A bit disappointed that this didn't cover the scan(s?)lation community that was the lifeline for anime and manga in the early 2000's before the categories became mainstream. How manga was translated was by rabid enthusiasts who would learn Japanese and scan and translate manga for little to no compensation, or even at cost to themselves. In that sense, it felt similar to the warez scene of the same era. Since then, anime has gone from a weird niche genre to cosplay supplies being sold in Joann Fabrics. I am still not used to how mainstream it is, but as someone who benefited from their work, I would love to hear more about those early scanslators.
The scene is still pretty alive and I cry for every manga that gets an official English release. Because at least the digital offerings often look visually worse than the fan scanlations. But they're the rights holders, so there's no competition to avoid insane compression artifacts or stupid localisation errors.
What does look visually worse mean? My (very limited) experience is the opposite: the quality of the translations and redraws were much higher in the official versions. And there’s way less “all according to keikaku”-type translations.
I've found baked in aliasing artifacts of tiny dots to be a common issue. A good scan would give you a high enough resolution that would allow you to zoom in far enough that it disappears.
Of course this is a bit of a double standard by me, there's a lot of lower quality fan scanlations, but I don't expect the same quality from a free fan service as from a commercial official one.
E:
Just one example: try reading the most recent Chainsawman chapter on the official app by Shueisha. There's very visible jpg compression artifacts around text and all straight lines have very visible blurring issues if you look at a panel in detail.
I remember the fan scanlations by, for example Jamimi's Box to be significantly higher quality than that.
There's a lot of parallels with quality issues in official anime streaming.
It's not unusual for those involved in the fansub scene to be obsessed with figuring out how to encode shows to look their best with a given codec, with settings often being dialed in per-episode, while services like Crunchyroll and Netflix seem to have one-size-fits-all settings that have issues that are readily visible even on cheap commodity screens. Most notoriously, they're bit-starved with dulled colors, smudged motion, and extremely bad banding in dark scenes.
Netflix has extremely elaborate compression setups to eke out every last bit, so if it looks bad on a reasonable connection then they're effectively doing that on purpose by being too aggressive. I would not call it lack of effort.
Crunchyroll seems to happily throw 6-8MBps at things, is that not enough with h.264?
I've been on gigabit fiber for years so I doubt connection speed is an issue.
Between the two Netflix is usually worse with dark scenes in anime. If I had to guess their settings are better optimized for live action shows/movies, which to my eye generally fare better with their encoding than anime does.
Crunchyroll is generally ok-ish but even at 6-8mbps it still sometimes struggles with high detail (grain for example) and motion, and as I mentioned earlier sometimes colors can be a bit washed out. Banding is also sometimes a problem but not to the extent that it is with Netflix.
Getting a higher quality from scans is not unexpected either, because it’s a fan service.
So for HQ teams there’s often a love of the work involved, and more importantly there are no ROI calculations. If the redrawer wants to spend hours on SFX (then rightfully moan about it in the credits) they usually can.
These days it's not uncommon for scanlation groups to be in it for the donation money. Well I say "group" but it's usually just the leader pocketing the money and recruiting bored / enthusiastic teens.
The excuse is that they need money for buying books to scan or digitals, but imo unless they're working on many series, spending 1~3usd per chapter (at 4~8 manhours / chapter), isn't bad enough to warrant begging for donations. Plus the digitals tend to be free if they know where to look.
On a lighter note, not every group is like this, like you mentioned, some people just genuinely think some stories are worth sharing. Others just wanted to practice reading / writing a foreign language and figured they might as well write the translation down and share that.
I was a translator for an anime fansubbing group, so I have (good!) stories to tell about the fansubbing and scanlation community.
That's for another time though. For now, I'll just cut to the chase and copypasta something I wrote a couple years ago regarding translation quality between professional and amateur work:[1]
Japanese has a very rich history and is also significantly more expressive than English is, so academic teaching of the language can only do so much. You really need to immerse yourself in Japanese literature/media and partake in just ordinary day-to-day conversations to simply begin to understand the depth of it all.
Translation is also hardly an appealing job unless you find love in the art of translating itself, speaking as someone who has done translating as a hobby (both FGO and otherwise). Almost nobody knows or cares about the effort that goes into translations, most readers only care if it "sounds right" in English. Keep in mind that the vast majority of people who read the bad translation pointed out in the OP won't realize it's bad. It's very hard to keep yourself motivated when your readers don't and can't give a crap about your work.
