Welcome to our 3rd Beastly DevLog! Where we give all our wishlisters and patient fans a little sneak peak into our process of localizing this game into English, as we speed closer and closer to launch day.
TWO BEASTS OR NOT TO BEAST IS COMING MARCH 20TH!
Next up, Fakku's own Image Editor & Letterer Extraordinaire, Randall, has popped by to give a little behind the scenes insight into the work he did on the game…
Visual Friction & Graphic Immersion
Hello, this is Randall (often credited as Uranohoshi Graphic Design Club), I did the Image Editing and Logo design for Two Beasts or not To Beast. Normally, I usually work on stuff like manga editing, but a lot of those skills transfer over easily to games where large swathes of the graphics require Japanese text to be replaced with English text. Two Beasts or not To Beast was probably the most complicated games project I've done yet, with some of the original images from Japan not having layered files. This isn't exactly uncommon, and honestly a big part of what I do for work already, but it introduces a larger problem: Faithfully recreating/redrawing some of the assets to match the Japanese ones in order to not ruin the experience for the player.
In my opinion, what I like to call "Visual friction" is one of the biggest issues with making a localized piece of media (whether it be manga, games, or anime) a seamless and immersive experience for the person experiencing it. Think things like shoddy redraws, a bad subtitle font, text too big or too small, colors being wrong, etc. While most people might not notice them or know how to single them out, they create a sort of mental stumble to the reader/player/etc., even if they don't consciously realize it. With this in mind, I approach all of my projects the same way: Adapt everything I can, as closely as I can to the original, with as little "visual friction" as possible. Of course, as many people in this field can tell you, there are some compromises you must make just due to the size of assets and fitting English text into a small space while keeping visual consistency with your previous assets. One particular example I wasn't particularly happy with was the "Achievements" button, which is a long word in English versus the Japanese word for achievements. (I do not know what this word is, because I am simply the image editor. But I know for sure it isn't nearly as long as "Achievements". It's probably like 2 kanji max.)
My biggest hurdle was the loading/eyecatch screens by a wide margin. These were, quite frankly, a nightmare. As previously mentioned, these weren't layered, and were maybe the worst files to not have layers for. They are very visually beautiful images I think, with the parchment colored background and brown line work giving a very warm vibe in their presentation, and the red accents at the top and bottom envelop the whole image in a nice composition overall. Matching these colors was The Big Problem to overcome, as they had Japanese text that had to be replaced completely on the top left portion of every image, where there was usually some line work under the text. This required me to manually redraw the lines that *should* be under the text, using a lot of context clues and trial and error until it looks "right". This is a very time-consuming process to make it look "right", as even just a few mistakes here and there can make the effort look a little too obvious.
Here's what one of the "eyecatch" images look like before and after. There were around 20 of these, and all of them had to be edited.
I remember it being a long period of experimentation and color matching until I got it as close as I possibly could, and was pleased with the results. The brown line color changed ever so slightly under the pink accent via a multiply layer style, which was my main source of headaches and experimentation. If the color was just a little bit off, it would be incredibly obvious to people, and I don't want them to be thinking about that. If you're looking at the image and thinking, "What the heck is up with the color here? Why is this line not the same color as this one over here?", it distracts you from the cute anime girl on screen. A horrible thing to happen to anybody, and an experience I hate to have myself.
A screenshot from my experimentation PSD. I'm pretty sure I ended up using math somehow to find the correct color, and then changed its opacity until I found the correct value. I don't know math, so don't ask me how I did this. Even though these images show up for a few seconds on your screen, I need those seconds to be as frictionless as possible. This is, and always has been, my work ethic with everything I do.
Another memorable part of working on Two Beasts or not To Beast was the sound effects/onomatopoeia that shows up on screen during the erotic scene segments. Depending on who you ask, this is probably where my work shines the brightest, as it is among maybe the most recognizable things that I do. (You have probably seen my handwriting in font form without realizing it, but I sure notice it every time it's used.) There are 165 of these files, all individually done by hand with no images repeated. I always find it pretty distracting when a sound effect is obviously a repeat but just rotated/repositioned/resized in a creative way, even if I do find it a valid way to save time. I think probably 99% of people that aren't me don't notice when it happens, but I do, and that's a good enough reason for me to go out of my way to really make each unique. After all, I want to enjoy this game myself, and what's better than doing something for yourself in the future?
Come on down to the handwritten sfx store we got all your sex noises in different sizes and varieties
One of my other favorite images to work on was the "Mind break" screen that comes up when you fail one of the mini-games (which people definitely only fail because they're hard minigames and for no other reason). The text is [理性 決壊] (lit. reason collapse), which is is typically translated culturally to "Mindbreak". The "cut up" effect is quite simple: Rasterizing the text layer, using polygonal lasso to just "cut up" the area I think would look best, and moving the pixels aside to give the desired effect (I remember learning to do this specifically for making forum signatures for the halo 2 clan I was in!). For this text I ended up using the font BestTen-CRT, a free Japanese font that I'm pretty sure was the original font to begin with. However, this leaves the text in the middle awkwardly missing. The visual nature of it having a tapered curve in the middle reminded me of something pretty obvious, as I'm sure you are already thinking "haha I know that". The "destroyed" text is the same font as previously mentioned, using a similar style to move the pixels aside (rectangular marquee in this case).
see me in +r btw (i will lose)
I hope you enjoyed reading my little devlog where I was able to convey my thought processes, even just a little bit. I put a lot of thought and care into a lot of this stuff, so it's kinda fun to put those thoughts into tangible words where hopefully I don't sound like too much of a maniac. Working on this game was certainly a challenge, but I think the payoff was pretty good.
Guest Character Spotlight
Meet Kaede! The even-tempered, helpful store clerk hiding her true, feral self.
And Chifuru, the shy neighborhood friend and secret lover of all things BL.
Thanks for reading, see you next week!
Thanks for reading, see you next week!
