Scanlation Thread

I bet most fan "translators" wouldn't even pass N3.

Dumb frogposters.

Crying because your got sniped is a time honored scanlation tradition.

How to find interesting obscure authors? I feel like between Seirikogeisha, Garo, Beam and Ikki there's not much else.

>deleted

Attached: a5b887d8084366a7507370f5e6ed1e71.jpg (374x347, 39K)

non-anime/manga pic

Wew lad, since when jannies enforce that "rule"

I'm the only one with this issue?

Attached: 1557355308326.jpg (1440x810, 137K)

Apparently most translators in these threads actually know the language enough to read without OCR.

I'm not a translator, but I do have an untranslated manga I want to read, so your question is relevant to me. I personally just don't know what steps I need to do to properly look up kanji, since I lack the knowledge of what tools to use and what components to look for to identify with.

Attached: 1536058333411.jpg (640x480, 16K)

you can draw into google translate as well
or use a different OCR program
but yeah translators that have to transcribe have a hard time of it

Just learn radicals, those are the components you are looking for. You can use jisho.org to search by radical.

I don't read japanese but I do know chinese so you can just look up the Kanji on Chinese-English dictionaries which might be able to help.

There are websites that allows u to draw the strokes as well to identify the words.

I don't believe that.
There are always some Kanji that you don't know/forget, so you need OCR to recognize it.

Kanji tomo for PC, Yomiwa for android.

I never use ocr. I either take a stab at typing it out based on what I feel the kanji looks like it should be read as and if that fails I draw the damn thing into google translate.

You never need OCR to recognize it, you just need to learn how to look up kanji. Learn how kanji are constructed and written so you can use the radicals and the stroke count to look them up.

ocr is simpler, faster and more efficient.

Usually even moderately rare kanji have furigana. If not, you can probably type them by onpu. If it fails as well, you can always draw in google translate/search by radicals/ask a native.
Go back to anki you dekinai.

I second KanjiTomo. Also use ejje.weblio.jp for a good Japanese to English dictionary.

That's fucking retarded, jannies are fucking retarded.

Also, to all porn scanlators, stop fucking over leveling shit, or whatever it's called, you are just making censorship worse you fucks.

Example:
files.catbox.moe/2fs6wv.jpg
files.catbox.moe/gzcgwg.jpg

The second one doesn't even look better than the original you fucking retards.

>Not studying Heisig
Never, ever gonna make it

>anki meme

KanjiTomo app is quite heavy and laggy

Heisig makes remembering Kanji HARDER with its nonsense.

>That's fucking retarded, jannies are fucking retarded.
It's literally the first Yea Forums rule.

Attached: 5e2.jpg (554x439, 38K)

>It's literally the first Yea Forums rule.

It doesn't make it any less retarded.

And then you're fucked when it doesn't work correctly like OP is whining about.

>And then you're fucked

If you are fucked by something of that level you shouldn't be translating in the first place, fag.

Weblio's fine if you only have a browser handy but it uses volume 4 of the chujiten and volume 5 did a bunch of improvements to the example sentences so I'd use that over it. Plus I prefer a combo of the Daijiten and Advanced Favorite anyways.

The post in said
>need OCR to recognize it
>need
so pointing it to OP answers. You don't NEED OCR. Yes, it's easier, but it is not a NEED. People have done it with a dictionary before OCR is a thing.

>People have done it with a
Sure, but why do I need to regress my self? We're in the age of technology. I expect google to have a functional OCR, which I need for speed and convenience.

You don't have to look for obscure authors. Even popular guys like Atsushi Kaneko only have one series scanlated despite raws for other stuff being easy to find.

can you guys give me some tutorials on how to translate with google OCR? i have a couple mangas that noone has picked up for 12+ months.

If looking up unknown kanji is of any burden to you then you're shit at Japanese and you should stop translating.

You can't stop me.

bump

Digital raws are a curse. Everybody's too lazy to make a proper high quality scan these days.

Jisho
>jisho.org/
It's a godsend for finding any kanji I don't know.

