Show HN: Translate Japanese manga and Korean manhwa with Chrome extension (pawakalabs.com)
If you are a manga or manhwa lover, you must understand the feeling of waiting for your favourite series being translated into English or sometimes your native language.
Now, you can translate them in real-time with Fakey Chrome extension!
54 comments
[ 3.9 ms ] story [ 99.7 ms ] threadAlso this is a manga thread now. I will go with Oyasumi Punpun as one of the best mangas I have ever read ;)) and I recommend it to you all. Dont look up anything about it just read :)
There's way too much nuance in both Korean and Japanese for this to reliably work. Chinese, from my limited study is a bit closer to English in grammar and structure so that might work
But you trust scanlation groups? Neither will give you perfect, professional-level translations.
>Chinese, from my limited study is a bit closer to English in grammar and structure so that might work
Mandarin is full of nuance, and it's no closer to English than Japanese is. It has the Subject-Object-Verb grammar structure, just like Japanese and Korean.
This isn’t correct from what I’ve studied in both Japanese and Mandarin.
https://lptranslations.com/learn/chinese-vs-japanese/#:~:tex...
> For example, Chinese verbs are not conjugated and only have one form, whereas Japanese verbs have a wide range of conjugations and particles. Plus, Chinese is an SVO (Subject+Verb+Object) language just like English, so sentences are easier to make and interpret. Vice versa, Japanese is an SOV (Subject+Object+Verb) language, meaning you do not say: "I eat sushi" but "I sushi eat".
Mandarin often moves the object to the front for emphasis, creating an OSV or SOV structure (e.g., 寿司你吃, "Sushi, you eat" or 你寿司吃了吗 "You sushi eaten?"). This isn’t true SOV grammar but highlights how meaning shifts through word order in ways English can’t replicate without rephrasing.
The nuance in Mandarin often comes from particles that take on very different meanings depending on how they are used (e.g. 了, 的) and contextual cues rather than conjugation. For instance, 吃 "eat" becomes past tense with 了 (吃了), future with 会 (会吃), or continuous with 在 (在吃)—no verb changes needed. But if you say 要吃了, it actually means future tense of "will eat soon"!
Meanwhile, Japanese relies heavily on verb conjugations (食べる→食べた) and postpositional particles (は, を) to mark grammatical roles, in a way making its structure more rigid and easier to interpret. Personally I found Tae Kim's interpretation of "Japanese isn't SOV, it's actually V!" to be useful.
Both languages share subject-drop tendencies (like omitting "I" or "you" when contextually clear), and compound-word formation in both languages from the use of Chinese characters (kanji) adds another layer of contextual interpretation.
I often prefer fan level over professional level because they are targeting different audience. As far as quality goes, there is a range and sometimes I skip something because the quality is too low, but I see plenty that does a good enough job.
Part of it is that there is no such thing as a perfect translation because there isn't an exact equivalent in another language. For someone with no knowledge of the original culture or language, there is some translations that will probably work best, but the more one knows about the language and culture, even a small amount picked up just from consuming other items, the the more likely a different translation works better. For a definite concrete example, how should one handle honorifics like chan, san, and kun.
Just because neither are perfect doesn't mean they are equally bad, though.
Not in manga space, but Microsoft had been notorious for nonchalantly shipping crazy MTL errors for past 3-5 years, e.g. "Copilot Child Adoption Kit", "Reply in cost estimate", or "Print in scenery". For manga and entertainments, more likely modes of errors would be wildly fluctuating pronouns, genders, personas, formalities almost in styles of Monty Python satire.
These are less likely to manifest with language pairs that are closer together like English and German, and less likely with human translators who can trivially go through pages to read with full context and/or write with consistency.
For a while now my coworkers have been free of charge when they're available. It's pretty insane
For old translation systems, you are absolutely right though.
And context isn't always clear. It depends on where the conversation is taking place, where the speaker is looking, where speech bubbles are positioned, and if the author intends to mislead readers, the information necessary to understand the context might even be revealed in future episodes.
The fact of the matter is there's going to be some nuance lost in even a good human translation. For example as far as I can tell in Korean martial arts novels a word is used which can mean either "demonic" or "unorthodox" to describe a group in opposition to the "orthodox" martial artists and a human translator has to basically pick one and go with it removing whatever nuance is in the original.
I'd prefer an official translation of course but the AI one is "good enough" if none is available.
Formal vs informal speech is another thing that doesn't seem to translate- but I don't think you need to understand every nuance to enjoy what is in many cases a straightforward genre story.
Manga is much trickier since the AI if reading the page may not be able to pull from the context of other pages the way it could in a novel.
I think it might be a tad premature to say it's over for them.
That Time I Got Reincarnated as a Slime, one of the more recent lightnovels I've been reading (via "Slime Reader") has an extremely noticeable drop in quality when they moved to machine translations (and that's with post-translation editing).
However, the speed at which things can be translated is a huge positive, and I'm sure translation tech will get better, so it's probably just a matter of time.
I have always had huge respect for both manga and anime translation groups.
"How does Fakey work?"
"In your dreams" (pricing)
It's cool it replaces the text in the page, makes sense, wonder if it's an overlay or literally modifies the image on the page
Maybe it uses contour finding for the speech bubbles
The other Chiyo
It uses overlay. Tried to render the text directly on the image but it is two to three times slower.
