I released a musical jam every day in January

Jamuary is a project started in 2016 by True Cuckoo, a musician known for his positively quirky synthesizer music. The challenge is to record and share a jam every day of the month of January. In 2023 I managed to complete it, with 31 jams up on my YouTube channel. Here’s an overly detailed blog post summing up the experience!

I first tried out Jamuary last year when I published six videos over the course of that month. You’d think that experience would make me prepared for doing the full 31 days this year. In practice, I was pretty naive about the amount of work it ended up requiring. The rate of publishing that Jamuary requires is frankly crazy, but I felt that the experience gained was worth it as long as I set some quality bar for myself that I didn’t want to fall under.

How I approached Jamuary in 2023

I intuitively knew what I wanted to do from Day 1. I ended up summing up my ideas for my personal journey publicly early on in the month. To summarize here:

  • all entries are YouTube videos
  • posts are between 3:00 and 5:00 in duration
  • there must be a live performance component
  • use unique equipment setups between days
  • use 3/4 or 6/8 meters on days divisible by 3
  • use a variety of filming techniques
  • switch music genres
  • generative (or “AI") elements are allowed but cannot dominate
  • …unless I implemented the generator myself
  • put the effort in
  • but don’t sweat the composition
  • read the daily prompts on jamuary.org but don’t worry about following them
  • post on the day of but the day ends when I go to sleep, don’t fret that “midnight passed”
  • post on all 31 days of January

In the end, I broke some of the rules, for instance, there’s one piece below 2 minutes in duration and another one clocking at 7:30. But most of them really clock between 3 and 5 minutes, which is already longish for the modern viewer’s attention span. Another rule I broke was that I couldn’t really use unique setups every single day as there’s simply not enough gear. The point of that rule in particular was to push myself to dust off the less used stuff. To try out ideas I had in the past but never gotten to. That I succeeded in.

Most importantly, I stuck to the most important rules: I managed to post on all 31 days, and every post was a YouTube video with a significant live performance component. I avoided phrases that rationalize lazy publishing, like “I explore” or “I experiment” or “I barely scratch the surface” or something is “my first attempt at”. I put in the effort.

How I recorded a new entry

There were multiple phases each day.

The first phase is almost the most painful: figuring out which gear setup I even want to touch. I was nearly paralyzed by choice at this point, but also wanted my output to be diverse so that if anybody actually goes and listens to the entire playlist (doubtful but possible!), they won’t be bored by consecutive tracks being essentially rehashes of the same thing.

So I tracked my hardware usage in a notebook and deliberately tried to pick up my less-used gear every now and then. Otherwise, I would be tempted to just do “31 Days of Iridium”, and that would be terribly boring. It still ended up to be the most used piece of equipment in my videos, but it’s not in 31 videos but “just” 9.

I even tracked how many pieces included drums and how many didn’t, to try and get those evenly split. In the end, 15 entries included drums, while 16 did not. As close to 50:50 as I could get. Deliberately thinking about this sounds silly, but if I hadn’t done that, it would have been too easy for me to just do solo piano stuff. But that’s less interesting for people to listen to and would provide me less growth.

Then phase two was setting up the hardware, and figuring out what it says through me on that particular day. Once I knew what it said, it was time for phase three = recording the live take. That sometimes took hours to get right. You see, while most of my videos have cuts, and I could have easily made several takes with short time spans, I treated all of Jamuary as an exercise in music, as practice, and thus I only accepted complete takes.

Once this was done, phase four was a round of overdubs, usually just some pad, modulation, and delay… but sometimes also drums and an entire new melody line. Sometimes the overdub ended up being so integral to the piece that I went back to phase three to record a live take of the other instrument, too.

When all tracks were laid out and mastered, the final stage was to move video footage to Davinci Resolve and combine everything. This on its own easily took an hour of video editing, creating the YouTube video cover art, uploading, and copywriting the description.

In the end, every single day I spent 3-5 hours on Jamuary. It got old pretty quick but I stuck to it out of stubbornness and the desire to complete the challenge at least once.

New things I learned

There are a lot of new things I tried and learned through work on the jams. Here’s a laundry list, roughly grouped, and admittedly mostly useful for my own future reference. However, just looking at how long it is, it’s pretty clear this project clearly was worth doing.