Absolutely. I suck at Kanji and know I'm too shit to scanlate or translate anime. Unfortunately a lot of existing translations are made by people who are even worse than me.

Comment systems were a mistake.
People should just run their own self-hosted aggregators.
Fuck """communities"""

Why are you blaming the system instead of manchildren scanlators?

When you think that you've created something that cannot be ruined by manchildren, nature invents a better manchild.
The long and short of it is that manchildren are going to manchildren, but there is a certain global optimum system design that has yet to be reached which can mostly curb manchild activities, and our current system encourages rather than inhibits it.

its more easy try to learning basics and then use the graphic translator

jajajaja you fools, lol
who needs something translated just call me
for a few bucks ofc

Don't.

jajajaj no, Learn how to wrote a kanji thats all
OCR is for noobs likes you

Sharex's OCR function is pretty good.

Sharex uses a web service. You might as well just say that that web service has good OCR.

What are your rates?

Which role in a scanlation group tend to have the whiniest people Yea Forums?

>redoing a chapter that another group already scanlated
Why do people subject themselves to this?

PR/QC
mostly because they're impotent if no one cares to listen

How could physical scans possibly be higher quality than digital raws?

Just be a Chink to read the kanji.

Why would a scanlation group need Public Relations?

proofreading, numbnuts

Higher resolution technically. Digital raws you might be stuck with 959x1400 or worse, while you could scan a printed pages at a higher resolution like 1500x2250 or larger. That gives you a lot more options than smaller resolutions, especially if the scan itself is in good quality.

What's the difference between PR and QC?

Literally nothing. They're both useless unless the translator is also useless in which case they're slightly less useless.

are you scaling an image and loading it in the app or taking a picture of a screen? the picture method never failed for me, it even works fine in artist cg;s (you have to make it black and white first)

Which role requires the most effort to fulfill, assuming you have the skills for it?

It won't detect it well when the focus of the image is that blurry.

Many digitals are low resolution, have artifacts, and are blurry for whatever reason.

diff.pics/7bzDzUou9pny/1

11 years and I still suck. Does this outer glow look fine or should I make it darker?

Attached: fk me.jpg (1250x1014, 218K)

go away, commie

Most people are either too lazy or too incompetent to clean nowadays so they rely on digital raws to avoid having to do extra, tedious work.

redraws

Redrawing is very tedious, especially if you replace the sound effects.

Translating is also quite difficult depending on how complex the original text is and any unusual terms or kanji that show up.

i'd go darker
the stark difference with the blue in the middle isn't great either. it might be best/easiest to just stick with a flat purple

>Everybody
But I do all my own high quality scans all the time.
See the following example.

I do that some times if the other version is really bad and has a fair amount of TL errors.
exhentai.org/g/1040767/ec5a281e4b/
exhentai.org/g/156449/56ead39e1f/

I rather keep the blue because I like keeping things faithful to the original text. How about this?

Attached: 58bc7a4c7cf97b84afca8079857cd284.png (427x355, 146K)

Technically a proofreader is only checking the script for errors while QC is checking for image quality, misaligned text, missing redraws, awkward translations, etc. It can also indicate whether they do their work before or after the typesetter touches it.

Also, it's nice to have someone on public relations so you have a POC for readers to submit feedback/questions to or recruits to contact. Having someone do just that is silly, but it's still a good idea for someone to take the position, especially if part of the group wants to stay anonymous and not deal with that shit.

Why are you using two colors when the original is a gradient with at least six distinct colors? Also, it falls off way too quickly. The center part should be much harder.

Honestly, because I have no idea what I'm doing. I haven't done much gradient glows and so far I haven't received much help on how to properly do this without just using two colors.

bump for help

The glow overlaps the text, just use a gradient overlay.

I redraw for a group and translate a little for my own projects and I'd say translating takes more out of me. There's a big difference between understanding the original text and being able to express it again in competent English that doesn't read like a grade-schooler's essay. But I've been using Photoshop to draw for years before, so redrawing is almost as mindless as typesetting to me.

How do I rip from here?