So combine a language that is very different from English with people who probably aren't all that motivated anyway, and most professional translations end up being shit. The work is thankless and giving more than the absolute minimum quite sincerely doesn't pay off because the readers don't know and can't care.
Good translations require passion from the translators and the editors, passion that simply just isn't there outside of the hobbyists and enthusiasts who enjoy the art of translating rather than doing it for their paychecks.
I grew up watching fansubs and reading scanlations. I still prefer them.
As one simple example the word "nakama" is almost impossible to translate properly into English. It's like a combined "comrade, crew, squad, friend, family" type of concept that has no English equivalent. Scanlations will typically have explanations for certain things on the side which adds a lot of depth to many stories.
> Japanese [...] is also significantly more expressive than English is
My first reaction was to doubt it because I remember reading [0] that spoken languages have roughly the same bit rate ( 39.15 bits/s on avg ). However, bit rate is far from everything (and this is also written not spoken), so could you try to explain what creates the expressiveness compared to other languages you know?
One of my biggest problems when translating from Japanese to English is figuring out how to convey everything. It sounds easy, but it really isn't; usually the English text becomes far too long and unwieldy.
For a specific example, "itadakimasu" (literally "I will take it.") spoken at the start of a meal is a complete bastard to translate to English. "Thank you for the food." would be a translation of the core spirit of the phrase, but it lacks the elegance and respect of the way it is conveyed in Japanese. "Amen" would be a western cultural/religious counterpart that most westerners could relate to, but this adds religious undertones that are not in the original Japanese.
Not to mention, "Thank you for the food." also overlaps with "gochisousama" (literally "Mr. Feast", "-sama" is an honorific) spoken at the end of a meal. Two very obviously distinct and different Japanese phrases, but the English translation ends up being more or less the same.
None of this is to say English isn't expressive, but Japanese takes the expression game to a whole new level; yojijukugo[1] are the epitome of this using just four words to construct and convey what would be entire sentences or even paragraphs in English.
Isn't it just problem in general translation rather than specific problem with Japanese? I couldn't think of any good translation for "Amen" in Japanese that is as compact either. Or something like "Bless you", whose concept doesn't exist in Japanese.
Definitely, langages differ in plenty of nuances, especially cultural. With the possible exception of the very technical and matter-of-fact stuff there’s a huge amount of taste and artistry involved in pretty much every translation, with the exception of very close languages (commonly close enough to be debated as dialectal).
That's like saying most programming languages fall within similar runtime performance, but translation is also like porting codes. Doing that line by line or per-function especially for human reading can be hard and/or weird and/or borderline wrong, like you might have to stuff class definition for a struct, guess variable types from name, explain "// onload means entrypoint in HTML", things like that.
Japanese is, for the most part, not more expressive than English, it's differently expressive. Same for every other language. The bit rate you mention is mostly a limitation of how quickly speakers can speak and listeners can listen. The cohort of listeners and speakers mutually agree on how to bucket the spectrum of experiences into a finite number of words, but different cohorts can bucket experiences differently.
In practice this means there's a lot of English words that get smushed together in Japanese[0], and a lot of Japanese words that get smushed together in English[1]. Japan also has a lot of shared cultural experience that Americans or Brits won't recognize, and vice versa, which translators have to spend time explaining.
Japanese tends to be vague, nonspecific, and the Japanese dislike directly saying what they mean. Since English should be specific and direct, there's a lot of interpretation and elaboration that needs to happen as part of the translation process.
While I get why the exaggerated "all according to keikaku" is poor translation, I've always been a fan of things like translators notes and not trying to translate things like jutsu and rasengan (anyone remember spiral blast????)
Maybe commercially it sells better to localise, but when you're consuming foreign media you're also sub/consciously discovering their culture. So having little notes about the little sneezing trope brings the world a bit closer one step at a time...
There's some anime and manga that you can't localize, you can just explain. Monogatari has so many cultural references and language jokes that its impossible to localize without losing everything that makes it what it is. Or something like Azumangha Daioh which has a character with many jokes built around their accent, you can't localize that.
As I recall, they localized Osaka with a Texas accent, which worked perfectly - the same drawling and also associated stereotypes about businessmen/greed/money/nouveau-riche/big-things.
"Localization sells better" is downstream of having a properly translated work. Deliberately not translating "jutsu" or "rasengan" only works because Naruto is a show about ninjas... debatably. If we translated everything in Naruto then everyone would realize that it's actually a show about ramen.
