What's up. I'm Soldev, a 3rd-rate developer. Even though I'm 3rd-rate, I've been doing the bare minimum a guy who makes a living through dev work can do—projects like long-awaited Korean fan translations, ROM hacks for imaginary characters, and building handy software tools.

AI Translated Image

Vampire: The Masquerade – Bloodlines (VtMB) 1... it gives off that musty classic scent. It's known as a tragic masterpiece, but (1) you'll want to toss it in the trash if you don't install the Unofficial Patch because of the bugs, and (2) with VtMB 2 flopping in October 2025, the demand for the previous titles has dried up. Because of these reasons, it's become a classic game that only those 'in the know' talk about. It has a massive amount of text (about 600 A4 pages), the content is complex, and since it's based on the World of Darkness TRPG setting, it's mostly for hardcore fans. Plus, the barrier to entry is high. I also played it briefly as an undergrad and then just forgot about it as I spent my youth. Then, late last year, I saw a gameplay video that inspired me...

AI Translated Image

Someone was actually playing it by using MORT for real-time OCR translation. Basically, a game from 2004 hadn't received a multi-language patch for 21 years. I checked, and there was no Korean patch either. There were attempts in the past, but since this game used a modified early version of the Source Engine to show off expressions (which was wild for the early 2000s), the trade-off was that they completely broke the Unicode support. I looked through old references and previous attempts, and started researching around Christmas last year. I posted the technical details on my dev blog, but man, it was a series of trial and error. The first part I succeeded with was the in-game dialogue, which was relatively easy to handle with an injector:

AI Translated Image

AI Translated Image

AI Translated Image

That part alone was pretty emotional, but after that, I started feeling like I was catching the Malkavian madness. First off... the default output uses CP949 bytecode. But the problem was that cutscene subtitles had a completely different output method:

AI Translated Image

AI Translated Image

AI Translated Image

Then there were the quest and item display issues...

AI Translated Image

AI Translated Image

AI Translated Image

AI Translated Image

Anyway, I've done the rough translation, but I still need to fix the alignment, lore terminology, and parts where the phrasing is too long. I got totally burned out at this point, so I ran it through an LLM and somehow managed to solve it.

AI Translated Image

AI Translated Image

I figured it's playable enough now with the basic translation, so I'm sharing the 1st patch. This patch covers rough translations for dialogue, cutscenes, loading screens, and pop-ups. The 2nd patch is planned for the first half of 2026.