Attached: Screenshot_20190602-033727.png (720x1280, 352K)

The pics clearly get saved somewhere I just can't tell where

Attached: Screenshot_20190602-035849.png (720x1280, 784K)

The app name might help. If it's mobile only I'm not even going to bother though.

Manga king

mangaking.jp/indies/#about
To add in this has a bunch of horror related stuff I cant find anywhere else

Hi

He said
>jajajajaja
he'll probably translate you a whole volume for a meal.

ok

>Heisig makes remembering Kanji HARDER
Then you're doing it wrong. 1-2 hours a day with Heisig's method and you'll know em all in 2 months.

OP should stop wasting his time and actually study instead.

>wasting time with remembering individual kanji
lmao

Or you could spend that 2 months doing something useful.

I actually learned the language so I don't have that problem, sorry. Back when I was starting I used KanjiTomo a lot though, and I found it pretty helpful. I recommend trying that if other OCRs are failing you. Really though, you should try to get off the OCR crutch as fast as possible. If you really want to read Japanese you need to start a consistent kanji and vocab study regimen. It doesn't really matter what, Wanikani, Heisig, Anki decks, just make sure you stick to it.

Maybe this is the case if your literacy level is very low, but for someone reasonably literate in the language there's never a need for OCR. Kanji have semantic and phonetic components which often indicate the reading and meaning, and combined with the other kanji / kana forming the word and the overall context most of the time that should be all you need to make a good guess as to the reading and meaning of a word. If there's a word I'm not sure of I just make a guess and double check in a dictionary (I like weblio.jp and jisho for J-J and J-E respectively).

If all else fails, go to jisho and click on the "radicals" button and just enter the components until you find the kanji you're looking for.

Attached: 1510719043189.jpg (432x328, 90K)

I "translate" manga. That is, I just write whatever I think is written there, because I don't even know Japanese. Nobody has called me out on it yet, everybody thanks me for "translating" and they don't even realize what they're reading is bullshit. I mean, people are ready to thank some pinoys for some unreadable garbage like the early chapters of that slime isekai, they will eat everything lol.
Pic related. It's the smug face I get when reading people's comments, on how they are grateful for my work. Lol.

Attached: 1469111702324.png (780x720, 162K)

based isekaitrash translator

Turns out I can see the folder where images are downloaded but I can't open the jpg's

Attached: Screenshot_20190602-060829.png (720x1280, 705K)

Images won't load for me. Is it because of the emulator or the app is region locked?

They're obviously encrypted.

Its either because the app doesn't work at all in emulators or it's because most emus have root by default (I had to use a hiding app on my phone)

Found out, they're application/octet-stream format

It was a region lock. Works fine on a Japanese VPN.

The app downloads zip files with a password. Requests to mkapi.jungleapi.net/3/Mangas/downloadChapter.json with chapter_id parameter in the POST field and x-mkapi-app headers return the chapter data. The "data" url is the zip file, for example data=mkfile.jungleapi.net/1/file/book/chapter_prod/52422/0af83bf355b9084daaa444b47b60748374135c9009acd02565431348e170bc0c. There is a password field in the response but it's base64 encoded and encrypted probably. I don't have time to reverse this at the moment but it doesn't seem too hard, you only need to find how the zip password is obtained, there probably is a hardcoded key in the app that decrypts chapter passwords.

Sure. But in addition to that getting tanks and especially individual magazines is bothersome. Unless you do a lot of stuff shipping costs will usually be higher than the product itself. And ordering is a pain in the ass when you don't speak moon

I zoomed that in post-detection

It's funny because they take donations and the sniper doesn't, so they're improving the scanlation scene and fucking over Hatigarm's income.

KanjiTomo and Capture2Text.

I really only bring out google's OCR for calligraphy and handwriting which, shockingly, it gets right quite often.

Alphapolis used to upload chapters to their website at 500x711. I'd absolutely take a mag scan over that shit.

Great thing about capture2text: grabs fucking everything at once, in text form, AND it's rather accurate.

>inactivity rule
What?

If a group dies and doesn't release a chapter for six months, the upload restrictions on Mangadex are removed and anyone can upload their chapters there.

