May 2026 progress... I will make game? (Patreon)

This month, I tried a lot of different things. Actually, maybe too many.

I spent about half the month on my main work (making Pantra-chan videos), but my scheduling was poor and the fingering video still isn't quite finished — sorry about that. That said, the experiments this month were important work that will shape the future direction of seismic, so bear with me.

A picture of Pantra-chan running out of steam. And so (?), here's my activity report for May.

Searching for a way to make games like my old works without Flash (about 1 week)

I've mentioned this here a few times before, but I've been agonizing over what form to give my current creative project. At first I thought I'd make it a pure video work, but if there's more to it than just a girl showing up and having sex (which honestly isn't bad — it has that doujin-game feel 😅), I'd need a fair amount of runtime to explain her backstory and so on.

On the other hand, following the format of my old works — a video-based game — means dealing with the Flash problem. Speaking of which, earlier this year Adobe had that wild announcement that they'd make Adobe Animate (formerly Flash) unable to even launch. They walked it back within a few days, so it's not unusable, but…

Local file read/write is restricted. Save data is tied to the file path structure, which makes it awkward to use.

The video format flv is old, so the quality isn't great.

Even if a game doesn't work for some completely unrelated reason, I'll get emails forever saying "it doesn't work because it's Flash!?!?"

On top of all that, I just can't motivate myself to learn ActionScript 3.0 — a language with no future outside of Flash — while trying to make something new.

I had tested Unity and Godot as alternatives, but neither gave me that "scratches exactly the right itch" quality that ActionScript 3.0 does when it comes to video handling. And frankly, if I'm going to spend time on experiments that might not even have an answer, wouldn't I be better off just making an erotic video? That feeling was hard to shake.

Having AI suggest approaches and write the code

So I asked an AI service for suggestions, and it turned out that rather than a well-known game engine, a software development framework called openFrameworks might actually work better for video handling — so I said "alright then, you build it" and had my AI teacher put it together.

There are many advantages to having AI do the building, but the big one is that it lets me run experiments while saving my own time. It's perfect for tasks that might not even have a solution. Meanwhile, I can be making 3D CG erotic videos.

And the conclusion was: openFrameworks looks promising. It's C++, so it runs fast; of course it can save files locally; and it can scan data folders, making content expansion easy (ActionScript 3.0 won't even allow that).

It also supports HAP, a video format with an alpha channel, which makes it possible to overlay cross-section views or sub-videos from a different camera angle on top of the main video. The image quality is naturally better than flv too.

So layouts like the one above — cross-section positioning — are easy. It's also a good move away from H.264 and H.265, which have become somewhat legally risky for smaller studios to use in games.

And the key feature from my old works — preparing lip-flap videos, syncing two videos together, and toggling display to match the voice in a brute-force lip-sync 😅 — that worked too.

The yellow-green graph in the top left is the voice volume; the red line is the lip-sync threshold. You can see that in the first pass, the threshold is well-tuned and the lip-sync is working nicely. Since Patreon still doesn't allow videos in adult-oriented posts, I've used a GIF instead.

This connects back to last month's "Wolf girl with you Together Multi-Language Project." Even though the update is already nearly complete, if I build the game framework in openFrameworks instead of insisting on ActionScript 3.0, I could update my older games beyond just "Wolf girl with you" — and stack future "Pantra-chan projects" on top of the same foundation.

Exploring making a game together with AI (about 1 week)

Beyond recreating the same format as my old works through different means (AI + openFrameworks), I was also personally interested in making something more game-y — a more traditional eroge where you actually play a proper game and get rewarded with an erotic video for clearing it.

May rolled around, and with it Golden Week. My brain had gone to mush, so as a little independent study project, I ran some experiments in that direction too.

I started with some light tests in the development environment I'd already set up for the Flash replacement work, and then — out of sheer opportunistic curiosity, thinking "hey, if I can make a game on the side, why not?" — I told my AI buddy "do it !!" and this is what came out.

It’s still a bit rough around the edges, but it’s mostly playable, so I’ll be distributing it to members on paid plans (including the $1 plan). I’ve been thinking about setting up a Patreon premium page, something I’ve wanted to try for a while. However, the minimum price for a premium page was $3 🤣. It’s actually cheaper to sign up for the $1 plan.

This game was made as a test, with the specific target of keeping everything — difficulty, level editing, all of it — within what I could manage on my own.

Quite a few people ask me, "Is your next project going to be a video or a game?" What they probably mean by "game" is something where the gameplay is almost entirely oriented toward the erotic content — definitely not the kind where you "battle through stress to finally earn your reward" like this game. Yeah, probably not that.

But… honestly, if I feel like making something, I can't predict which way I'll go. If I'm not enjoying what I'm making, I can't keep going.

Part of why I've been leaning toward games over videos comes down to something simple: even if it's a straightforward game, I want that clear narrative of "fight and win" or "chase and capture." That's all I really need.

If there's a sexy female enemy character you defeat, or someone you corner and catch — that's enough for an introduction. After that, the video can just be an erotic video from the very first frame.

But don't neglect the main work

So… I spent about half the month on all the experiments above, and as a result the video work slipped a little in the second half of the month. The cross-section animation where fingers move in and out is nearly there — just a bit more to go before it's done.

This is a test clip; a version closer to the finished product is available in the paid section.

Free ticket code

So — lots of experiments, lots of results — but the multi-language support, the post-Flash testing, the weird little dungeon game… none of it really counts as fulfilling the rewards that come with supporting this fan site.

Since I wasn't able to deliver a proper erotic video this month either, I've already sent out free tickets. Please use them without hesitation. I'm hoping to release the fingering video early in June.