New technical skills

  • Norns: Producing actual music with Arcologies. Installing samples for Arcologies and using them. Programming Arc support in Lua for Euclidigons.
  • Roland E-4: Using E-4 scatter on vocals. Using E-4 scatter on line-level input with TRRS. Sending audio to E-4 via USB from the computer. Harmonizing E-4 with a live controller.
  • Ableton Live: Dialing in an Aphex Twin slightly distorted 808 sound (key insight: hi-hats + toms together are this sound). Using IAC MIDI and BlackHole to use AUv3 applications in Ableton Live (enable Drift Correction on all secondary devices when using Aggregate Devices!). Putting Spectravox on Ableton’s reverb to get a choral sound. Realizing that reverb on grand piano with the dry signal removed has choral qualities to it.
  • FM synthesis: Patching ID700 from scratch. Creating an FM piano patch from scratch on Opsix Native and on Iridium.
  • Video editing: composition with filters (Organelles #8) and backgrounds (Mono Stations #11). Making LUTs in Davinci Resolve. Changing the FPS of an already made video in Davinci Resolve (don’t render in 3:2 as it creates interlacing; instead create a new Timeline with different FPS and copy contents over). Setting up my tripod in “low mode” to capture the CP300 keyboard exactly from above.
  • Other notable experiences: Using standalone MIDI and polyphonic aftertouch on Launchpad Pro (I haven’t realized or have forgotten it had it 🤦🏻‍♂️). Playing algorithmic music live on piano (Palindromes #16). Using live granular on Iridium. I also learned the basics of Melodyne and that my voice apparently can’t keep one pitch very well, even for a second.

Epiphanies

Apart from technical experience, over the course of the month, I realized a bunch of things about my music-making.

  • Some of my jams started out with a generative part as means of speeding up the process. In the end, it never sped up production, but it also never ended up overwhelming the piece. I still employed a lot of live performance each time. Conclusion: starting with something generative is a good means to unblock creativity and set up constraints for a piece.
  • I set out with an explicit goal to produce with varied software and hardware setups. I could make do with most things I got, but I noticed I gravitate towards some choices more than others. In particular, I surprised myself with how much I like to perform with Euclidigons on Norns.
  • On the other hand, while I used the Moogs a few times, I only used my modular skiff once. I find the time needed to set the patch up pretty cumbersome, doubly so that then I have to clean it all up afterwards.
  • I also used E-4 quite a few times because like what it did for me: it got me singing, something I wanted to do but was afraid of for literally years. Having the voice processor with auto-tuning, vocoding, and harmonizing allows me to hide behind those things while my vocals are imperfect. But the way to get the vocals to be better is to actually sing. The E-4 gave me the confidence to do it.
  • On the other hand, the E-4 is a rather frustrating device to use. Roland stripped it to the bare minimum needed to be musically useful, and while they kept all the necessary ingredients for the full package to be musically useful, there are a few severe limitations and bugs. I won’t be listing them here but I stumbled upon something every time I used it.
  • I thoroughly enjoy the TR-606, TR-707, TR-808, and TR-808 core kits in Ableton Live. They allow enough variety to be musically useful without wearing the listener out but are simple enough to use that a drum track can be added without much fuss to any track. Some of the macros provided allow dialing in cult classic sounds.
  • Just through Jamuary 2023 I released as much music as I did throughout the entire previous year. There are quite a few pieces I made in 2022 that I’m happy about, and there’s definitely value in giving productions more time than a few hours. Surprisingly though, I find many of my 2023 jams to be just as interesting. The challenge as a forcing function ended up achieving quality through quantity. I published over 3 hours of music during Jamuary 2023. I’m sure there’s at least 30 minutes there that is close to album release ready.

Things I wanted to do but didn’t end up having time for

At the beginning of the month, I expected myself to have time to generate visualizations for the music. This was naive. With the time spent just doing audio + video, there was no way to spend even more coding generative graphics.

I did end up coding a few things to help me with the jams, and I had plans for more. In particular, Subharmonicon accepts MIDI but does not natively allow providing chords to the six oscillators and sub-oscillators. There are CCs you can pass to tune the subharmonics. I want to make a MIDI processor where you play a chord, and it translates it to the required CCs so that Subharmonicon plays your chord.

