"The 11 crude vignettes of variously woozy summertime jams are made with largely malfunctioning and repurposed instruments..." - Jennifer Lucy Allan in The WIRE magazine No.317, in a review of King Kong Ding Dong album Youth Culture Index.
"Dunebuggy" and the other Marvelry Skimmer track ("Beachball") posted up here on Workbench (and at FMA) were inspired by something I read in a music review in The WIRE magazine, that one above. That they can serve as inspiration for more music is the primary reason I dig music reviews; a good turn of phrase can be great fuel for the imagination. Not that I'm necessarily prompted to turn whatever spark I can get off a review into a concrete project, but in this case it did happen that way. I think what I had been working on previously had been wearing me down, and the review presented an idea that seemed like straight up fun, easy fun. Yeah right. It did turn out to be pretty fun -- especially "Beachball" -- but it was also a helluva lot more work than I thought it would be.
So, the assignment: make 11 "summer jams." I made two and started on a third (which turned out to be not so much a summer jam) and by the time I got that far the summer was over. Blame it on "Dunebuggy:" it took a long time to make. Or at least it sure seemed like a long time.
Here's the blow-by-blow.
August 10. "Beachball" is posted to the Free Music Archive. I had already started thinking about what the second track should sound like; I thought it would be acoustic, campfirey.
August 15. A Sunday. I had set aside some time to start on the second "jam." The first thing I did was listen to "Beachball" and try to hear what should come next. The acoustic idea I had had a few days ago was gone; I couldn't really recall it. So I listened to "Beachball" a few times, trying to casually sneak up on the ending; to be loose enough to hear something start in after it.
This is probably an unusual way to work; often you won't write songs that are "pre-sequenced" like these two are. You'd just be writing a bunch of songs and dealing with the sequence afterward. But in this case, I knew what I was working on would follow directly after "Beachball" and I wanted to make sure the transition was a good one.
After a while I thought of the keyboard sound, the note "B", the fifth degree in "Beachball." And then the jangly guitar from "Beachball" over it, and thought that whatever happened next, it would happen in the key of B.
I opened a new ProTools session (called "Jam 2") and imported the jangly guitars and bass tracks from the "Beachball" sessions. To fix the tempo, I recorded an electric guitar track - just percussive noises, while listening to the bass. That way I could get rid of the bass -- which I wasn't going to use -- but preserve the same tempo in my new song.
I placed the jangly guitars about 20 seconds into the start of the song -- which at this point consisted only of the "tempo" guitar track -- leaving that empty time to introduce the keyboard notes. I found two sounds on the Casio keyboard I bought the week prior for $20 off Craigslist, and record the "B"s. It wasn't going to be as simple as that though. The keyboard's cheap digital technology sounded harsh and unmusical, and there was no pitch wheel to soften up the sounds, or any other options either. My solution was to use a few digital delays on the notes, riding the time settings to wobble the pitch slightly.
That was about it for the day's recording, but I needed to figure out what the actual body of the jam was going to be.
I started recording into my cheap mono cassette deck; my usual method of coming up with ideas. At this point, I still thought the jam was going to be acoustic, so I pulled out the steel string, tuned it to an 'open B' chord, and luckily I happened on a riff that I liked. Two riffs actually, one more complex, one less, and practiced them to get them smooth enough to get the idea on tape. Later on in the day, when I was doing something entirely different, I realized that the part would be better on bass than guitar. Here's the recording I made of the riff, from the practice cassette:
Dunebuggy (acoustic sketch) by James Beaudreau
August 16. Around 6 a.m. A brief morning recording session before work: I printed the keyboard tracks with the delay effect so I could start the song on a pitch wobble.
August 16. 9:00 p.m. I picked up the bass and tried the riff I wrote the other day, and it worked. I opened up the ProTools session and listened: at this point al it was was the keyboard intro, the "jangly" guitars, and the scratch electric guitar keeping time. I tried to judge where the bass should come in, and, deciding on that, set up to record the part. Since I wanted to record this song without a click track, it would be up to the bass part to keep the rhythm strong. I put the bass in "direct" through distortion and EQ pedals and did my best to play it solid for around five minutes. It didn't come off though.
August 17. 6:50 a.m. The bass performance hadn't been good, but the idea was solid. And I could tell at this point that there would need to be additional sci-fi sounds to announce the entrance of the bass. Which made for fun recording. I also now knew that the bass wasn't going to work without a click track, and that the tempo needed to be quicker than the scratch guitar tempo.
