Patternflow

---
title: "What Is Patternflow"
date: "2026-06-29T21:30:00+09:00"
excerpt: "Where am I headed"
cover: "/journal/what-is-patternflow/crowd-supply-pre-launch.png"
coverCaption: "The recently opened Crowd Supply pre-launch page"
highlight: true
draft: false
---

I recently opened the Crowd Supply pre-launch page. A month after signing the contract. I really want to market hard and hit 150 subscribers fast. I emailed the 100 people on the waitlist to tell them it was open, and as of today, about two days later, it looks like under 20% have converted to subscribers, which is a letdown. What went wrong? Or is this just how it goes? If it were me, and I opened a link for a product I wanted to buy on Instagram only to be met with a subscription to a crowdfunding platform I'd never heard of, what would I do? Would I readily sign up and buy? How do I make that transition smoother? Maybe I should write down why Patternflow had to choose that platform. Rather than making people go verify for themselves that it's a trustworthy platform, it'd be good if I laid it out and handed it to them. Yes. Let me get to that soon.

Then I should share progress again on the Arduino subreddit. I don't know if it'll get the kind of response those first posts did. So I'll just have to try. I also need to organize the blog posts so far and write something for Hackaday. Honestly, the act of writing and submitting a piece about Patternflow with my own hands isn't something I welcome. Tailoring it to a platform's character, and combing through a massive amount of information to pick out the most fitting bits and shape them into a piece, takes no small amount of energy. So part of me had quietly hoped Crowd Supply and the like would write the pieces and do the marketing for me, but of course, this is my job. No one knows Patternflow as well as I do, so in the end I have to do it.

Still, I'm curious how it looks from the outside. Because I know it too well, I can't tell how it reads to someone seeing it for the first time. There's so much information that I can't see it. Analyzing traffic earlier, I found that a place in Japan called FabScene had run a piece on Patternflow. It turned out to be put together from the two posts that had made a splash on the Arduino subreddit. I'd love it if things got made like that, on their own. There's truly so much material on the blog to draw from! I just noticed two pieces went up on Hackster.io as well. So grateful! I hope others feel free to take the material and write too. Personally, I'd love to be featured in an art-side magazine like CAN (Creative Applications).

For marketing channels, I plan to keep focusing on Instagram as I have been. A post about Patternflow someone put on X went viral once. Amusingly, my name was written as Seungheon rather than Seunghun. I guess "hun" reads as "heon" to some. I tried posting from my own Twitter account, but as usual there was no response at all, so I'm not going to bother with that side. Social media really does reward consistency, it seems. Just managing Instagram is all I can handle, so let me focus only there. Even that isn't easy, though. It's been over a month of steadily posting patterns, and I still can't get a read on what elements people like. Today I filmed the same pattern being manipulated in different ways and experimented using the pilot reels feature. I think I have to go like this, forming hypotheses and testing them one at a time in a controlled setting. Thinking of it as experiments makes it feel like it might even be a little fun. Keep at it long enough and maybe I'll develop a sense for what people like.

<figure className="journal-md-figure">
  <img src="/journal/what-is-patternflow/viral-x-post.png" alt="Patternflow post that went viral on X (by @Inspector_9)" />
  <figcaption>A post about Patternflow that went viral on X (by @Inspector_9)</figcaption>
</figure>

I want to start studying sound soon, too. It's been a good two months since I set out to make content pairing Patternflow with sound work. I keep putting off the sound study with busyness as my excuse, and I feel a bit pathetic about it. But it's also true that I genuinely don't have the energy for it, so I need to change my approach. I have to change myself so I can take sound work not as study but as rest, as play. I've been unhappy lately about YouTube and reels stealing my downtime anyway, so let me try it little by little. No, actually, this isn't the kind of thing you do little by little. Don't overthink it, just start. Starting is everything. Going this direction, I'll be able to collaborate with various musicians, or supply them with the product, and market naturally that way. And my intuition says this is the market where it sells best.

The other direction, the wish to treat Patternflow artistically, is the strongest of all. It can't be helped, since it started as an artist debut in the first place. I've gotten messages from various artists and people in the field. A lot of them wanted to collaborate, but as we talked I just couldn't get a feel for how the collaboration should actually work. That's because my own thinking about Patternflow still isn't cleanly sorted. So for now I'm responding with a let's-talk-slowly approach, using the crowdfunding as my excuse. I want to collaborate a lot too, but not knowing how it should work is frustrating.

