to my guitar.  I couldn’t make it work.  I muted the drums and bass and tried to play drum beats along with my guitar and vocals.  I couldn’t do it.  Mr.Clockwork, it seems was seriously speeding up during the refrain.  And not just speeding up but playing a kind of syncopated beat that would not work with the regular drum and bass beat.

Option 1: Re-record the whole song.  Not an option.  After about six months of recording, I was in no mood to re-do an entire song.

Option 2: Do something radical using the capabilities of my software without compromising my belief that recorded music should be recordings of real musicians playing real music.

Here’s the way I worked Option 2.  The song is called “Deep Well,” and it was inspired by the BP well disaster.  The bass and drum parts are incredibly regular, drums playing half notes (kick beat, snare beat, kick beat, snare beat, and so on) and bass playing quarter notes, with some slight variation in the one-line refrain at the end of each verse.  I’m finger picking the guitar and playing lots of notes.

Figure 1 : Snare drum track from original recording marked with Audio Transient feature.

I used Sonar X1 Producer’s Audiosnap feature to find out the average tempo.  I set up the metronome  at that tempo.  I learned to play the refrain on-beat, and then I re-did the guitar playing along with the metronome.  Then I redid the vocals.  Faced with re-recording the bass and drums, I tried an experiment.  I used the Audio Transient feature to mark the kick drum beats in the kick drum track, and set it to be sensitive enough to mark the snare and hi-hat bleed in between (see figure 1).

Figure 2 : Snare drum track from original recording split into separate clips for each strike.

I then used “Audio Snap Split Beats into Clips” to split the entire snare drum track into one clip for each beat of the snare and one clip for the bleed in between (see Figure 2).  I then repeated this process for the kick drum and bass guitar.  I decided to jettison the overhead drum mic tracks and settled for the room mic to take care of the hi-hat as well as provide the room sound for the whole kit, so I repeated the process for that track. Luckily there are only two cymbal crashes in the song, and no ride cymbal, or this wouldn’t have worked.

I then proceeded to cut and past each snare drum, kick drum, and bass guitar note into my new recording (see Figure 3).  For the snare and kick drum this meant that I took just the main beats for each track and left the clips with the bleed in between the main beats.  Then I pasted in the room mic clips—on that track I occasionally had to move a hi-hat strike a little to get everything on beat.  This process was time consuming but relatively easy since I could line the old notes up with the vertical grid tempo lines in the new recording.  Finally, I pasted in the two cymbal crashes. 

Figure 3 : Bass and drum clips from original recording pasted into new recording lined up with tempo grid lines.

So all of the kick drum, snare drum, and bass notes were the notes that Jim and Nancy had actually played and in the order they played them in, but they were lined up with the new recording of my guitar and vocals.  I had to cross-fade between notes to get rid of clicks.  But it worked really well.  The one-line refrain at the end of each verse was a little tricky.  I used Audio Transients to mark the beginning of the guitar beats, and then lined up the drums and bass accordingly.  And since the overall feel we are trying to get with the drums and bass is a pounding regularity, with the guitar providing the musical ebbing and flowing, having the drums and bass pretty much right on the beat gave us the sound we wanted in the first place.

Jim added a keyboard part, and I added slide electric guitar in the refrain, but the primary instruments that the three of us play feature the notes we actually recorded, and the notes we play when we perform the song live, only organized a bit differently.

Would I do this again?  Probably not, and it probably wouldn’t work on songs with a more complex rhythm section.  Still, it was an interesting experiment, and it worked for the mood I was in at the time.  And, by the way, going through this process  improved our live performance of the song.  Take a listen and see what you think.

Craig Clifford
November 2012

Share on Facebook
Share on LinkedIn
Tweet about this on Twitter
Email this to someone
share with your friends

© audio times 2018. All rights reserved.

Website by Small Business Marketing