I decided to open a new ProTools session ("Jam 2b") to record the rest of the track so that I could set up the tempo without reference to the tempo in "Jam 2;" I'd edit the two parts together later.
By 9:15 I had the new session set up, the tempo chosen, and I stared the bass part. By 9:37 the part was done, I had a mean blister on my index finger (since I only play bass lately when I'm recording), and it was time to go to work.
August 18. 10:00 p.m. I put the Casio through an EQ and distortion pedal and found a sound that reminded me -- with some imagination -- of a Fender Rhodes through a ring modulator. By 11:15 I had recorded the keyboard lick in the verses and processed it with a flanger plug-in. At this point the shape of the tune was coming into view.
August 19. 6:00 a.m. I had known that I was going to want to have some bridges to break up the verses in this tune, and this morning I figured out where they needed to be. There would be chord changes in the bridges, and I knew the bass would need to be different in those parts. So I tried a couple of fancy ideas, but before long I accepted that I already had a part in mind -- I had been ignoring it, thinking it was too easy or simple. I muted the bass where those bridges would be. I looked for some keyboard sounds, something halfway between "ballpark" and "haunted house". Finding them, I recorded two keyboard parts in the bridges. By this point it was 7 a.m. and time to get ready to leave for work.
August 20. 7:20 a.m. A quick session to fix some iffy moments in the bridge keyboard parts.
August 20. 7:00 p.m. Now I needed to figure out the bass part under the bridge. It came together fairly easily, though it took a bunch of takes before I could get it down well enough to keep. The keys in the bridge didn't sound bent enough, so I ran them through a wide chorus effect.
August 22. It had been a week by this point, and it was time to get to the drum tracks. I don't have a proper drum kit; partially because of NYC apartment living space restrictions, and partially because I never seem to have the money lying around to go pick up a bass drum and a hi-hat. So, for the time being, I've made do with what I've got: a snare drum, a floor tom (rescued from the curb by Philip Lynch), and a cymbal. Though I since haven't done it this way, on "Dunebuggy" I recorded each drum separately.
At 1:30 I started with the "tom-tom" part. One of the "toms" is a detuned snare drum without the snares engaged; the other is the floor tom tuned up tight. I knew the kind of beat I wanted: a latin beat, something like the pattern played by Elvin Jones on the Joe Henderson track "El Barrio" (from the Inner Urge album). So I listened to that track, trying to figure out what elements I could take out of it and in a limited way, replicate.
The toms didn't take too long to record, but because they're on the very edges of the beat -- and I'm not exactly an expert drummer -- I had to shift some hits around in ProTools later. At 2 p.m. I re-tuned the snare drum and recorded the rim-shots.beat I have to shift some hits around in ProTools later. At 2 p.m. I re-tune the snare and record the rim-shots. Here are the isolated tom and rimshot tracks, each recorded separately.
Dunebuggy (toms and rimshots) by James Beaudreau
August 26. 11 a.m. - 12 p.m. I was working from home. I had more drums to record, but it's not something I can do too early or too late, so I squeezed in the bass drum and ride cymbal parts at lunchtime. I took the floor tom, put it on its side, detuned it, stuck a pillow in it, and voila, there's the bass drum. After I recorded the part I detuned it further by a few steps in ProTools to differentiate it from the tom hits and to give it more bottom end.
The ride cymbal track went pretty well. I recorded a take, decided it was too busy, and recorded a second take with a straighter rhythm. It turned out that the first one was the keeper. Here's the full "drum kit" mix:
Dunebuggy (full drum kit) by James Beaudreau
August 29. Weekends: generally better for making some noise. At this point it was Sunday, and time for electric guitars. Lately I've been playing a Parker P-44 guitar through a Vox amp. At around 1:30 p.m. I found a tone I liked using this set-up plus an EQ pedal, put an SM57 on the amp and did a few takes, improvising a part.
After a little while it was feeling like a waste of time. I couldn't seem to come up with anything good, and I started to think that the electric guitar wasn't the right texture. Called it quits at 3:00 p.m.
August 29. 5 p.m. After doing something else for a couple hours and clearing my head, I returned to the track and found that some of the guitar parts were pretty good after all. One of the takes was better than the others. I prune it down, editing in ProTools, and it's good to go.
