› About the vocals

‹ About the vocals

Vocal samples come from two sources (both OGL):

1. Liz Truss' resignation speech (20 Oct 2022)

2. Boris Johnson's resignation speech (7 July 2022)

3. (Update, 26 Oct 2022: Is anyone else intrigued that Liz's speech already has almost twice as many views as Boris's?)

I'd previously chopped Mr Johnson's speech during week 27 (Resigning Party).  The original was still on my SD, and―by coincidence―I'd revisited it on Wednesday night and re-chopped him to rap/beatbox (the initial "a-backbench'n-na-na-aback-aback").  12 hours later, the Prime Minister of the United Kingdom announced her resignation. (Coincidence?  You decide).  It was clear what must be done.

It was much harder to convince Liz T to rap.  She's comparatively softspoken and her speech was deplorably focused and short.  Time was tight, so at first she just  beatboxed hi-hats with hard "T" consonants.  But this is her song, so she got additional coaxing to solo some funkier scat.  She keeps just about level enough of a tone now and then to harmonize her with herself, which was fun but a bit tricky because I didn't use any time-stretching hacks to keep them in sync at different pitches.

Background noise was a major complication during her performance.  The air behind Boris's speech had all sorts of noise (yelling, chanting, and at a few points hints of what might be a brass band in the distance―I still wonder about that).  But the sound engineer must've been really on point―or Boris is just so loud―that the din was mostly inaudible as he spoke.

Not for Liz T: camera shutters on burst mode are audibly clicking at ALL TIMES as she speaks.  It's impossible to filter them out from her voice (at least, not directly on the M8) and they have a machine gun rhythm of their own that can clash at any speed.  This effect becomes particularly vexing while trying to harmonize two copies of Liz T at different speeds―it sounds like she's rapping inside a typing pool.

› About the drums

‹ About the drums

The main beat is mostly chopped from track 8 of Barrys Ultimate Fakes and Beats Vol 1 (Sample Pack), by "Barry Beats" a.k.a. Si Spex. 

Barry/Si's idea of fakes" is brilliant: he finger-drums authentic-sounding "vintage" drum tracks using virtual drum kits and processes them to sound like they could have been ripped from vintage records―but in fact they are completely original and royalty-free.  It doesn't hurt that he's master of emulating the sound and groove of almost any genre, with a particular flair for the "sound library" sound of the 60s/70s.

Once WB is over, I want to try making a few fakes of my own.  They open up some fun possibilities, and it might make my Komplete license from a few years back seem like less of a waste.

› About chopping samples on the Dirtywave M8

‹ About chopping samples on the Dirtywave M8

For breaks and vocals, I generally try to:

1. Keep sampler instruments in FWD mode

2. Jump to different locations using STA

3. Use DEL/KIL to adjust timing and articulations

This lets me chop non-destructively by jumping around bits of the same file (and saves time/space while bundling).  But there's a drawback: STA can only specify 256 points (00–FF) and their increment size is proportional to the length of the sample.  This forced me to go back and crop longer clips into small sections a few times in order to get enough precision.

That strategy backfired when I discovered―quite late―that the main break's kick was a bit late on the first beat of each row.  It couldn't be fixed non-destructively, because the actual hit fell in between STA 7F and 80 in the sample.  When the mix was busy, the mismatch just read as syncopation.  But during sparse moments it felt jarringly out of sync with the sub.  By that point it was too much trouble to fix (part because that same break had been STA-jumped into itself at many points―fixing the first beat would lead to a cascading need to readjust the relative timing of the chops after that, leading to a similar precision problem at each point down the line).  The quick answer: take out the sub during the sparse parts, leaving just the break―its timing sounded fine on its own, and the relative timings of the later chops were left intact.  And it turned out to the arrangement sound less static.  Lesson learned?