Having this script would allow me to use Subharmonicon as a vocoder carrier for Spectravox, a combo I always thought would be perfect but which is hard to use out of the box without a lot of additional patching with external modules.

I also have to say I had a big expectation that experience with the recording and production would lead to me becoming more time-effective at those things. Sadly, it turned out that this was not the case. My growing experience made me notice more things I ignored at first, and while I figured out an effective workflow to end up with a satisfying product at the end, I never found a way to significantly cut down the time required to produce it.

I never properly touched the guitar throughout the month. I intended to, but since this is not my forte, I decided against it because I needed to have something worth sharing every day. I will likely overcompensate in February. This will allow me to try Iridium’s live granular with the guitar, too, something that I’m sure will yield interesting results.

Finally, there’s a bunch of Norns scripts I had my eye on but decided against picking up and learning just now. In particular, the breakneck pace of Jamuary made me decide against trying out zxcvbn, which might be the most complex music production environment for Norns at the moment.

Will I be doing Jamuary in 2024?

You’ll have to ask 2024 me. I think it isn’t very likely, especially in this year’s form. Maybe if I have a concrete artistic goal behind it? Last year I had a more focused idea with “Polyend Tracker combined with…”. This year it was a much more general “record every day with a live component without it becoming repetitive”. The end result was expectedly all over the place, as part of the goal was to return to things I already own but don’t use very often.

Don’t get me wrong, it was well worth doing this year. I learned a ton. However, doing it again would feel derivative and frankly simply too much work. Especially if you look at it from a YouTube publishing perspective.

Nobody watches your jams

Yeah, if you thought that “a video per day will make my subscriber stats go through the roof!” then I have bad news. Publishing at this pace is clearly diminishing returns.

I didn’t record the numbers properly, but I recall I started the year with something like 420 subscribers to my YouTube channel. Now at the end of January, with 31 new videos clocking at over 3 hours of material, I’m at 462. Barely over a single subscriber per Jamuary video.

The videos themselves don’t get any views from the #Jamuary or #Jamuary2023 hashtags, and YouTube itself doesn’t seem to show them to people much. You can get views by cross-posting your videos to dedicated synthesizer groups, but those are limited single-video boosts.

In concrete terms, all my Jamuary 2023 videos only gathered a little over 2,400 views on YouTube. That’s 85 views per video on average, but in truth, the bulk was at the start of the month. Similarly with likes on the videos. The 31 videos gathered 194 likes for a paltry average of 6.5 per video. Again, most of them in the first half of the month.

It’s not that I started phoning it in with my submissions. Quite the contrary! Sure, I clearly didn’t always publish what the viewers were expecting. That said, I suspect they simply got tired of Jamuary over its course. I can believe it, the amount of content the challenge generated is staggering. The Jamuary 2023 topic on the Lines forum gathered over 700 posts. The challenge is also scattered across different media platforms, with some musicians preferring YouTube, others Instagram, and the rest using audio-only platforms like SoundCloud or Bandcamp.

To compare, in the second half of 2022, I published three videos of me playing Iridium solo. Those three videos combined are currently at over 3,100 views, with 157 likes shared between them. That’s for barely 7:07 of content combined. Much less effort for arguably better viewer engagement. Similarly, my six Jamuary 2022 videos gathered 2,038 views (340 views per video) with 106 likes between them (17.6 per video). Compared to this year, that’s 4X better engagement in terms of views and 2.7X better in terms of likes.

Sure, comparing old videos to new ones isn’t exactly fair, as older stuff does accumulate views over time. However, save for a black swan viral video event, those views later are a long tail. The bulk of my views on my videos come in the first month or so.

Summing up

To end on a positive note, this was a challenge that I successfully completed, and that brought me lots of joy and fulfillment. The YouTube viewer stats above are only provided matter-of-factly. The main goal wasn’t to please an audience; it was to practice my craft, try out things I always wanted to but never found a good excuse for, and to push myself.

In the end, this opens up an interesting 2023 for me. I’m sure I’ll find all the new experience gained through this project useful in the upcoming months.

#Music/Jamuary/2023