August 29. 6 p.m. The rhythm guitar success had improved my mood considerably, and it felt like a good time to give the planned guitar solo at the end a try. I used the same equipment, but changed the settings on the guitar and EQ pedal, recording five takes in quick succession. I usually enjoy recording guitar solos and this day was no exception. The five takes took ten minutes to do.
Next it was time for some sound effects to ratchet up the freakout around the guitar solo. I chose one guitar take at random (take 5) to play back while adding the best sci-fi sounds I could get out of the Casio. Later I made the final guitar solo out of an edit of Take 4 (the first half) and Take 5 (the second half).
August 30. 8 a.m. To boost the end section I added an additional drum part (played on the Casio). I figured I'd put it in the left side of the mix, with the main drum part leaning to the right. This was pretty easy: I recorded the part with my left hand once through with a fix to the opening figure.
August 30. 6 p.m. Tambourine track added. Then I deleted most of what I had recorded, only keeping one spot in the beginning, and the ending. It would go with the Casio drums on the left. Here's a short sample that isolates the drum kit, Casio drums, and tambourine:
Dunebuggy (drumkit and casio drums) by James Beaudreau
September 1. 9:45 p.m. I tried adding some electric guitar fills here and there on the track, mainly in the second verse. None of it really worked, though I liked one particular lick which subsequently made it through to the first mastered version of the track before being dropped before the second mastering. Here are the excised guitar riffs; all of them were discarded the same day they were recorded except for the first one which made it through to the first mastering session.
Dunebuggy (guitar outtakes) by James Beaudreau
The recording phase was finished at this point.
September 2. The intro section ("Jam 2") and the main section ("Jam 2a") were combined into a new ProTools session ("Jam 2b") and I started mixing.
September 4. The "final" mix was finished. It was much more compressed than what I normally produce, partly because I was determined that the first, second, and third (the ending section) parts should all have their own intensities. And partly because on my monitors it was sounding good squashed. But the heavy compression would prove to be a bad idea. Here's a snapshot of what the final mix looked like.
September 9. 4 p.m. Today Scott Hull mastered the "final" version of the track. I was at the studio (I work there) when he asked me to come into his room to take a listen because some elements weren't sitting right and he wanted to check that it was sounding like I had intended it to.
In fact, it's wasn't. And here, fellow home recordists, is why you should always get your music mastered, even if -- ESPECIALLY if -- you're recording in a less than "pro" setting. There were problems with the mix that were fully inaudible to me on my home monitoring systems. In Scott's room the tom part (the detuned, snareless snare) was super loud, with a crazy bloom on it that obscured most of what else was going on. Additionally, some of the relationships between the instruments sounded skewed, and worst of all, there wasn't much that could be done to fix any of it because of the heavy limiting I had already applied to the track. Scott made an heroic effort to carve out some of the aberrant tom-tom and to glue the rest of the mix together, but in the end too much EQ surgery had to be done and the track sounded, ultimately, not worth all the work I had put into it. I took careful notes while listening in Scott's room.
September 9. 9 p.m. It was time to remix from the ground up. I removed the limiting, adjusted balances between some of the instruments, tweaked some EQ settings, and reconsidered the panning. The guitar lick I had recorded on September 1 got discarded. There was one downside to the rebuild: in my haste to correct all the deficiencies I forgot to add the a big wooshing flange that I had at the very end of the track on the previous mix.
September 10. 8 a.m. I discussed the mix via email with Scott and he agreed that it was a good idea to try to re-do the track. Before uploading my new mix to the Masterdisk server, I listened to it one last time and made a few adjustments. This time, I'd have it seperated into two stems: the drums in one, and everything else in the other. Scott mastered the new mix later that day with beautiful results: the final released version.
September 13. I heard the track over the weekend, and it sounded good to me, but my confidence in my home monitoring had been shaken. I was glad when I debriefed with Scott this morning and he confirmed that the new mix/master sounds good.
And that wraps it up... almost. The nights of September 13-14, from around 10 p.m. to 1 a.m., and the early morning of September 15 were spent prepping the Workbench post, choosing the artwork, and writing these notes (later edited and updated in December).
The process of building the track had it's ups and downs, but mainly, looking back, it sure seems like it was a lot of work. At the end of it, after writing the first version of these notes, I couldn't imagine wanting to do it again. But, as I wrote at that time, "I probably will."
And I did: the next track I finished took longer!
PREVIOUS POST: Marvelry Skimmer "Beachball"