Upload restriction as in the group's upload restriction? What's the point?

Some groups have blocked their chapters from being uploaded to Mangadex entirely, by restricting uploads to group members only and then ignoring it.

Sniping is shameful and you should be ashamed for supporting it. Enjoy having no one to translate your manga, noobs.

Allowing yourself to be sniped is shameful

Stop asking for donations, Hati.

I mean, if you really think you can do the job better than the group translating it - at least give them a message first. Uploading out of nowhere just to fuck with people is despicable. You usually end up with the group dropping it, the sniper fucking off and no translations ever. Also for most people it's just a hobby, so naturally there will be some delays and such, I mean you've got at least 2-3 people working on it, possibly more (yes, you can do it all alone, but that's usually not the case). It's sad to see how little respect is given to the people who provide you translated manga to read for free.

>Also for most people it's just a hobby
They take donations, so I'm holding them to a higher standard. The sniper doesn't, so I'm not. It's not hard.

Well I agree, if they accept donations then they should do it properly.

>get a series completely caught up handed over by _anq for nothing in return
>fall behind and delay shit like total retards sometimes even with shit scans
Fuck Hatigram, if a one man team does a way better job than you, your "team" doesn't even deserve existing lmao

>hobby
>constantly begging money

I guess that depends on how long it's been since the last chapter. If it's been months since the last chapter I'm going to assume they're fucking dead and just take over.

Jisho is great for trying to figure out what retarded bullshit symbolism the mangaka came up with this time.

I hope manga licensing will kill all the fan translations, just how it happened to subbing.

Does anyone happen to have that Sanic image saved that was the sniper's credit page?

There's too many manga for that.

That'd be delightful. Much less work for us.

>image
English was a mistake.

That's Japanese you idiot.

Yeah, written in retarded English "phonology".

Literally impossible.
there are thousands of ongoing manga.
maybe even over 10k?

Not like groups are doing any obscure stuff anymore and all the popular stuff gets licensed now. Hell, even some more lesser known stuff is getting license.

Why are all scanlators such attention whores and drama queens?

Only the medium and big groups are.
The small ones shut up and do their thing while maybe putting a link to their paypal/patreon on the credits page without constantly bugging you to throw them money.

What about the guy whose mother has been sick for 50 years and he has to take care of her?

Medium group. Or a gray zone if you want, since he's been translating for a very long time so everyone knows him.

I'm not.

Because you only see the loud ones.

I'd add to this that google translate's kanji handwriting recognition is actually pretty good, it can usually reliable detect quick scribbles too, especially if you keep stroke order in mind
It's definitely faster than searching by radicals on jisho in my experience
OCR is also prone to small mistakes, like mixing up the 亻and忄radicals, from what I've seen

Attached: 00-9.png (869x425, 18K)

Most of the time is because money is involved. The rest is because they just are.

he's very active and scanlate a lot of manga everyday. I won't consider him "small" group.

Why do people even translate manga?

>how do translate with furigana?!?!

Please don't try translating when you can't even understand it yourself, user. These are not 'advanced' kanji.

It's the kind of procrastination that makes me feel like I'm kind of productive

for most it's to learn the language
and for a lot of those they pass the TLs off to people who typeset it. and those people do it because they want to read the manga.

Apparently you can't even read.

money

If you want to machine TL something for your own personal enjoyment, that's fine, but don't upload it somewhere just because poor translations are rarely called out unless something is obviously wrong.

Are you using a packet sniffer on PC or Android native? None for Android for that well for me

You can't stop me.

Nope, it's more complicated than that because requests use SSL. My setup is as follows:

>Vpngate for VPN
>Memu emulator
>Fiddler proxy on Windows
>Root certificate needs to be exported from Fiddler and installed on the emulated device, just copy the .cer file on the device and install it
>Go to the Wifi options, hold the touch on the network connection, go to the advanced options and set a proxy, now this is depended on the network config but the IP should be your local one of the host PC, usually 192.168.0.x, default proxy port in Fiddler is 8888, when in doubt run the android browser and try address:8888 to see if it's accessible
>Reboot the emulator