O'Reilly    
 Published on O'Reilly (http://oreilly.com/)
 See this if you're having trouble printing code examples


In many ways, music technology has become a silly numbers game. Today's digital instruments are sold on gigabytes of samples, kilohertz of frequency response, and ever-rising bit depth, polyphony, and even more esoteric specs. In this episode, George "the Fat Man" Sanger goes back to basics, gleefully playing everything from an ancient test-tone generator to a spicy Excaliburrito to show that one glorious note may be all you need. (DMI 07-21-2008: 15 minutes 10 seconds)


Production Notes: Recording

I reached the Fat Man at his Texas studio, Abbey Trails, by SkypeOut, recording our conversation on my Mac with Ecamm Call Recorder. Because that meant his side of the call was still going through the cruddy telephone mic and harsh Skype compression (my voice went directly to my computer through my Logitech 250 USB headset), I had Fat record into his computer as well with a second, high-quality mic. After we finished, he sent me his side of the conversation as an MP3 so I could sync it up with mine, using his Skype track as a reference.


The Fat Man also recorded two additional tracks: stereo Moog and mono Excaliburrito synths, and I loaded those into my multitrack editor, Ableton Live, as well.

Minimoog Voyager
The Minimoog Voyager slots perfectly into Sanger's studio. (Click to enlarge.)

To synchronize the local and remote voice tracks, I panned the Skype reference track 100% left and the direct vocal 100% right. Listening on headphones, that made it easy to tell when the two tracks were in sync and which was slightly ahead. (The brain can detect ear-to-ear differences of a fraction of a second; it's one way we deduce the direction of sounds.)

Theoretically, I should have been able to drag the remote vocal left or right on the timeline until it lined up with the reference vocal. But as I've discovered in previous two-ender recordings, the two parts often run at a slightly different speeds, requiring some warping (time-stretching) in Live to sync over the long run.

This time, however, I couldn't find a warp setting that worked. No matter what I did, the Fat Man' direct vocal would be behind the Skype reference at the beginning and ahead of it at the end. Finally, I realized he’d stopped the recording briefly in the middle, creating a “ripple in time.” Scanning through the waveforms visually, I found the place where the direct signal was missing a few humps. I then split the clip into two parts and slid the second part later in time so its beginning lined up again. With the missing chunk restored, I found that slowing the direct signal 0.02 bpm put the two recordings back into usable sync. (See Figure 1.)

Fig. 1: Waveform sync
Fig. 1: About 40 minutes into our conversation, the Fat Man accidentally paused his recorder, making his side of the conversation fall out of sync with mine. Splitting his track and sliding the second part to the right restarted sync (top oval). Altering the tempo slightly (bottom oval) compensated for the remaining drift.

Arrangement and Editing

After marking the sections of the conversation I wanted, I exported them as stereo WAV files (my voice on the left, Fat's on the right) for precision editing in BIAS Peak. I snipped out "ums" and used BIAS SoundSoap to remove some of the fan noise my mic picked up from my Mac.

EIKO Oscillator
This ancient EIKO oscillator was a gift from the Fat Man's father, who wired it up in the days before digital.

As usual, I then used Apple QuickTime Pro to record my voiceover. (The CPU hit is so low it doesn't trigger the Mac's fans.) For a mic, I again used an SE Electronics USB2200a condenser. After snipping out P-pops, tongue clacks, and false starts in the voiceover with Peak, I imported everything into a new Ableton Live session. I slid the elements around on the timeline to get the best flow, enhanced the vocals with Izotope Ozone, adjusted levels with envelopes, and rendered the mix into a stereo AIFF file. Finally, I converted the mix to an MP3 in iTunes, where I added the cover art.

The Digital Media Insider theme music came together in Live as well. The opening sound effect is a compressed mouth noise spliced onto a tone cluster I generated in Native Instruments Reaktor. The main groove is from Steinberg Xphraze. (Jim Aikin turned me on to both virtual instruments in his article "My Five Favorite Soft Synths.") The piano is from the Garritan Personal Orchestra, which I discovered when we interviewed Gary Garritan.

The theme also features a few percussion samples dredged from my hard drive. Altogether, it took just six tracks. Effects processing was courtesy of Live's default plugins and Freeverb.

Links

You can find all 26 episodes of Digital Media Insider on the show's home page, http://digitalmedia.oreilly.com/podcast-audio.

Copyright © 2009 O'Reilly Media, Inc.