Officially licensed translations generally promise some degree of consistency, but are often (though not always) mediocre. Scanlations are more variable (of late, one not infrequently sees machine translations...), but peak higher.
I disagree. Official translations are quite often worse. They try to localize far too much. The cultural peculiarities (non-western) are the kinds of things why many people prefer manga/anime over more western media. Removing them is silly and probably the reason the fan translation scene is and was popular. While an official translations usually means that you have no fan translation alternative for the work, but you can just go watch/read something else instead.
In Yokohama Kaidashi Kikou the images are dark to the point of making it hard to see things, in both the physical and digital editions. This isn't the case for the Japanese versions (including the new digital edition) or the fan versions.
I suspect that even if an official release is available (with free samples) potential buyers tend to check out the scanlation first. Mostly since checking for updates on websites that aggregate fan scans are more convenient than checking 3ish publisher sites.
But anyways, I'm playing devil's advocate but I'd imagine a sloppy scanlation of chapter 1 has very real consequences on interest in the later / better official release. By sloppy I mean distracting grammar mistakes, poor image quality, etc. I know I personally wouldn't give the official a chance if I mistakenly thought the plot / art was meh from the scanlation. And that's bad because it doesn't support the authors, who tend be super overworked.
Agreed, as someone who did some scanlation back in the 00s and still follows the scene (which remains quite vibrant) that was a real disappointment. Fan scans lead the way on all the things they're discussing, with a lot of debates around various translation choices and work put into sound effects and redraws despite having no access to originals. I think it lead the way for commercial translators and showed there was a market and demand for getting close to the originals, and sense then has also been a source of effective market research in terms of what is popular. There are a number of wonderful series also only available to non-Japanese readers via fan scanlations. No sort of coverage of that was a surprise.
Even older than that, even in the 90's there was a large community of folks who had an Amiga and genlock and would translate and distribute subtitled VHS tapes of somewhat esoteric shows like Marmalade Boy and Kodomo no Omocha that did not fit into the genres that were getting commercial releases at that time.
Also, didn't mention the entire run of Mangajin[1], which was quite the resource for folks learning Japanese back in those days.
Ahhh yes, the original fansubs. The good ones came with liner notes, explaining linguistinc and cultural nuances.
And there was substantial debate in the fandom concerning the ethics. How much margin over the price of a blank tape was acceptable before it became a for-profit piracy operation? Should fansubs cease distribution when a show got an official release? Even if that release was dub-only or had a worse translation?
Official translations often (1) censor a ton in both the text and the art to satisfy Western Puritanical sensibilities and (2) have drastically less cultural context than scans do --- scans often will have notes about references to myths/stories that manga are referencing or influenced by like creation myths or relevant stories/concepts from Buddhism or random bits of cultural stuff like why a character might sneeze randomly after being talked about.
I've yet to read an official translation that's as good as scanlations, and the notes are often more creatively integrated into the scans in a way that are useful when encountered whenever possible.
I still subscribe to the official translations out of some sense of moral obligation to send some money back to the creators, but scans are faster and higher quality often enough.
That's because official translations follow the rule of “Readers in English should be able to enjoy the story without thinking about it being a translation,” - an impossible standard.
Think about what it requires - to cut out or replace any signs of the culture it came from, that would cause any friction with the target culture. To trick the reader into thinking the world is one big homogeneous mush, where works coming from the other side of the globe don't contain anything that would so much as confuse them. Nothing that would require a translator's note or explanation.
It's not translation, it's butchery. It's making a sausage come out on one end, no matter what parts come in at the other.
Yeah I understand the constraints. I just wish there was an official version that was targeted for people like me instead of scanlators getting lawyered into submission and we end up with less.
> Japan has a longstanding bias against lefties, and many left-handed children were trained to use their right hand. So manga artists and their fans weren’t crazy about seeing left-handed warriors in the U.S. versions.
That's probably not the angle I would've chosen if the intent was to garner sympathy.
Not really. Given that the article's argument was "The old method of localizing manga was flawed because X, Y, and Z", I don't consider Y to be a good supporting detail when it essentially boils down to, "Japan has an irrational hatred of left-handed people."
One is germane to the discussion. "I'm disregarding your argument because don't think your username is very professional" isn't, unless the discussion were perhaps about lack of decorum in online discussions.
Snapshot seems a bit incomplete. Portions of the text are difficult to read, and it ends rather abruptly. Thinking it was probably cut off before the end.
Not gonna lie, I was expecting something more in-depth, like the complete history of manga translation from the Studio Proteus/Toren Smith era and before, to now.
