Tuesday, March 4, 2008

Coalescence and Luxation

This is what I've been working on for the past week after I put my kids to bed. The music is realized entirely using my modular - every total element, one monophonic line at a time. The only non-modular element is Battery which I used for the drums. All of this audio was recorded into my DAW and severely processed and edited, but no other synths were used. This is the final version:

AAC (160k 7.2MB)
MP3 (128k 9MB)





Creative Commons License


This work is licensed under a
Creative Commons Attribution-Share Alike 3.0 United States License.

If you're not a synth nerd, stop reading now. I'm going to go into mind-numbing detail about each step of the process and provide audio examples of how the piece developed.

Of the modular patches, the four Cwejman modules I own (VCO-RM2, ADSR-VC2, MMF-1 and VCA-2P) served as the core of each patch, and account for about 75% of what you hear in the final product. I really love these modules. I don't believe they're to everyone's taste, but I think they're fabulous.

I was hoping to get to Where Are You? (part I), but the module I was counting on to realize that piece didn't... ah... function properly. So, I have a Flame Clockwork which arrived yesterday as an temporary solution. Hopefully I'll be able to share some of my impressions of this product at that time. Aching to work on music, Coalescence and Luxation started developing.

This piece started inside my head Saturday night as I was putting my daughter to sleep. It begins with a basic patch consisting of two triggers: one every two measures which fires off a long ADSR controlling the cutoff of a high pass filter, and another series of triggers going to a shorter envelope controlling the VCA. This is all Cwejman.

Then I added another patch where the output of a triangle LFO is sampled by a sample and hold, and this is controlling the plan b model 11 filter. The vactrols naturally smooth out the output of the S&H. I really liked how this pattern interacted with the original chord progression.



I left the patch intact. There were several things I wanted to change. 1) put the VCA after the filter to obtain a better signal to noise ratio. 2) record at 24-bit and 3) change the key and clean up the chord progression.

Sunday was interesting. Once the patch was altered and changes made, I could arm the next track then tend to family matters while the system dutifully recorded that pass and stop automatically. So, by the end of Sunday, I had something that sounded like this.



The fidelity had improved, but some of the character was lost, but it would make working on the rest f the piece a lot easier.

Monday evening I started processing the snot out of what I recorded with the Lexicon PCM80, doing all sorts of OCD things like stop recording after every chord change, wait for the delay lines to empty, and record again at the chord onset. Then, I started editing like crazy, creating the glitchy first half of the piece.

Tuesday, I used Battery as a sound source and recorded a drum track. Recorded the track in one take. Quantized, but one take. Rendered each element onto its own track while reading The Amber Spyglass. The drum tracks are the only non-modular-generated element. Vocoded the drum tracks with the existing tracks. Again, did the OCD-delay recording thing. I wish I had a software delay with an automatable delay kill. Spektral Delay did this, I'm not sure if it was automatable, but that wasn't a 'normal' delay, either.



Wednesday night I wanted to add some lines that took notes from the chord progression and trigger accelerating or decelerating patterns. For each line, I selected a different flavor of filter, cwejman MMF-1, plan B LPG, plan B model 11, Livewire Frequensteiner. I used the Cjweman VCO-2RM as an LFO that triggers the short envelope on the ADSR-VC2 since I could voltage control both the frequency and the pulse width - using the long envelope to create longer gates times as the sound progresses.

The composition is taking form, but by the time the original motifs come in around the halfway point, it is sounding too produced. In general, it sounds too rich, with too many layers. I want to cut down the complexity, focus attention on a sound at a time and play a little bit with the perception of meter and rhythm via some extreme editing.



On Thursday evening, listening to the track, the match=true flag goes off in my brain and I'm compelled to drop in The Carpenders "We've Only Just Begun" at a specific spot. It worked and sounded really creepy. I spend the rest of the evening refining this idea for no particular reason. Interesting and weird as this was, I wanted to release this piece as creative commons and didn't want to have any legal issues, so I trashed all this. Ah, well.

Friday night I came to the conclusion that subtractive wasn't the way to go on the first section and decided to re-render the idea using FM. That was ok, but the Doepfer A-137 Wave Multiplier was exactly what I was looking for. I ended up using it on all six tracks. Got everything ready to record and went to bed. Saturday I rendered the results. I spent the rest of the weekend editing audio and playing various processed versions of mix back through the Cwejman VCAs, using a MIDI keyboard as the trigger.

Monday night I created a proper bass patch and passed a mix of the first section into my favorite plate program on the PCM80 and recorded the result. Then I added another pass of drums and processed it though the Malgorithm. After that, some final editing, clean up and mix. At this point I have 7GB of audio data related to this single project.

5 comments:

bughouse said...

very cool - great sounds/comp/production and i really appreciate the "mind-numbing detail". thanks!

intersect said...

Very cool sounds, I love the way you fit the beats into the mix. I know from experience it can be hard to sync everything up. I want to hear more!

Richard Lainhart said...

Very beautifully done - thanks!

What are you using to generate the visuals?

stretta said...

Adobe After Effects CS3 and the Particular particle-generator plug-in.

I exported the individual audio parts into After Effects and extracted key frames from the audio amplitude. Then I wrote expressions in After Effects that used this data to animate the particle generators for each part.

The finished video took 20 hours to render on an 8-core machine.

The blog post about the video is here:

http://stretta.blogspot.com/2008/03/coalescence-and-luxation-video.html

copepod said...

as I listen more, I have to say that I like everything except the beats - which I feel imply a more rigid form than I was appreciating before that happens. It's still a really nice track though - and a lovely video.