On October 27th of 2019 in the pre-pandemic Terra, just a few days after I got the Volca Modular and a week or so back from a show in Seattle, a little sequence sprang up while I was messing about, and I knew I had a song. Rather than sit at my kitchen table and record a quick take, I had this desire to just get out of the house and do it remotely... 
Why not go to an abandoned building, set up, and record the song straight to the camera there? I packed up my GH4 and the gear I might need. This included headphones, a passive mixer, the Kaosssilator, and a lapel mic. There would be no computer. I had decided to try something "experiential", and I headed out to a place I knew would work to get the creative flow going. The area I was interested in was past Adelanto, CA where there is an abandoned army town that I like to call, Neverville.
George Air Force Base was a United States Air Force base located within the city limits, 8 miles northwest, of central Victorville, California, about 75 miles northeast of Lost Angeles. I have gone there many a time to shoot as it is the closest thing you will see to a post-nuclear dystopia, devoid of almost all life. It is a place that has grown somewhat darker every time I visit. During that October stint, I felt it was darker than usual (premonition perhaps). For one thing, there were signs of gang activity, strange warnings, and many other things that one would consider it best not to go alone as I had. But there I was. And I felt alone. 
For the recording session, I would play the sequence on the Volca live, while simultaneously signing into the lapel mic, all feeding the mixer and being recorded straight to the camera. The mic clip quickly broke when I got there. Because there was no way to get the mic to sit right without a clip, I used some tape (I had some in the car) to literally tape it to my chin. This allowed me to focus on the Vocla movements and the Kaossilator to my right. The experiment proved to be quite harrowing, and while I got some incredible footage of the town itself, the building I was to try and record in yielded nothing good. The being in my head did not make me feel comfortable, and any sound that got to my headphones felt threatening to me. I scrapped the idea after a few extra hours of filming random things and went home. The filming did yield good stuff, like finding an entire room filled with tumbleweeds that seemed to be alive and watching. And there was a tree going through a window...
It wasn't until I got home that I decided to try recording the song again in my own bathroom, which sports a very old heater. The feeling was better since I was not constantly looking over my shoulder as I had been in the abandoned building I'd set up shop in (similar look though). But to keep things consistent with the other recordings I'd made, I once again taped that dollar microphone to my face and did a few more takes of the song. One of those was OK and I edited together a video that I posted to YouTube. Don't bother looking for it as I have scrubbed it.
Besides the obvious feeling of being in a bubble in a dangerous place, I was not used to using headphones (especially the cheap ones I had on hand), the mic being taped to my face was terribly uncomfortable, and the entire thing was off-key because the lyrics were new to me. I was horrified, but I let it go as I had gigs, there were art shows happening, and there would be time for re-recording the thing at a later date.
I made a few videos explaining that I would be re-sampling the Volca and doing my best to record the song properly when I had the chance. Meanwhile, without much fuss, I managed to do a few overdubs and re-uploaded the video with extra sounds. I did not have a separate vocal track, so that bummed me out. I did a sampling session but the truth is that I was never quite able to recapture the sound that I got in the field, which eventually turned me off to hardware altogether. Here's a video of the sampling session I did a few weeks later, which is still on Facebook:
https://www.facebook.com/endpop/videos/1056720098006655
By now I had done a talk for an art group about my liquid rubber technique, and I was doing my best to deal with something that was going on with my mouth/speech, resulting in having to consult about possible surgery. And of course... looming in the corner just as I got back from an El Paso / Chicago weekend was...
The Pandemic
The pandemic took the entire world and turned it upside down in every way imaginable. While I can say that the pre-pandemic Claude was perhaps dealing with the usual creative bouts of ebb and flow versus nothingness and despair, the pandemic created an aversion to being outside for that first month. I remember it well. It wasn't until I wisened up that I came to my own notions about what was really happening, but by then I was doing the Quaranstreams, doing my best to stay in my apartment, dealing with the surgery bit, and re-discovered my love of all things crypto in the process.
I literally forgot about this song. Besides, it is nothing. It is just a song. What does it mean? But something made me want to revisit it and see what could be done with it in the end. After all, the footage itself is nice. And I have been wanting to look into doing more with my footage of that place.
So I have started to mess with it today. Thus far, I discovered the footage is quite good, and aside from the terrible vocal and overly loud noise which could perhaps use a bit of automation there is quite a bit of information in the audio to make it all work and still keep the video. So that's my project for this week among other things that happened over the weekend. I am going to use some AI tools to see how much I can separate the tracks, hopefully, re-do the vocals and mix the thing back into the video. If that's all good to go, we should have it by Friday.
Monday Update:
Morning Update
I'm missing two files, one is called White.mp4 and the other is of course Black.mp4. If I am unable to locate them, I will have to splice the video from the video master reference. I found the file, but I do believe I made a clip of it inverted. In any case, I have found that bit. As for the Reaktor intro, I have it on audio, but the video exists only on FB. That's a drag, but I will deal with that later.
The next bit to find was the Ableton project, and in going through the footage, I found it although it was called Time Bound back in 2019. That's such an odd title. No wonder I had trouble locating it.
Evening:
I have the mix from the camera isolated and have to render it out making sure there are no effects on the main bus. This file that I render will be the one that I can correct with Wavelab and the AI to separate the main things. Mostly whatever I can get so long as I have the vocals and the main tracks. Drums would be nice too, as there is a flub I can fix, but it's not a deal breaker. Even if I don't fix it, there is a great effect I have in mind to help in this area.
Strings and bass synths were isolated as overdubs, and they are not perfect, but they have a charm that goes with the video performance, so I am keeping them in. I may nudge a note or two.
Vocal is an oddity to me. I will have to hear it raw, but has moments where it is interesting as a performance, so maybe a fix? We will see how the isolated track sounds first. Otherwise, the mic I will use is the Shure SM63 to track it over again. Besides, it will be fun to sing it if I ever plan to play it. Best to get some practice time. Tomorrow is the artilect's turn to help me do the mundane work of making some stems. I'd do it now, but the body is complaining. I've been up since 7:02 this morning, so I'd say I put a good day in.
Last-ditch before bed
Well, it worked. I was going to sleep on it, but I was playing around with different algorithms that analyze audio, and I successfully separated the audio raw audio I had in the camera into four very usable stems. The vocal can be cleaned, but I am still opting for re-singing the track. But at least I have a reference vocal I can clean up. The 'drums' were my favorite stem. It misinterpreted the little clicks as kick-like and leaves quite a bit of room for some play. I was able to even fix the annoying mistiming at the end that was bugging me! The 'modular' stem is the Volca, but also the noise stuff from my yellow Kaosillator. I got lucky in that if I want, I can replay the noise parts cleanly in place. I might do it depending on how far I get with the rest. It's really about the vocal now. If I cannot make the song resolve, it is an exercise in technology flexing and nothing more.
Phew. Rough day at the studio, but there it is. Ready to be worked on in a meaningful way before I add it back to the video. The Artilect did a good job. Now it is up to me to finish it.