IIRC Japanese usually defaults to a generic pronoun that means "that person" rather than more specific pronouns for "he" or "she," so there were some times where I think the AI had to guess on whether "he" or "she" was the appropriate equivalent and just guessed wrong. In at least one instance it might've also been that a character was intentionally using feminine address to refer to an outwardly gay man, which could've been an interesting nuance if translated more clearly, but as things were it just left me confused.
It is even worse than what I described so far since if the topic of conversation is people’s health, someone might say “Alisu wa” to ask how Alice is and another person would reply “genki (desu/da)” to mean “Alice is well”. “Alisu wa” is just the subject. It is in that context saying “[Is] Alice [well]?”, while only saying the subject. The “wa” part marks it as a subject/topic. English has only a vestigial concept of subject forms of words in I/he/she/they/we. Japanese has much stronger concept of a word that is a subject by appending “wa” to the end of it (or “ga”, but that is a rabbit hole). Thus “Alisu wa” alone is really just saying a subject. In any case, if you are not paying attention, you might mistake the speaker as claiming to be well rather than the speaker claiming someone else is well.
Machine translation often processes text sentence by sentence. That is incapable of determining the subject in a multitude of Japanese sentences since the information needed to determine a subject is in a prior sentence spoken by another person. The machine translator thus must guess and it is often wrong. You will have better luck if you dump text into a high end LLM and ask it to translate since then it will consider prior sentences and have some idea of what each subsequent sentence actually means.
Also, Japanese does not really have more specific pronouns for he and she. Like Chinese, they only have 1 third person pronoun “kare” (he/she/it). In recent years (around the past 160 years from what I have read), it has become popular to use it mainly for males, while the word “kanojo” (girl) is used in its place for females. However, “kanojo” is not a true pronoun and is literally the noun girl. At least, whenever I read “kanojo”, I read girl, not “she”. I suspect machine translators would also read it that way, as it is what is really being said.
That said, I have studied a little Japanese, although I am far from an expert. I replied mainly because your understanding of the translation issue differs strongly from reality. I hope this reply is helpful.
I saw a YouTube comment '中国要素隠す気もないの草' translated to "I don't have any intention of hiding the Chinese elements". The actual translation is "[Ubisoft] aren't even trying to hide the Chinese elements [in Assassin's Creed]". The Japanese sentence has no subject so you need explicit context (in this case the whole content of the video) to translate it correctly.
You also get craziness like 私は美味しい meaning 'I [find the food] delicious' or 'うな重が食い逃げした' meaning '[the person who ate] eel over rice fled without paying'
That said, 私は美味しい illustrates how you really need an entirely different way of thinking about language in order to understand Japanese. If you want to report you are tasty as an English speaker might misread that upon hearing は is a subject marker, you would want to use “ga” (が) to mark what is being described as tasty. If you bizarrely want to say you are tasty, like how an English speaker would interpret that upon hearing は means subject, you would want to use が instead of は.
In a slightly more normal but still quite bizarre situation, if you wanted to say you are not tasty, you would say 私が美味しくない. That would be useful if you wanted to tell a Japanese speaking cannibal that you are not tasty. It also would look extremely weird to an English speaker in comparison to the version that declares yourself to be tasty, since two characters were added to the end of tasty to negate it, unlike English where “not” is added before the adjective, but that is how Japanese works.
I think GP meant うな重は食い逃げした. This sentence can be interpreted as the same as above or "[the person] ate eel over rice [but not other dishes] and fled away".
But... come to think of it, even a native speaker may not be in a position to deny someone's finding about interesting similarities between Japanese and English sentences and possibilities of interpretation.
As I am still learning, I have been trying to be open minded about possibilities. That helps when encountering things like 風邪です and 空気がおいしい. The former as you know literally means wind, but also can mean having a cold (and it would be very socially awkward if someone’s name was 風邪). The latter literally calls the air tasty, but means the air is fresh. The downside of being open minded is that someone can tell me something wrong and I will try to interpret it under the assumption that it is correct.
So you're saying that this sentence makes no sense unless は is used?
>よっば
Just the fact that it misrecognizes this character means at least 10% of it's translations are going to be nonsense. That the creator puts this video up as a demo doesn't instill any confidence.
For any real translation app you will need something like deep research with a much longer context window and a much lower hallucination rate, better FIR and a slate with the input words and your confidence in them. Arguably a conscious robot is required.
Until OCR is a solved problem manga won't be a solved problem, and even then because I guarantee there is more wacky shit in comic books than in government and corporate documents, it still won't be 90% and that's just OCR
If you actually understood the scope of the problem, and were acting rationally, you would just go learn japanese. Of course if you compare current AI to bad localization you can say it's a million times faster and more accurate even though in reality it's worth than any bilingual person reading it and reporting the meaning. As I ramble perhaps pretraining with "this is the meaning" at various clause, sentence, paragraph and page levels for every popular or useful media will make Gemini or ChatGPT seem like magic.
That said, learning Japanese is an extremely slow process and even if one does learn enough to understand a novel, one’s reading speed will be extremely slow, even if one is lucky enough to remember enough that it is not necessary to use a dictionary every other kanji. Even upon learning Japanese, it is quite tempting to dump the text into a LLM to have it translated.