Re-Laid Plans

Currently reworking a track I wrote in 2002 called "Best Laid Plans". It's a song I absolutely adore in its lush nakedness. Like "Solace", it was written completely on the fly with no attention paid to meter or metronome. At the time I recorded it, I was using a borrowed MIDI controller and a software synth, the Yamaha S-YXG50 (free with my sound card), to make sound, and some freeware sequencer to capture the MIDI data as I played. The problem with the S-YXG50, a problem which still exists to this day even on fast machines, is the unbelievably high 500ms latency between MIDI note-on events and actual sound output. This makes playing fast melodies practically impossible -- but it's OK for long chords.

One night, I hit Record on the sequencer and started playing a really nice chord progression with a patch called CCPad, and really hammering on the keys in frustration at the lag. I got some cyclical motion on the bassline, but no regular beat formed. When I played it back, it sounded kinda nice -- until I doubled up the sound by layering on another patch on the same MIDI channel called Dream Piano. When it came in, the thickness and hunger of the sound washed over me, and I wept in adoration. I knew I had something special.

Within a few days, I had the MIDI imported into Acid Music and made my best attempt to render the MIDI output into a wave. For some reason that's lost to the sands of time, I recorded both the pad and piano on the same take instead of independent tracks. I also used the crappy reverb and chorus effects built into the S-YXG50, even though they were noisy as hell, because they sounded awesome for the material and Acid's Music's effects busses were limited by design to one effect at a time.

The track needed something extra during the quiet lull in the middle, so I grabbed my external modem, installed some voicemail software, and used my phone to record a monologue ad lib. Some guy speaking personally, frankly, about being alone (too bad it's my voice). The sound coming through that phone and into the crackly compression of the voicemail gave it a palpable feel. The on-the-spot performance gave it a currency that practice would've lacked.

After some mixing, the song was complete enough. I did a quick but crappy render to MP3 to check the sound, and that's the version of the song I've been listening to for years. Well, the goal of writing music is for others to hear it, and I've been hoarding it to myself for a long time. Now I want others to hear it, but not in its original state. The sound is too awful, and the words are too raw, too personal, almost whiny. So last week I set about re-recording each patch into a separate, isolated, and dry (no effects) wave track.

The plan wasn't without hitches, of course. Nothing worth doing is easy. I destroyed the OS on almost two systems. One, my ancient XP desktop, where I originally produced the track, had some driver problems, and I screwed it up while trying to fix it. The other system, my late-model laptop with XP, I almost broke when the newer of the two versions I have of the S-YXG50 software exhibited a serious incompatibility with XP and caused explorer.exe to crash after startup (the synth isn't even a VST - it's Win95~Win98-era stuff). I managed to save that system by uninstalling that version and installing the older one. And then I had to wrangle with and edit the MIDI data that I had in archive, because I can't remember where the MIDI that went into the first recording actually ended up. And Acid would crash or inject bad data into the MIDI track. And then the MIDI patchbay software (MIDI-OX) would drop note events during recording of the wave tracks. And the Windows mixer on the laptop isn't lossless, but resamples the master fader, and so on.

Honestly, I can't believe I ever got anything done back then on any of this old and busted equipment.

I imported all the waves out of Acid and into Sonar (superior to Acid in every way). Added all the necessary effects to bring the dry tracks back up to the noisy, breathy, airy warmth of the original but without all the clipping and excessive static. For instance, the S-YXG50's wavetables had a high bit depth, but its effects were 8-bit to reduce processing overhead, which raised the noise floor and gave it that airy sound that Sonar's effects couldn't match; my cure for simulating that is to use a bit-depth dithering plugin in the chain to crunch it to 8 bits. And tonight I stood in my walk-in and recorded the rewritten spoken words with the crappiest consumer mic I could find to help simulate the original telephone sound. Vox needs more enhancement work, but the phrases are arranged in the right spots now.

I have a rough mix and so far it sounds awesome. I'll deal with mastering once I'm happy with the final arrangement. Can't wait to share it with you guys.