And then I come back to this question. What is Patternflow? Is it this one product being funded? Or is anything that combines an LED panel and knobs Patternflow? Or, more abstractly, is Patternflow the purpose, carrying on Nam June Paik's spirit to bring interactive media art to everyone? On pattern generation, right now I only make patterns at 128×64 resolution, but someone says they'll change it to a different resolution and work that way. Welcome, of course, and later I'll design a structure to make that easy. So could a really, really huge media façade be Patternflow? Conversely, could something small enough to hold in your hand be Patternflow? My current sense is roughly: "an LED matrix carrying a creative-coding pattern whose parameter values change according to sensors," that's Patternflow.

But does there have to be a sensor? Can the pattern, on its own, not be Patternflow? The reason I'm chewing on this is that it becomes the criterion for whether to scrap or keep the community goal I newly set last week. Infect every ugly LED panel in the world with Patternflow. The thing is, most such LED panels have no sensors. It means taking the ugly signboards you see on the street, the ones that only deliver information, and dyeing them with beautiful Patternflow. That way people can easily find cases to apply Patternflow in daily life, and even make money through it. I want it so that not just me but anyone who's a member of Patternflow can use this to earn, too. Seen this way, the signboards along the road look like a gold mine.

But then, since there's no sensor, no matter how much it runs in real time, visually it's no different from baking a video and dropping it in. Which collides with my earlier decision to remove the video feature, declaring that "simply baking a video and putting it on the panel is not Patternflow!" What's the solution? Should I be more generous and just say "video is Patternflow too!"? Or, "it's a pattern made with Patternflow, so it's Patternflow!" That has some logic to it. Or, when putting it on a sign, should I make the parameters change periodically, so I build an LED signboard that keeps shifting over time? That looks fun too. Let me think it over slowly. To begin with, I need to find out how ordinary LED signboards actually work. If they're compatible with the ESP32 that'd be easiest, and if not, some extra work for expansion will be needed. And slapping a pattern on isn't automatically good, so I'll also have to research patterns that don't ruin the information delivery but actually heighten it.

I just got back from the gym, and partway through, my thinking sorted itself out. Removing video and adding patterns to signboards don't conflict. Because I'd already been counting the audio-reactive feature as Patternflow. It was never the case that something had to be hand-operable to be Patternflow. So if a pattern is being generated in real time and its flow can change by some means, then it's Patternflow. That's it.

At lunch today I stopped by the Seoul Design Incubating Center. I got to meet founders running their own brands, and there'll be more networking with them ahead. Chances to collaborate will arise naturally, and how will that go? I'm curious. I also think I'll need to find a professional technician or an electronics supplier going forward. Well, that's not urgent. It's a secondary piece, so let me not rush it.

And then, naturally, I found myself thinking about the military. I have to enlist within three years. (In Korea, military service is mandatory for men, and putting it off risks losing the alternative-service option and being assigned to active duty.) That naturally becomes an ending for the Patternflow business. The Patternflow community or spirit won't disappear, but I'll have to let go of the power to hold the reins and lead it as a business. It doesn't feel only bad. I was never going to do only Patternflow forever anyway; it's just one of the works of my life. And power disappearing means responsibility disappears too, so there's something fun in getting the chance to start a new project unburdened.

What do I have to keep doing right now? Improving the PCB and the 3D-printing model are work I've always done, nothing new, and I have to keep at them. If there's something I newly tried last week, it's without question the efforts to revive the Discord. I started right after writing the last entry, and looking at it now, it feels like it's in a decent state. The fact that four people did the self-introduction, one of those mechanisms, is both surprising and moving. Honestly, since it's an online community and not an offline gathering, I'd thought people might find sharing personal information uncomfortable. My own older brother, in fact, showed some reluctance about the introduction for that reason. But I really wanted it in. Because I want this to be not just an online community but something closer to a tightly connected gathering. I worried a lot, but there were people who did it anyway, and I'm relieved.

For others to contribute easily to expanding Patternflow, there's a precondition that has to be met. The leader, meaning me, has to share the view I'm seeing. What future I'm picturing and what has to be done for it, and clearly where the parts are that you can freely touch versus the sensitive ones. The thing is, I'm not certain either. I like this direction and yet it also feels like it'll change soon, and I'm wary because once I say it out loud I feel like I'll be bound to it. For now I'm putting it off, but I trust that through the process of writing like this, the outline is slowly taking shape.